NEXUS NETWORK JOURNAL Architecture and Mathematics

Aims and Scope Vera Spinadel

The Mathematics & Design Association
Founded in 1999, the Nexus Network Journal (NNJ) José M. Paz 1131 - Florida (1602), Buenos Aires, Argentina
is a peer-reviewed journal for researchers, professionals E-mail: vspinade@fibertel.com.ar
and students engaged in the study of the application of
mathematical principles to architectural design. Its goal Igor Verner
is to present the broadest possible consideration of all The Department of Education in Technology and Science
aspects of the relationships between architecture and Technion - Israel Institute of Technology
mathematics, including landscape architecture and Haifa 32000, Israel
urban design. E-mail: [email protected]

Editorial Office Stephen R. Wassell

Editor-in-Chief Department of Mathematical Sciences
Kim Williams Sweet Briar College, Sweet Briar, Virginia 24595, USA
Corso Regina Margherita, 72 E-mail: [email protected]
10153 Turin (Torino), Italy
E-mail: [email protected] João Pedro Xavier
Faculdade de Arquitectura da Universidade do Porto
Contributing Editors Rua do Gólgota 215, 4150-755 Porto, Portugal
The Geometer’s Angle E-mail: [email protected]
Rachel Fletcher
113 Division St. Instructions for Authors
Great Barrington, MA 01230, USA Authorship
E-mail: rfl[email protected] Submission of a manuscript implies:
• that the work described has not been published before;
Book Reviews • that it is not under consideration for publication elsewhere;
Sylvie Duvernoy • that its publication has been approved by all coauthors,
Via Benozzo Gozzoli, 26 if any, as well as by the responsible authorities at the
50124 Firenze, Italy institute where the work has been carried out;
E-mail: [email protected] • that, if and when the manuscript is accepted for
publication, the authors agree to automatically
Corresponding Editors transfer the copyright to the publisher; and
Alessandra Capanna • that the manuscript will not be published elsewhere
Via della Bufalotta 67 in any language without the consent of the copyright
00139 Roma Italy holder.
E-mail: [email protected] Exceptions of the above have to be discussed before
the manuscript is processed. The manuscript should be
Tomás García Salgado written in English.
Palacio de Versalles # 200
Col. Lomas Reforma, c.p. 11930 Submission of the Manuscript
México D.F., Mexico Material should be sent to Kim Williams
E-mail: [email protected] via e-mail to: [email protected]
or via regular mail to: Kim Williams Books,
Robert Kirkbride Corso Regina Margherita, 72,
studio ‘patafisico 10153 Turin (Torino), Italy
12 West 29 #2
New York, NY 10001, USA Please include a cover sheet with name of author(s), title or
E-mail: [email protected] profession (if applicable), physical address, e-mail address,
abstract, and key word list.
Andrew I-Kang Li Contributions will be accepted for consideration to the
The Chinese University of Hong Kong following sections in the journal: research articles, didactics,
Shatin, N.T. viewpoints, book reviews, conference and exhibits reports.
Hong Kong S.A.R. China
E-mail: [email protected] Final PDF files
Authors receive a pdf file of their contribution in its final
Michael J. Ostwald form. Orders for additional printed reprints must be placed
School of Architecture and Built Environment with the Publisher when returning the corrected proofs.
Faculty of Engineering and Built Environment Delayed reprint orders are treated as special orders, for which
University of Newcastle charges are appreciably higher. Reprints are not to be sold.
New South Wales, Australia 2308 Articles will be freely accessible on our online platform
E-mail: [email protected] SpringerLink two years after the year of publication.
Nexus Network Journal

RECALLING EERO SAARINEN

1910-2010

Kim Williams and J. M. Rees, Editors

VOLUME 12, NUMBER 2

Summer 2010
Nexus Network Journal
Vol. 12
No. 2
Pp. 163-362
ISSN 1590-5896
CONTENTS
Letter from the Editors
163 J. M. REES and KIM WILLIAMS
Recalling Eero Saarinen 1910-2010
167 ROBERT OSSERMAN. How the Gateway Arch Got its Shape
191 DAVID M. FOXE. Saarinen’s Shell Game: Tensions, Structures, and
Sounds at MIT
213 OZAYR SALOOJEE. The Next Largest Thing: The Spatial Dimensions of
Liturgy in Eliel and Eero Saarinen’s Christ Church Lutheran,
Minneapolis
239 LUISA CONSIGLIERI and VICTOR CONSIGLIERI. Morphocontinuity in
the work of Eero Saarinen
249 TYLER SPRAGUE. Eero Saarinen, Eduardo Catalano and the Influence of
Matthew Nowicki: A Challenge to Form and Function
259 RACHEL FLETCHER. Eero Saarinen's North Christian Church in
Columbus, Indiana
Other Research
271 R. BALASUBRAMANIAM. On the Modular Design of Mughal Riverfront
Funerary Gardens
287 MARYAM ASHKAN and YAHAYA AHMAD. Discontinuous Double-shell
Domes through Islamic eras in the Middle East and Central Asia:
History, Morphology, Typologies, Geometry, and Construction
321 GIULIO MAGLI. At the Other End of the Sun’s Path: A New
Interpretation of Machu Picchu
343 TESSA MORRISON. The Body, the Temple and the Newtonian Man
Conundrum
Book Review
353 ARELI MARINA. The Symbol at Your Door: Number and Geometry in
Religious Architecture of the Greek and Latin Middle Ages by Nigel
Hiscock
Conference Report
357 EIR GRYTLI. Architecture and Mathematics. A seminar to celebrate
Professor emeritus Staale Sinding-Larsen’s 80th birthday
Erratum
361 TOMÁS GARCÍA-SALGADO. Erratum to: The Sunlight Effect of the
Kukulcán Pyramid or The History of a Line
 LETTER FROM THE EDITOR
Eero Saarinen changed my life in a building. Since I am an architect I am
susceptible to such things, and they are difficult to write about and conventionally
excluded from our reasoned discourse about architecture. However; when Kim
Williams announced this project – a collection of essays honoring Saarinen on the
centennial of his birth – I jumped at the chance to assist. I cherish Eero Saarinen’s
work and hope to contribute, in some small way, to public conversation surrounding
the built works.
Projects such as this special section of the Nexus Network Journal for
Architecture and Mathematics call for an approach that is both bottom-up and top-
down. Bottom-up, because one must always work within authors’ particular interests;
top-down, because the particular contributions must be integrated into a networked
whole. The challenge is to craft a collection of given parts so that the assembly hangs
together and reads separately. Surprisingly as this project developed, the top-down
and bottom-up approaches morphed and then melded into the collection before you.
In top-down mode I was hoping for something relating architecture and music in
Saarinen’s work. It is a commonplace of architectural commentary to ally these arts
since Matila Ghyka borrowed von Schilling’s dictum “architecture is frozen music.”
As in most common notions, there is more than a grain of truth in the general
observation, but general observations only go so far. Besides, there is a fluid quality to
Saarinen’s mature work that denies any kind of “petrified” form. Luckily David Foxe
stepped forward with an essay entitled “Saarinen’s Shell Game: Tensions, Structures,
and Sounds at MIT”, featuring a nuanced reading of Eero Saarinen’s buildings at
MIT. Based on the performance of music in two very different structures, Foxe's
conclusion supports the complimentary roles of sound and structure which fosters
both “external clarity and internal discovery.”
Rachel Fletcher's contribution on Eero Saarinen's North Christian Church
provides a different point-counterpoint. Juxtaposing quotes from Aline Saarinen's
book of Eero's thoughts and drawings of the church, often superimposed with
diagrams describing the architectural geometry, Fletcher develops the “merits of
geometrical proportion.” Most interesting however, is her conclusion that Saarinen
responds “organically to the situation at hand.”
Designing this issue on Eero Saarinen, I also hoped for an essay leading “where
angels fear to tread.” I am referring to the relationship between father and son: both
designers, both visionaries, both strong personalities. Discussing this relationship,
Susan Saarinen recounted a telling story. When Eliel Saarinen and Eero Saarinen
decided to enter the competition for the Jefferson Memorial separately, the studio
they shared was divided in half with a wall of blankets. Father and son actually used
separate entrances for the duration of the competition phase of the project. The well
known but often misinterpreted end of the story is also telling. The telegram
notifying the winner was addressed to “E. Saarinen.” It was the family that assumed it
was addressed to Eliel.
These stories speak to competitive threads in the relationship between old-world
father and new-world son. This is to be expected and indeed, who is surprised? But,
it is not the whole story. Clearly the son defined himself in contradiction to the
father, as sons are wont to do. Equally obvious is that there would have been no Eero

Nexus Network Journal 12 (2010) 163–166 Nexus Network Journal – Vol.12, No. 2, 2010 163
DOI 10.1007/s00004-010-0035-3; published online 27 May 2010
© 2010 Kim Williams Books, Turin
 without Eliel. In this often told tale there are rivalries (the Jefferson Memorial
competition, the General Motors site plan); there is the knowledge transfer between
mentor and mentored (how to enter and win a competition, how to make “no small
plans”) and then there is the dialog between masterful equals. We will never be privy
to the actual details of family dynamics yet we can conjecture some of this
conversation in the built form of Christ Church Lutheran. This is exactly what Ozayr
Saloojee does in his essay on “The Next Largest Thing: The Spatial Dimensions of
Liturgy in Eliel and Eero Saarinen’s Christ Church Lutheran, Minneapolis.” His
conclusion, in part, is “as a tribute to Eliel Saarinen with the education wing, Eero
has managed to reflect his own attitude about architecture in a way that honors his
father, but also his own vision.” There is no higher compliment between two makers;
father and son shine in the twice-reflected glow of a strangely symmetrical project.
The essay by Luisa and Victor Consiglieri, “Morphocontinuity in the work of
Eero Saarinen,” also speaks to strange symmetries: between form and structure,
architecture and urban environment, inner and outer, visual perception and
emotional movement. Their project, to give such notions a mathematical expression
while retaining the messy complexity of agency, reminds me of the work of George
David Birkhoff in Aesthetic Measure. Even though the effort is less than perfectly
consonant (what discursive piece ever is?) the motivation – to join mathematics and
biology and architecture – is uniquely suited to the Nexus Network Journal, which is
after all a journal created to host just such difficult experiments.
The essay by Tyler Sprague, “Eero Saarinen, Eduardo Catalano and the Influence
of Matthew Nowicki: A Challenge to Form and Function,” is also devoted to a kind
of experiment, this one about the way Matthew Nowicki speaks in the work of Eero
Saarinen and Eduardo Catalano. As modern architecture developed, fresh
vocabularies of form and of discourse emerged. Matthew Nowicki was in the
vanguard of these developments. Nowicki supplied instances of each in the Livestock
Pavilion (Raleigh N.C.) and in an essay titled “Origins and trends in Modern
Architecture.” Using these sources Sprague charts Nowicki's influence in works by
Catalano and Saarinen arriving, appropriately enough, at conclusions regarding “how
architectural discovery is often contextual.”
Finally, the essay by Robert Osserman, “How the Gateway Arch Got its Shape,”
is an insightful romp through the entwined histories of science, mathematics and
architecture. From a top-down editorial point of view “The geometry of the Gateway
Arch” was exactly the essay I had already decided not to accept. As a native of the
state in which the arch stands, I had read too many comments like “its shape is a (fill
in the blank) catenary.” In fact, much of the excitement in this project for me was in
rediscovering other masterworks by Eero Saarinen: John Deere Headquarters,
Milwaukee War Memorial. I thought I knew all about the arch, having effectively
grown up with it. Boy was I wrong. We hold in special regard those whose works
expose us to our own unfounded prejudice. This is what Professor Osserman’s essay
did for me – not only for the geometry of the arch but also for Galileo (he really did
understand the difference between a parabolas and catenaries) and Robert Hooke
(who was not closed-minded in the arrogant sort of way like I had always assumed).
Beyond simply correcting my prejudicial ignorance Bob Osserman’s essay couples a
word picture of the arch’s form – flattened catenary – with a mathematical formalism,
effectively joining an intuitive understanding with a technical description in a
scrupulous presentation that takes care to call reader’s attention to open questions.

164 Kim Williams – Letter from the Editor

This essay is altogether satisfying and I can only hope that it, along with all the other
works in this section devoted to Eero Saarinen on the centennial of his birth, find
wide readership.
At the end of this process I want to thank all the essayists for their hard work and
express my gratitude to Kim Williams for picking up the ball. Regarding Saarinen, I
am again reminded of my debt to him. I can honestly say that he has changed my life
twice: once when I was a student of architecture, through the TWA terminal, and
then again as a more mature practitioner reviewing a life in design well lived, cut
tragically short. Two things Eero Saarinen repeatedly demonstrated will always
inspire: the courage to approach each project as its own functional and aesthetic
problem – its own opportunity, style be damned, and that space is a rooted surface at
play in light. Thank you E.S. We owe you much.

About the guest editor

J. M. Rees, colorist, architect, writer, is editor of The Sixth Surface: Steven Holl Lights the
Nelson-Atkins Museum of Art (2007) and Urban Stories of Place (2006). He practices
architecture in Kansas City, Missouri. Exhibitions include Conceptual Play at the Greenlease
Gallery (KCMO), the virtual nomad at the Linda Hall Library for the History of Science and
Technology (KCMO), and Manhattan Miniature Golf at P.S.1 (Queens, NY). He enjoys
collaboration and believes that works of imagination are always more effective when they
respond to multiple conditions with layers of signification.
1618 Summit St.
Kansas City, Missouri 64108 USA
[email protected]
* * *
The celebration of the centennial of Eero Saarinen’s birth 1910-2010 provides a
welcome opportunity to examine the work of this great twentieth-century master
builder. Other research included in this issue takes us to different parts of the world
and different epochs, and demonstrates a link between the ancient and modern
attempts to unite architecture and mathematics.
Indian professor R. Balasubramaniam, who sadly passed away before he could see
this issue in print, studied the two significant funerary gardens of the Mughal period.
In “On the Modular Design of Mughal Riverfront Funerary Gardens”, he proves that
the modular designs of these Mughal funerary gardens were based on Arthasastra
units, and sets forth a novel mathematical canon to analyze the dimensions of
Mughal architecture.
In “Discontinuous Double-shell Domes through Islamic eras in the Middle East
and Central Asia: History, Morphology, Typologies, Geometry, and Construction,”
Maryam Ashkan and Yahaya Ahmad present a developed geometric approach for
deriving the typologies and geometries of discontinuous double-shell domes in
Islamic architecture. Common geometric attributes are created using a corpus of
twenty one domes that were built in the Middle East and Central Asia, beginning
from the early through to the late Islamic periods.

Nexus Network Journal – Vol.12, No. 2, 2010 165

 Giulio Magli takes us to Peru during the age of the Incas. In “At the Other End
of the Sun’s Path: A New Interpretation of Machu Picchu,” he challenges the
standard interpreation of Machu Picchu by means of a critical reappraisal of existing
sources and a re-analysis of existing evidence, leading to a new interpretation of this
mysterious city.
Tessa Morrison examines “The Body, the Temple and the Newtonian Man
Conundrum”, a fascinating look at a little-known aspect of Isaac Newton: his interest
in architecture, and especially, his knowledge of Vitruvius.
This issue is rounded out with a book review and a conference report. Areli
Marina reviews Nigel Hiscock’s The Symbol at Your Door: Number and Geometry
in Religious Architecture of the Greek and Latin Middle Ages. Eir Grytli reports on a
conference held in Trondheim, Norway, in honor of Professor Emeritus Staale
Sinding-Larsen’s eightieth birthday.
Finally, I am very proud to announce that the Nexus Network Journal has been
accepted into Thomson’s ISI database for scholarly journals, and will be listed in the
Art & Humanities Citation Index from 2009 forward. The Arts & Humanities
Citation Index is a multidisciplinary index to the journal literature of the arts and
humanities. It fully covers 1,160 of the world's leading arts and humanities journals,
and also indexes individually selected, relevant items from over 6,800 major science
and social science journals. The NNJ will receive an impact factor rating starting in
2011. A special thank you to all those who have contributed to the NNJ, from its
founding in 1999 to the present, for establishing the high standard of scholarship that
characterizes the journal. I am very grateful to you all.
Whether your own interests are historic or modern, you’ll find fascinating reading
in this Summer 2010 issue of the Nexus Network Journal.

166 Kim Williams – Letter from the Editor

 Robert Osserman Research
Robert Osserman How the Gateway Arch
Mathematical Sciences Research
Institute Got its Shape
17 Gauss Way
Abstract. Robert Osserman examines Eero Saarinen’s
Berkeley, CA 94720 USA
Gateway Arch in St. Louis in order to shed light on what
[email protected]
its exact shape is, why it is that shape, and whether the
Keywords: Eero Saarinen, Gateway various decisions made during its design were based on
Arch, catenary, parabola, weighted aesthetic or structural considerations. Research included
catenary, Robert Hooke, Galileo discussions with engineers and architects who worked
Galilei, history of mechanics with Saarinen on the project. The paper concludes by
noting some questions that are still unanswered.

Introduction
Much has been written on the subject
of Eero Saarinen’s most widely known
creation and architectural landmark, the
Gateway Arch in St. Louis. Nevertheless, it
is difficult to find complete and
authoritative answers to some of the most
basic questions:
1. What is the shape of the Arch?
2. Why is it the shape that it is?
3. Of the various decisions that had to
made, which were based on esthetic,
and which on structural considerations?
These questions raise issues of a purely
mathematical nature that are of
independent interest. A number of them
have been treated elsewhere [Osserman
Fig. 1. Eero Saarinen’s Gateway Arch, St.
2010] and will be referred to as needed.
Louis. Photo courtesy of Historic American
Engineering Record, Library of Congress, Here we focus on the basic questions that
Prints and Photograph Division are our central concern.

The Shape of the Arch

The terms most often used in describing the shape of the Gateway Arch are parabola,
catenary, and weighted catenary. We shall discuss each of those terms in some detail later
on, but let us note that they all represent mathematical idealizations of physical
phenomena. A parabola, as Galileo demonstrated, is the shape of a path traversed by a
projectile subjected only to its initial impetus together with the force of gravity, in the
absence of air resistance. A catenary is the shape assumed by a hanging chain or a flexible
cord of uniform density. A weighted catenary is the shape assumed by a hanging chain
whose links vary in size or weight, or by a flexible cord of variable width, or variable
density material.1

Nexus Network Journal 12 (2010) 167–189 Nexus Network Journal – Vol.12, No. 2, 2010 167
DOI 10.1007/s00004-010-0030-8; published online 5 May 2010
© 2010 Kim Williams Books, Turin
 An equally basic feature of the shape of the Arch is what we shall call, borrowing a
term from computer, television, and movie screens, its “aspect ratio”: that is, the ratio of
width to height of the inside dimensions of the smallest picture frame that can hold the
full frontal view of the Arch; or, in architectural drawing terms, a front elevation of the
Arch. Aspect ratios are usually expressed in a form such as “3-to-2” or “3:2.”
An occasional source of confusion stems from the fact that both the parabola and the
catenary extend to infinity. In both cases there is, up to scaling, a single curve, but arcs of
a single parabola or catenary can look very different from each other, depending on the
part that one chooses, and the corresponding aspect ratio of the given arc. In fact, in both
cases, one may choose arcs with any aspect ratio one wishes. Fig. 2 illustrates how to
obtain a variety of aspect ratios from a single catenary.

Fig. 2. Aspect ratios for a catenary

A further potential source of confusion is that Saarinen changed the aspect ratio of his
Arch design between the time that he originally won the competition, and the final
design used in the construction. The original design was for an arch that was 630 feet
wide and 590 feet high, whose aspect ratio was therefore approximately 21:20, or just shy
of square. By the time the arch was built, Saarinen had kept the original width, but raised
the height to 630 feet, exactly matching its width. According to architect Bruce
Detmers,2 who worked with Saarinen on the Arch, it was typical of Saarinen to favor
simple geometric figures, such as a square, and therefore not surprising that he would
modify his design to obtain the 1:1 aspect ratio. According to other sources, the principal
reason that Saarinen increased the height of the arch was that between the time that he
won the original competition in 1948, and the beginning of construction in the early
1960s, new and higher buildings had been added to the surrounding landscape, which
forced Saarinen to modify his own plans so that the Arch would clearly dominate its

168 Robert Osserman – How the Gateway Arch Got its Shape
surroundings. It is undoubtedly the case that both of these factors played a role in his
final choice of the exact 1:1 aspect ratio.
It may be worth noting that the 1:1 aspect ratio is a very old tradition in architecture.
In medieval times it was known as ad quadratum. The original 1386 specifications for
the cathedral of Milan is an example (cf. [Heyman 1999: 19]).
The contrast between the shape of a parabola and a catenary is clear if we choose
segments of each with a 1:1 aspect ratio (fig. 3).
In fig. 3, the parabola is the pointier inner curve and the catenary the rounder outer
curve.

Fig. 3. Parabola and Catenary with aspect ratio 1:1

History of the problem
Commemorative arches designed to be lasting monuments date back thousands of
years, with a number still standing from the days of ancient Rome. However, it does not
appear to have been until 1675 that the question was raised – at least in print – of what
shape an arch should be, from a mathematical and structural point of view. At that time,
one of the leading scientists of the day,3 Robert Hooke, provided his answer to the
question in the form of an anagram of the Latin phrase: ut pendet continuum flexile, sic
stabit contiguum rigidum inversum, or, “as hangs the flexible line, so but inverted will
stand the rigid arch.” Hooke was famous for his ability to devise physical demonstrations
and experiments illustrating both old and newly discovered scientific principles, and was

Nexus Network Journal – Vol.12, No. 2, 2010 169

specifically commissioned to do that for the recently created Royal Society. In the case of
the ideal shape of an arch, he realized that what one wants is that the slope of the arch at
each point exactly match the combined horizontal and vertical forces acting on that part
of the arch – the vertical component being due to gravity from the weight of the portion
of the arch lying above the point, and the horizontal force being simply transmitted
unchanged along the arch. At the apex of the arch, there is no weight above it, hence no
vertical component, and the force is simply the horizontal one of the two sides of the
arch leaning against each other. As a result, the slope at that point is zero. As one moves
down along the arch, the vertical force keeps increasing, and the slope of the arch must
increase accordingly.
One simple consequence of this general reasoning is that at the base of the arch, the
horizontal force continues to be present – a fact well known to builders who created
flying buttresses and other devices to counter that force – and as a consequence, the
bottom of the arch should never be strictly vertical, but rather, angled outwards. As it
happens, probably not by chance, Hooke recorded his dictum at the time when both he
and Christopher Wren were among the chief architects in charge of surveying the
damage, and of rebuilding London after the disastrous fire of 1666. Hooke shared his
insight with Wren, who immediately applied it to his design of St. Paul’s cathedral,
which features an interior dome that is the first to be angled out at its base, rather than
vertical (cf. [Heyman 1999: 40-41]).
But the main thrust of Hooke’s observation was that the exact shape of an ideal arch
could be obtained by simply hanging a chain (or “flexible line”) and recording the form
that it takes. The equilibrium position would be determined by a balance of vertical and
horizontal forces that are the mirror image of what they would be in the case of the arch,
with the role of gravity reversed: the vertical component at each point is determined by
the weight of the chain below it that needs to be supported, while the horizontal
component is simply transmitted unchanged along the chain.
Left unanswered by Hooke is the obvious mathematical question raised by his purely
physical solution to the ideal shape of an arch: “what is the equation of the curve that is
formed by a hanging chain?” Wherever this subject is mentioned in print, one is likely to
find a statement to the effect that Galileo was the first to raise the question, and that he
answered incorrectly that the shape that the chain would take was a parabola. Both of
those statements merit a closer look.
Galileo’s discussion of the shape of a hanging chain appears in print in his last, and in
the opinion of many, his best, book, Two New Sciences. The book takes the form of four
days of dialogue on a wide array of subjects. The first day introduces the first “new
science”: the analysis of the strength of beams and columns from a scientific and
mathematical point of view. It is often considered as the incubator of the modern fields
of structural engineering and strength of materials. During the second day, Galileo has
his chief protagonist, Salviati, propose two practical methods of drawing a parabola
[Galilei 1974: 143]. The first is by rolling a ball along a tilted metal mirror, and the
second is by hanging a fine chain, and marking points along it. In other words, he does
not pose the question of the shape taken by a chain, he simply assumes that it takes the
form of a parabola, based on both theoretical reasoning, using the decomposition of the
forces acting on it in the vertical and horizontal directions that led him to deduce the
parabolic path of a projectile, as well as on experimental evidence: drawing a parabola on

170 Robert Osserman – How the Gateway Arch Got its Shape
a vertical board, and observing the shape of a hanging chain that seems to follow the
same curve.
Few commentators note that Galileo returns to the same question on the fourth day
of his dialog, and explicitly states that the similarity is only approximate. He also explains
the analogy between the forces acting on a projectile, and those on a hanging chain. Here
is the exact quotation:
Salviati: The curvature of the line of the horizontal projectile seems to
derive from two forces, of which one … drives it horizontally, while the
other … draws it straight down. In drawing the rope, there is [likewise]
the force of that which pulls it horizontally, and also that of the weight of
the rope itself, which naturally inclines it downward. So these two kinds of
events are very similar.
…
But I wish to cause you wonder and delight together by telling you that
the cord thus hung, whether much or little stretched, bends in a line that
is very close to parabolic. The similarity is so great that if you draw a
parabolic line in a vertical plane … and then hang a little chain from the
extremities … , you will see by slackening the little chain now more and
now less, that it curves and adapts itself to the parabola; and the agreement
will be the closer, the less curved and the more extended the parabola
drawn shall be. In parabolas described with an elevation of less than 45°,
the chain will go almost exactly along the parabola [Galilei 1974: 256-7].
In short, Galileo never poses the question of precisely what shape is taken by the
hanging chain, and he contents himself with noting that it provides a close
approximation to a parabola, especially when the parabolic arc that one draws is near to
the vertex where the shape is relatively flat. For example, a 45° parabola would be
represented by the equation
y = ½ x2, -1  x  1,
which fits in a rectangle of width 2, and height ½, so that its aspect ratio is 4:1. Fig. 4
shows the parabola together with a catenary having the same aspect ratio.

Fig. 4. 45° parabola and catenary with the same aspect ratio: 4:1
It is often pointed out that unlike the case of a freely hanging cable which will take
the form of a catenary, when a cable supports a roadway that is much heavier than itself
and whose weight is distributed evenly along the horizontal rather than equally along the
cable, the cable will take the shape of a parabola. That will be the case for the cables
joining the two towers of a suspension bridge.

Nexus Network Journal – Vol.12, No. 2, 2010 171

 Fig. 5. The Golden Gate Bridge
In the case of the Golden Gate Bridge, shown in fig. 5, the central cables form curves
that make an angle of almost exactly 30° with the horizontal at the points where they
leave the towers. However, as fig. 6 makes clear (or as one can show by a direct
calculation), the parabola and approximating catenary are virtually indistinguishable in
this case.

Fig. 6. 30° parabola and corresponding catenary

The fact that the overwhelming number of references to Galileo’s treatment of this
subject declare him to have been in error, and ignore the fact that he ends up getting it
exactly right, would seem to reflect an extra degree of pleasure in catching him out in any
perceived slip-up. In fact, the first page to come up on a Google search of “Galileo
wrong” starts with an item from NASA entitled “Was Galileo wrong?” and ends with a
NOVA television program entitled “Galileo’s battle for the heavens: his big mistake.”
The first of these entries questions whether the acceleration of gravity is the same for all
varieties of materials – a fundamental tenet of science ever since Galileo, and certainly
true within the limits of measurements available to him (and many subsequent
generations). The last has to do with an actual mistake: his theory that tides were caused
by the rotation of the earth. In between these two items are a series of references to a
book entitled “Galileo was Wrong: The Church was Right,” devoted to proving that the
Earth in fact is motionless and at the center of the Universe, just as Galileo’s opponents
asserted at the time.4
Of course we now know that had Galileo posed the question of finding the exact
shape of a hanging chain, he could not have answered it, since the basic tool needed had
not yet been invented: the calculus. It was not until the end of the seventeenth century
that the problem was explicitly formulated and solved, the curve in question being
dubbed a “catenary” since it was formed by a chain – a catena.5 The brilliant but
cantankerous Bernoulli brothers, Jacob and Johann, produced a series of derivations of
the equation for the catenary along with a series of barbs at each other’s reasoning, such
as Jacob’s parody of an argument of Johann that he describes as somewhat like proving
that a pebble is stone by reasoning that “every man is stone; every pebble is a man;
therefore every pebble is stone.”6 Whatever the shortcomings of the reasoning, they did
indeed arrive at the correct conclusion. The equation for the catenary can be written in
the form we now call the hyperbolic cosine,
y = cosh x = ½ (ex + e-x)
7
in suitable choice of coordinates.

172 Robert Osserman – How the Gateway Arch Got its Shape
The design of the Arch
The Gateway Arch was only one component – although clearly the most dramatic
one – of Eero Saarinen’s winning entry in the 1947 open competition for the design of a
“Jefferson National Expansion Memorial,” dedicated to Thomas Jefferson – to his vision
of an America stretching across the continent, and his Louisiana Purchase, which roughly
doubled the size of America at the time. By a truly strange coincidence, the first use of
“catenary” recorded in the Oxford English Dictionary is by the future President Thomas
Jefferson, in a letter dated December 23, 1788 to Thomas Paine, recommending the use
of a catenary arch rather than a circular one for a 400-foot span iron bridge that Paine is
proposing to build. Jefferson would certainly have been inordinately pleased, as an
architect himself, to know that the principle of the catenary would form the basis of what
was to be the largest monument in America in the twentieth century, in honor of
Jefferson himself.
The Jefferson Memorial competition attracted virtually all the major architects in
America, including Eero’s father Eliel, a much more famous architect at the time than his
37-year old son. At least one part of the confusion involving the shape of the Gateway
Arch stems from the fact that the competition was held in two parts, first narrowing
down to the top five contenders, and then – after significant revisions were made –
choosing the winner among those five. In addition, there was a gap of 15 years between
the time that Saarinen’s proposal was chosen and when construction actually began,
during which time still further changes were adopted. And even after construction had
started, alterations were made, as one can see from different sets of blueprints in the
Saarinen archives at Yale University. Further confusion arises from a certain degree of
carelessness in terminology that can be seen, for example, in a dictionary of architecture
where a picture of an arch is accompanied by a caption referring to it as a “catenary or
parabolic arch” [Visual Dictionary of Architecture 2008: 41], not aware of, or not
interested in, fine distinctions. Whatever the reason, newspapers and architectural
journals that reported or commented on the shape of Saarinen’s winning design for the
Gateway Arch, referred to it without exception as a parabola.8 By chance, among the
letters that Saarinen received was one from H. E. Grant, head of the Department of
Engineering Drawing at Washington University in St. Louis, asking for more details
about the “parabolic arch,” since Grant wanted to use it as an example in a book on
descriptive geometry that he was writing at the time. In his response, dated March 24,
1948, Saarinen says:
The arch actually is not a true parabola, nor is it a catenary curve. We
worked at first with the mathematical shapes, but finally adjusted it
according to the eye. I suspect, however, that a catenary curve with links of
the chain graded at the same proportion as the arch thins out would come
very close to the lines upon which we settled.”9
In other documents Saarinen left considerable room for ambiguity. In 1959 he wrote:
The arch is not a true parabola, as is often stated. Instead it is a catenary
curve – the curve of a hanging chain – a curve in which the forces of thrust
are continuously kept within the center of the legs of the arch.10
Whether by “catenary curve” he means an actual catenary, or is trying to hedge a bit,
is not clear. In an unpublished transcript dating from 1958-59, he describes the Arch as
“an absolutely pure shape where the compression line goes right through the center line

Nexus Network Journal – Vol.12, No. 2, 2010 173

of the structure directly to the ground. In other words, a perfect catenary” [Eero Saarinen
2006: 343].
If one consults the Wikipedia entry for “catenary” one finds – at least at the time of
this writing (June 2009):
The Gateway Arch in Saint Louis … follows the form of an inverted
catenary. … The exact formula
y = -127.7 ft · cosh (x/127.7 ft) + 757.7 ft
is displayed inside the arch.
The equation given is indeed that of a catenary, and it has the desired dimensions of
630 feet in both height and width, but it is not the equation that was used in the
construction of the arch, and that appears on all the blueprints and specifications.11
At the Arch itself, a sheet provided by the National Park Service, under whose
jurisdiction the monument falls, is headed “Equation for the catenary curve of the
centroid of arch cross-section.” Below that is the same equation that appears on the
blueprints. We shall return to the equation shortly, but it is definitely not a catenary.
In 1983, an article entitled “Is It a Catenary? New questions about the shape of
Saarinen’s St. Louis Arch,” appeared in an architectural journal [Crosbie 1983]. The
author, Michael J. Crosbie, explained that despite widespread belief that the Gateway
Arch is shaped like a catenary, it is not. Instead, it is a “weighted catenary,” the form
taken by a “weighted chain” in which the various links are of different weight, rather
than all the same. In mathematical terms, it amounts to using a “flexible line” with a
density that is variable rather than constant. It appears to be uniformly true that any
description of the shape of the Gateway Arch that tries to be accurate will use the term
“weighted catenary” or something similar.
We are now in a position to state precisely the mathematical questions that arise.
1. How much information is conveyed by the expression “weighted catenary”? More
precisely, what range of curves may be obtained by suitable choice of a variable
density?
2. What was the basis for the choice of the particular “weighted catenary” chosen by
Saarinen? Was it primarily on esthetic or on structural/mathematical grounds?
3. The arch is designed so that its cross-sections are equilateral triangles. Again the
question whether that choice was esthetic or structural.
4. The size of the cross-sections varies according to a precise formula, with the sides of
the equilateral triangles growing from 17 feet at the apex to 54 feet at ground level.
What determined those dimensions, and what is the significance of the specific
formula used for the tapering? And once again, were those choices motivated by
esthetic or mathematical considerations?
The answer to the first question can be formulated quite simply: the term “weighted
catenary” conveys essentially no useful information; virtually any curve that one can
picture as a possible candidate12 for a weighted catenary can actually be reproduced by a
suitably weighted chain.13 Among those are, for example, a parabola and a circular arc, as
well as the catenary-like curve of the actual arch. In all of those cases one can explicitly
say how to distribute the weight so that the chain, or flexible line, will hang in exactly the
desired shape. In other words, saying that the shape of the Gateway Arch is a weighted

174 Robert Osserman – How the Gateway Arch Got its Shape
catenary is technically correct, but essentially devoid of content. Furthermore, it is not
descriptive, in that it says nothing about the actual shape of the Arch, but only about the
method used to produce it.
On the other hand, there is a very simple description of the precise shape of the Arch.
It is what we call a “flattened catenary” and consists simply of a catenary that has been
shrunk uniformly in the vertical direction by a given amount. It will have a vertical axis
of symmetry, just as the catenary does, and if we choose that axis to be the y-axis, then
the equation of a flattened catenary takes the form
y = A cosh Bx + C, (1)
where A,B > 0. Note that if we take the catenary curve y = cosh x, and scale it uniformly
up or down in size, we get the equation By = cosh Bx, while flattening it by a uniform
compression in the vertical direction would give the equation y = D cosh x, where
0 < D < 1.
In other words, equation (1) represents a catenary if and only if A = 1/B. The
constant C corresponds simply to a translation in the vertical direction, and we if choose
coordinates so that the vertex of the curve is at the origin, then C = - A. Putting all this
together, we see that by setting D = AB, equation (1) takes the form
y = D(1/B)(cosh Bx – 1) (2)
so that equation (2) represents a catenary with vertex at the origin that has been flattened
vertically by the factor D.
The centroid curve of the Gateway Arch is of exactly this form, where A and B are
specified numerical constants:
A = 68.7672, B = .0100333, (3)
so that the flattening factor D is given by
D = .69 (4)
In other words, the catenary is shrunk in the vertical direction by just under a third.
The effect is to “round it out” somewhat more at the vertex. The y-coordinate in
equation (2) for the Gateway Arch represents the vertical distance down from the vertex
of the centroid curve. The numerical values given in (3) correspond to distances
measured in feet.
It is worth noting that shrinking a curve in the vertical direction is exactly equivalent,
up to uniform scaling, to expanding it uniformly in the horizontal direction – precisely
what is often done to reformat a film done with one aspect ratio to fit it onto a wider
screen with a different aspect ratio (fig. 7).
It is also important to observe that since the flattening degree D for the Arch given by
(4) is a little over 2/3, and since the centroid curve of the Arch is intended to have an
aspect ratio of approximately 1:1, (“approximate” because it is the outer silhouette of the
Arch that is designed to have aspect ratio 1:1, and the same will not be true of the
centroid curve, as we discuss more fully below,) we must start with a portion of the
catenary having an aspect ratio of approximately 2:3 and either shrink it vertically by
about 2/3, or stretch it horizontally by about 3/2 (fig. 8).

Nexus Network Journal – Vol.12, No. 2, 2010 175

 Fig. 7. Wide-screen Mona Lisa

Fig. 8. Re-formatting to obtain a 1:1 aspect ratio

It is ironic that Saarinen has been attacked from both sides, with some commentators
criticizing the shape of the arch for being too purely mathematical to be taken seriously as
a work of art, while others take him to task for not being “pure” enough. Michael
Crosbie [1983] notes the fact that “the curve is not a perfect catenary,” adding “— a pure
form that would seem more attractive to Saarinen…”. Hélène Lipstadt is more severe:
“The purity is illusory and the explanation deceptive (if not deceitful), since he
knowingly had his engineers calculate an impure, weighted catenary” (quoted in [Merkel
2005: 244, note 36]).
Clearly, a catenary is no higher on the scale of purity than, say, a parabola or a
circular arc, which are “weighted catenaries.” The equation for the parabola is perhaps
the simplest of all equations for a curve, consisting as it does of a polynomial with a single
(quadratic) term. A circular arc is one step more complicated, involving square roots, and
hence, algebraic functions. The catenary, built out of exponential functions, which are

176 Robert Osserman – How the Gateway Arch Got its Shape
not even algebraic, but so-called transcendental functions, would seem to lie relatively far
afield in the domain of mathematical purity. Presumably what is meant by the criticism
is that the equation Saarinen used as a basis for his arch, on the suggestion of Hannskarl
Bandel, one of his engineers, suffers from being almost, but not quite, a catenary,
involving the insertion of rather arbitrary-looking numerical constants.
To criticize Saarinen for not using a perfect catenary, however, is to miss the point of
Hooke’s original insight. The catenary is the shape one obtains using a uniform chain or
“line” and is therefore the shape one wants for an arch of uniform thickness. For a
relatively small arch, uniform thickness would not be unreasonable, but for the kind of
monumental structure proposed by Saarinen for his Arch, it would be almost
unthinkable, both on esthetic and structural grounds. One would naturally want the
cross-section of the arch to be the least near the apex, where it has little or nothing to
support and where any additional weight adds to the strength requirements of everything
below. Conversely, the parts of the arch closest to the ground have the most weight to
support, and would naturally have the largest cross-sectional size. As a consequence, the
correct shape would not be the mirror image of a uniform chain, but of a
correspondingly weighted chain, as Saarinen points out in the quotation given above.
That principle was clearly understood early on. A famous illustration from 1748 in a
report by Giovanni Poleni on the cracks in the dome of St. Peter’s shows a cross-section
of the dome, together with the shape of a hanging chain that was weighted proportionally
to the loads of the corresponding sections. Poleni used the result to deduce that the
structure was basically sound, despite the cracks that had appeared [Heyman 1999: 38-
39].
In short, a “pure catenary” is the ideal shape for an arch of uniform thickness, and a
flattened catenary is the ideal shape for an arch that is tapered in a certain precise (and
elementary) manner. In the case of a weighted chain, we may express the amount of
weighting by a density function (s), where the weight of any arc of the chain is the
integral of (s) with respect to arclength s over that arc. One can show (see [Osserman
2010] that a given flattened catenary of the form y = f(x) = A cosh Bx + C is obtained
from a density function  which, when expressed in terms of x, takes the form
(s(x)) = (ay +b)/(ds/dx) = (a f(x) + b)/(1 + (f’(x))2),
where the coefficients a and b are determined by the coefficients A, B, and C of f(x), up
to an arbitrary multiple (since multiplying the density by an arbitrary positive constant
just changes the total weight, but not the shape of a hanging chain.) What this means
physically is that the weight of any portion of the chain lying over a segment of the x-
axis, given by (s)ds, is equal to (ay+b)dx over the given x-interval.
That brings us to the answer to our fourth question: how the shape of the Arch –
which is to say, the shape of its centroid curve – is related to the degree of tapering.
As we have noted, the centroid curve of the Gateway Arch is a flattened catenary
whose equation is of the form (1), where the x-axis is horizontal, the origin is at the
vertex, or highest point of the curve, and y represents the distance down from the vertex.
We denote by h the height of the curve, and by w its width. The fact that y=0 when x=0
translates to the condition C = -A, so that the equation takes the form
y= A (cosh Bx -1). (5)

Nexus Network Journal – Vol.12, No. 2, 2010 177

 Then the fact that y = h when x = ±w/2 implies
h = A(cosh Bw/2 -1). (6)
If we introduce the notation
R = cosh Bw/2 ,
then equation (6) becomes
A = h/(R – 1) (7)
while from the definition of R,
B = (2/w) cosh-1 R . (8)
The coefficients A, B that enter into the equation for the centroid curve given on the
blueprints and all the official documents for the Gateway Arch have exactly this form
(fig. 9).

Fig. 9. Official equations for the Arch.

Source: http://www.nps.gov/jeff/planyourvisit/mathematical-equation.htm
We see that their somewhat strange appearance arises simply from the condition that
the curve pass through the points (0,0) and (w/2, h). The only question is the value of
the constant R. By its definition, we have R > 1, and conversely, if we choose an arbitrary
number R > 1, then equation (5) with coefficients defined by (7) and (8) represents a
flattened catenary passing through the given points. We find further that the degree of
flattening D of the curve is given by
D = AB = (h/w)(2 cosh-1R) /(R-1) (9)
Or, equivalently,
D = (1/r)(2 cosh-1R) /(R-1) (10)
where we have denoted by r the aspect ratio of the centroid curve,
r = w/h . (11)
In other words, the flattening coefficient D is completely determined by the desired
aspect ratio r and the – so far – arbitrary constant R.

178 Robert Osserman – How the Gateway Arch Got its Shape
 We are finally in a position to spell out precisely the relation between the shape of the
Arch and the basic geometric decisions that were made. In particular we see that the
complicated-looking values of the coefficients A and B in the equation of the Arch are
direct consequences of a few simple choices. Those choices are:
(i) the centroid curve be a flattened catenary;
(ii) the orthogonal sections be equilateral triangles with one vertex pointing inward;
(iii) the triangles at the top and the bottom be 17 and 54 feet on a side respectively;
(iv) the overall height and width each be 630 feet;
(v) the constant R in equations (7) and (8) be given by R = Qb /Qv , where Qb is the
cross-sectional area of the triangle at the base, and Qv the corresponding area at
the top.
From (ii) and (iii) it follows that the cross-sectional areas at the vertex and the base
are
Qv = 125.1406, Qb = 1262.6651, (12)
so that according to (v), the value of R is a little over 10. Also from (ii) and (iii) we can
calculate the height h and width w of the centroid curve needed to make the outside
dimensions of the arch satisfy (iv), and they in turn determine the aspect ratio r.
Inserting all these quantities into the equations (7), (8), and (9), gives the numerical
values of the coefficients A and B that we noted earlier, as well as the degree D of
flattening of the curve.
The one remaining choice that must be made is a formula for the tapering from top
to bottom of the Arch. The decision there was to make the cross-sectional area Q (x,y) at
the point (x,y) of the centroid curve a linear function of y, interpolating between the
given values (12) at the top and bottom. In other words,
(vi) Q(x,y) = Qv + (Qb – Qv) y/h .
Conditions (i) – (vi) together provide a complete mathematical description of the
basic design of the Gateway Arch. Its actual construction, with an outside skin consisting
of flat stainless steel plates, requires dividing this basic shape into sections,14 each
bounded by a pair of equilateral triangles, and replacing the curved surface defined by the
equations by three flat pieces.15 The result is also completely defined mathematically, but
is an artifact of construction, intended to so closely approximate the curved surface that
one is not even aware, for example, that the inner curve of the Arch, its intrados, is not a
smooth curve, but a polygonal one. That is indeed the case, and we limit our discussion
here to the curved surface of the Arch that underlies the construction.
What remains to answer is whether the choice of the shape itself was dictated largely
by structural or by esthetic considerations. When this question was put to Bruce
Detmers, who worked with Saarinen on the Arch, his answer was unequivocal: Saarinen’s
choice was purely esthetic. According to Detmers, Saarinen worked with a large variety of
shapes, many of them made on the elaborated Hooke principle, by hanging a weighted
chain, or a piece of rubber cut out with a varying width to gain the same effect. The
resulting shapes were then used, upside down, to make sizeable models of the arch, and
Saarinen would examine the models from all angles. He ended up rejecting them all, as

Nexus Network Journal – Vol.12, No. 2, 2010 179

either too flat or too pointed, or otherwise unappealing. He then turned to the engineer
Hannskarl Bandel for further ideas, and it was Bandel who suggested the shape that
Saarinen found to his liking and that was the one actually used for the design of the Arch.
(As a parenthetical note, the key role of the engineers throughout the process, from
design to completion, is all-too-often forgotten in accounts of the Arch and its
construction.)
The question that arises, if one accepts that Saarinen chose the shape of the Arch on
purely esthetic grounds, is “What role, if any, was played by mathematics?”
Bruce Detmers notes two such roles. First, mathematics was critical as a means of
communication. It is one thing to draw a shape and to make models that are 10 or 12
feet high, but if one arrives at a pleasing design, then it is still necessary to convert that
design into specific instructions to provide those who must manufacture and assemble
the parts. There is no obvious way to make the transition from a sketch or a model that
must be scaled up by several orders of magnitude and described with absolute precision
in order to arrive at the full-sized structure. But as soon as the shape is defined by a
mathematical equation, one may build it on any scale that the available materials and
methods will allow.
Second, Detmers also alluded to Saarinen’s predilection for perfect geometric shapes
like a square or a circle. In the case of the Arch, as we have already noted, he decided that
he would like its proportions to be such that it could be exactly inscribed in a square.
An additional, and crucial role of mathematics was to calculate the ideal shape of the
fully three-dimensional structure whose central curve had the form decided on. That
comes down to a process that is the reverse of starting with a hanging chain and adopting
the form it takes in order to build an arch. Here one starts with a desirable shape, and has
to determine the degree of tapering that would be structurally ideal to go with that shape.
But as we have noted, to virtually any shape defined by a mathematical equation, one can
assign a weighting that will produce exactly that shape.
The history of the particular equation suggested by Bandel for the Arch goes back to
the mid-nineteenth century. It arose in trying to determine the ideal shape for an arch
that supports an earth bridge – a bridge formed by simply piling up dirt above the
supporting arch to make a flat road surface. The steepness of the arch at each point
should be determined by the balance of the horizontal and vertical forces at the point; the
horizontal force is just transmitted along the arch and is unchanged from point to point,
while the vertical force results from the weight of the column of dirt directly above the
point, and is therefore proportional to the height of the road above that point on the
arch. From that one can show that the ideal shape must be a flattened catenary (see, for
example [Heyman 1982: 48-49]).
In the case of the Gateway Arch, it does not support any other structure. However,
the effect of the tapering is that the weight being supported by each cross-section
increases steadily as one moves from the top to the bottom of the arch, just as it does for
an earth bridge. What emerges then is a simple answer to our second and fourth
questions regarding the shape of the Arch and the amount of tapering: the two are
inextricably entwined, as are the esthetic and structural decisions that have to be made.
The reality is that far more comes into play than these basic considerations. In any
project of the magnitude of the Gateway Arch, economic and political factors are
inevitably going to be involved. One striking example is the appearance of the George

180 Robert Osserman – How the Gateway Arch Got its Shape
Washington Bridge, with its two silvery towers of interlaced steel girders. According to
the original design, those towers were to be covered in stone in the manner of the
bridge’s famous forerunner, the Brooklyn Bridge, and the steel framework one now sees
were to be the invisible support. However, the money for the project ran out, and
generations of New Yorkers, along with millions of out-of-town visitors, have engraved in
their minds a very different-looking bridge than the one envisaged by its designers. The
Eads Bridge across the Mississippi that now forms a backdrop to the Gateway Arch was
as dramatic for its time as the George Washington Bridge across the Hudson several
generations later. In both cases, the length of the bridges and the size of the river spanned
broke new ground, and in both cases the bridges linked two adjacent states. In the case of
the Eads Bridge, connecting Missouri to Illinois, powerful political factors entered in,
affecting the final design of the bridge.16
Even on the structural level, many factors must be considered beyond the ideal shape
for supporting the weight of the Arch itself. There are many forces besides gravity that
must be taken into account, including those arising from potential gale-force winds, from
earthquakes, and from uneven expansion and contraction as the sun heats up certain
parts more than others. In addition, one must assure that the Arch is not susceptible to
vibrations at certain resonant frequencies that might lead to structural failure, as
happened in the famous case of the Tacoma Narrows Bridge.
Finally, there are many additional decisions that must be made, where esthetic and
structural decisions are intermingled. For the overall effect, the most important was
undoubtedly the choice of stainless steel for the outer surface, reflecting blue skies and
clouds during the day, shining brilliantly against a dark sky when lit up at night. And
right down to important details, such as the choice of whether to “brush” the steel
horizontally or vertically, and what method to use for attaching adjoining steel plates.
That leaves us with one remaining question to be answered – question 3 about the
choice of a triangular cross-section. The answer forms part of the subject of our final
section.
Some controversial aspects of the Gateway Arch
No project as monumental as the construction of the Gateway Arch is apt to come to
completion without arousing serious controversy. At the least, there are always matters of
taste and a ready chorus of critics whose job it is to criticize. In the particular case of the
Arch, there were four flags raised against Saarinen that are worthy of note.
The first was raised at the very outset, during the original competition: “What good is
it?” A great deal of (public) money is to be spent on a structure with no “purpose” other
than merely symbolic. (That is a criticism familiar to most mathematicians who engage
in and teach “pure” mathematics, often with no practical applications in sight.)
It should be said that the Gateway Arch was designed with the express purpose of
putting St. Louis back on the map after a long period of decline from its nineteenth-
century glory as the second largest port in America and the “Gateway to the West.” In
that capacity, it has been wildly successful, attracting millions of visitors each year, and
joining the Washington Monument and the Eiffel tower as among the most widely
known symbols of their kind.
The unqualified success of the Arch, both symbolic and practical, as a draw of both
attention and visitors to St. Louis, has pretty well answered this objection. Beyond that,

Nexus Network Journal – Vol.12, No. 2, 2010 181

many are struck by the beauty of the shape, and the issue then turns on the value of a
piece of art or architecture that is admired and gives pleasure of a purely esthetic kind.
The second controversy arose during the construction of the Arch, when serious
worries were expressed about its potential to fail and to collapse in what would have been
a major catastrophe. That conflict and its resolution have been beautifully described by
one of the principal figures involved, George Hartzog.17
The third regards our question about the choice of a triangular cross-section.
Saarinen’s original design for the Arch had a quadrilateral cross-section. Between the first
and second phase of the competition, one of his colleagues, the sculptor Carl Milles,
suggested that a triangular cross-section would be more attractive. Saarinen liked the idea
and adopted it in his final designs, but he failed to publicly credit Milles with the
suggestion, a failure that caused many hard feelings and much anger.18
The fourth controversy, and most interesting from many points of view, arose shortly
after the proposal submitted by Saarinen and his team of co-workers was awarded the
commission to build the monument. An article in the New York Herald Tribune
displayed side-by-side pictures of the design submitted by Saarinen and a poster for an
International Exposition that was planned for 1942 in Rome. The headline was “St.
Louis Arch For Memorial Called Fascist.” In fact, the Rome exposition was explicitly
billed as commemorating the twentieth anniversary of Fascism in Italy, and Mussolini’s
“March on Rome” in 1922. The poster features an arch designed by the Italian architect
Adalberto Libera that appears to be on much the same scale as that envisaged by
Saarinen, and seems to resemble it strongly in shape. Saarinen vehemently denied
knowing anything about the earlier proposed arch, which in fact was never built, due at
least in part to the advent of World War II. He further noted that a “simple form, based
on the natural laws of mathematics,” could no more be co-opted by a political movement
than the Washington Monument could be faulted for the fact that its design is that of
ancient Egyptian obelisks that were built by slaves.
Ironically, Saarinen’s defense – that the form of his Arch was a standard one that
could well have been used by anybody – has the effect of diminishing the originality of
his design. On the other hand, it provides fairly convincing evidence that he knew
nothing about the earlier design. The resemblance on the poster turns out to be purely an
artifact of the particular angle at which the Libera Arch is depicted. The actual plans
submitted by Libera show that his proposal was for a strictly semi-circular arch, with no
resemblance to Saarinen’s, other than the overall similar scale, and of course the fact that
both arrived at the same idea of a monumental arch as a symbolic gateway – in Libera’s
case, to the proposed Exposition and the associated newly rebuilt area of Rome, still
known as EUR: Esposizione Universale Roma – and for Saarinen, to the broad Western
expanse of America across the Mississippi River.
An additional irony is that when Saarinen’s design acquired renown after he won the
competition, Libera accused him of plagiarism, whereas Libera had himself been similarly
charged, because another team of Italian architects had suggested much the same design
for an entry arch to the EUR somewhat earlier than Libera.
The final irony is that even if World War II had not intervened, it is not clear that
Libera’s arch would have been built, or that if built, it would have been able to resist
collapse. In fact, a semi-circle is far from an ideal shape for a monumental arch. Serious
doubts were expressed at the time. Some of those doubts concerned the choice of

182 Robert Osserman – How the Gateway Arch Got its Shape
materials and structural details of Libera’s proposal. Others are intimately connected with
Robert Hooke’s original insight connecting the shape of an arch with that of a hanging
chain. As we noted earlier, one consequence, put to use immediately by Christopher
Wren, is that one wants the base of the arch to hit the ground at some angle, and not
vertically, as is the case of a semi-circle. The larger the size of the arch, the more vital
such considerations become. That is true for an arch of uniform thickness and the
corresponding chain of uniform weight. But it is equally true for a weighted chain of any
sort.
We have also noted earlier that although it may seem completely counter-intuitive, it
is possible to weight a chain in such a manner that it will hang in the form of a perfect
circular arc, provided only that it is less than a full semi-circle. That weighting will in
turn indicate the ideal tapering of the arch for maximum structural stability. However,
Libera’s original specifications for his proposed arch give no indication that he was aware
of the mathematics needed to determine the correct weighting or degree of tapering.19
Concluding remarks
The paintings and the book by Jasper Johns devoted to catenaries [2005] make it
clear that the shape of a hanging chain evokes an esthetic response quite apart from any
attempt to analyze that shape by mathematical means.
In the case of architecture, the need to combine esthetic values with mathematical
and structural principles is one that dates back to antiquity, and is certainly apparent in
the surviving Greek temples. In some cases, mathematics and esthetics were viewed as
interchangeable. As is apparent from our discussion here, in the case of the Gateway Arch
they are, if not equivalent, then at least inextricable. It is clear from his letters and his
various explanations, that Saarinen was aware at a very early stage that the structural
requirement for maximal stability translated into keeping the line of thrust aligned as
nearly as possible with the slope of the Arch, which in turn determined the relevant
shape. Having begun with that principle, and tried out the resultant shape, he turned it
around, modified the pure catenary shape “according to the eye,” and then left it to his
(superb) engineering team to apply the mathematical principles that would produce the
desired shape and satisfy the structural needs.
Viewed more broadly, the role of the architect is one that has provoked a surprisingly
wide range of responses. At one extreme, there is the assessment reported in the book
Brunelleschi’s Dome by Ross King [2000: 157-158], that both ancient and medieval
authors “assigned architecture a low place in human achievement, regarding it as an
occupation unfit for an educated man.” He cites both Cicero and Seneca in support of
that view. At the other extreme, we have the view that “it is hardly surprising that, for the
ancients, the image of the architect has demiurgic connotations” [Benvenuto 1991:
Introduction, xix]. The author, Edoardo Benvenuto, goes on to say:
In a famous dialogue of Paul Valéry20 this “nearly divine” aspect is expressed in
the following words of Phaedro to Socrates when speaking of his friend the
architect Eupalinos:21 “How marvelous, when he spoke to the workmen! There
was no trace of his difficult nightly meditation. He just gave them orders and
numbers.
(To which Socrates responds, “God does just that”.)

Nexus Network Journal – Vol.12, No. 2, 2010 183

 A twentieth-century version of this story is reported by Bruce Detmers, whose duties
as a young architect working with Saarinen on the Gateway Arch included long hours
spent translating the engineers’ formulas for Saarinen’s Arch into long columns of
numbers generated, in the absence of a modern computer, on an old Marchand
calculator borrowed from the accounting department. These numbers were then passed
on to those responsible for manufacturing the various components needed in the
construction of the Arch.
Detmers recounts one incident that may resonate as a distant echo of Phaedro and
Eupalinos. He was working away at two or three o’clock in the morning, when Saarinen
wandered in, and asked what he was doing. When Detmers explained, Saarinen’s laconic
response was, “Keep going.”
Additional remarks and open questions
We elaborate here on some of the points noted briefly in our earlier paper [2010],
and conclude with some remaining questions.
1. It is often noted that viewers of the Gateway Arch tend very strongly to see it as
being taller than wide, despite the fact that its height and width are both exactly equal at
630 feet. This effect is generally described as an optical illusion. But is it, in fact, an
illusion?
While it is true that the outside curve, or extrados, is the same width as height, one’s
eye is as likely to fix on the inside curve, or intrados, or on something in between. In
those cases, the curve one sees, or extrapolates, is in fact considerably taller than wide.
The reason is simply that the arch is much thinner at the top than at the base, and that
even further, the base width gets subtracted off twice between the outer and inner curves.
In precise numerical terms, the inner curve is 615.3 feet high and 536.1 feet wide, so that
the height is 15% greater than the width. It is likely the case that the impression of the
Arch being taller than wide is partly due to this fact and partly a true optical illusion.22
2. It is clear that Saarinen was aware of Hooke’s dictum relating the shape of a
standing arch to that of a hanging chain, and that the goal was to build an arch where the
direction of the line of thrust at each stage of the arch is as near as possible to the slope of
the arch itself. He was further aware that the principle applied whether the chain was
uniform or variable, forming a “weighted chain.” In choosing the shape of a flattened
catenary, rather than a pure catenary, he was able to satisfy two distinct desires
simultaneously – first, to arrive at a shape more pleasing esthetically to him, since is was
more rounded at the top, and second, to have the corresponding arch satisfy the
structural needs of greater thickness and strength toward the bottom, while thinner and
more slender near the top. If one compares both the pure and the flattened catenaries
with a parabola, one sees that the parabola is even pointier at the top than the catenary
(fig. 10).

184 Robert Osserman – How the Gateway Arch Got its Shape
 Fig. 10. Parabola, Catenary, and Flattened Catenary with Aspect Ratio 1:1
It is fortunate (or is it something deeper?) that Saarinen found the parabola less
pleasing from a purely esthetic viewpoint, since following Hooke’s dictum in that case
would require greatest weight at the vertex, and therefore an arch that was thickest at the
top, and slimming down toward the bottom—hardly what one would want from a
structural point of view. In fact, it is clear that if one starts with a uniform chain that
hangs in the shape of the catenary, then to make it pointier at the vertex one would want
to increase the weight in the middle, while to make it rounder or flatter near the vertex,
one would increase the weight near the ends.
Fig. 10 shows all three curves together, each with aspect ratio 1:1. The parabola is the
innermost, pointiest curve, and the catenary the intermediate one, while the outer curve
is a flattened catenary with the same degree D of flattening as that used in the Arch
3. There are at least two ways to arrive mathematically at the shape of a hanging
chain. The first is the one we have been invoking, in which the object is to keep the
thrust line always in the direction of the chain. The other is by using the method known
as the calculus of variations to determine the shape of the chain that will give the lowest

Nexus Network Journal – Vol.12, No. 2, 2010 185

center of gravity. Both methods yield the same result – that the chain should be in the
shape of a (possibly weighted) catenary – and both methods make good sense physically.
However, a strange consequence is that in the case of the arch, the shape of the catenary,
while best possible from the point of view of the thrust line, is in a certain sense the worst
possible structurally, since it will have the highest center of gravity, and therefore be the
least stable. Partly to counter this paradoxical fact, and partly for other structural reasons,
the engineers responsible for the Gateway Arch decided to fill in the space between the
outer stainless steel surface and the inner carbon steel surface with concrete in the bottom
300 feet of the Arch. The weight of the concrete alone in the bottom 300 feet is more
than double the total weight of all the steel in the Arch, while the underground
foundation weighs more than double that of the above-ground concrete. The effect is to
considerably lower the center of gravity, in addition to giving extra strength to the
bottom half of the Arch, providing resistance to cantilever and torque effects from the
wind, and limiting vibrations.
4. The fact that the method of construction changes abruptly roughly halfway up the
Arch means that if one models the Arch as a weighted catenary, then the weighting will
undergo a singularity at the point corresponding to the end of the concrete filling. It is as
if the first 300 feet on both sides represent a pedestal, and the Arch proper sits on top of
that. From the outside, of course, one sees a continuous curved surface, as well as a
continuous variation in the size of the cross-sections. The principal question to which I
have not found an answer is the reason for choosing the variation in the size of the cross-
section in a way that makes the areas vary linearly with height. But the abrupt change in
the weighting at the 300-foot level means that no single formula can be optimal for both
of the parts – those lying above and below that level. Clearly the Arch is designed with a
sizeable margin of safety beyond what the simple choice of a curve and weighting would
provide. One would need such a margin for safety if for no other reason than the
necessity for the two legs to stand on their own, like a pair of giant leaning towers, during
the greater part of the construction. Only when they were over 500 feet high did they
reach the point where they could be joined in mutual support.
5. The final questions to which I do not know the answer are: 1) Was the choice of
54 feet for the dimension of the triangular cross-section at the base of the monument an
esthetic or a structural decision? (The dimension of 17 feet at the top, I have been told,
was simply the smallest felt to be possible, allowing for the construction of an observation
platform where visitors can stand up and admire the view from the top through the
windows built into the Arch.); 2) What is the basis of the choice of the ratio Qb /Qv for
the parameter R that enters into the basic equations (7) – (10)? I hope to return to these
questions at a later occasion.
Acknowledgments
The most interesting and accurate mathematical discussions of the Gateway Arch that I
am familiar with are those by William V. Thayer [1984], and a chapter (in Portuguese)
of a volume of computational activities and projects designed to highlight applications of
the calculus [Figueiredo et al. 2005]. I am particularly indebted to Bill Thayer for
providing a wealth of documents as well as a number of significant leads in my
investigation of this problem; also to Charles Redfield, one of the main engineers who
worked on the Arch and who provided me with a number of relevant documents, Bruce
Detmers, an architect who worked with Saarinen on the Arch, John Ochsendorf for
enlightening discussions of his paper [Block et al. 2006] on arches and related matters,

186 Robert Osserman – How the Gateway Arch Got its Shape
Jacques Heyman for prompt responses to a number of questions arising from his many
books and articles about arches, Hélène Lipstadt and Jack Rees who offered a number of
further valuable leads, and Robert Moore of the National Park Service, historian of the
Gateway Arch. I also thank Jennifer Clark, archivist at the Old Court House in St. Louis,
and Laura Tatum at the Saarinen Archive at Yale University
Notes
1. One can also treat a cord from which are hung a series of literal weights, in which case the cord
will take a kind of polygonal shape. However, we shall not consider that case here, since it is
not too relevant to the curve of the Gateway Arch.
2. This and all further references to statements by Bruce Detmers come from conversations with
him in March 2008.
3. Robert Hooke had the great misfortune to be a contemporary of Isaac Newton, and on two
counts. First of all, Newton’s brilliance and monumental achievements simply overshadowed
any and all potential rivals among scientists. But perhaps even more so, because Newton
developed one of his notorious lifetime grudges against Hooke, and did everything possible –
which was quite a bit – to denigrate and belittle Hooke and his achievements.
4. See also [Huerta 2006]. The issue here is that although Galileo is precisely correct in his great
insight that one cannot scale up any structure an arbitrarily large amount, since the weight goes
up as the cube of the scaling factor and the cross-sectional area of the supports – say columns
for a building, legs of an animal, or trunk of a tree – increases by only the square, nevertheless
he is “wrong” not to note that on the scale of normal buildings, strength is not the key issue,
but stability, as in Hooke’s dictum, and in that case what counts is the geometrical shape,
invariant under scaling.
5. Strictly speaking, the term used was the Latin word, catenaria, and the English word “catenary”
did not occur until somewhat later; this will be discussed shortly. An excellent and detailed
history of the “catenary problem” is given by Clifford Truesdell in The Rational Mechanics of
Flexible or Elastic Bodies, 1638-1788; Introduction to Leonhardi Euleri Opera Omnia, Vol. X
et XI Seriei Secundae, Lausanne, Orell Füssli Turici 1960; see also further references in
footnote 7 below. (Surprisingly, Truesdell also fails to cite Galileo’s correct description of the
catenary as an approximation to a parabola.)
6. “Tout homme est pierre; tout caillou est homme; donc tout caillou est pierre.” From a letter of
August 11, 1697; Jacob Bernoulli, Opera, pp. 829-839; reprinted in Die Streitschriften von
Jacob und Johann Bernoulli. Birkhäuser Verlag 1991, pp. 356-364 (see p. 361). See also pp. 1-
114 of this volume for an overview in English of the contents, and pp. 117-122 for a list of
“The Polemic Writings of Jacob and Johann Bernoulli on the Calculus of Variations” that are
reproduced in the book.
7. Other names that should be mentioned in this brief history are Philippe de la Hire and David
Gregory. De la Hire’s Traité de méchanique from 1695 addresses the question of how to
distribute weights along a chain to attain “a figure curved the way you wish it to be”
(proposition 123), and he makes explicit the connection between arches and variably weighted
chains. David Gregory discovered independently the relation between weighted chains and
arches and described them in a letter later published in the Philosophical Transactions in
August 1697. For more on both of these major contributions, see [Benvenuto 1991: 321-329].
Another good historical reference is [Bukowski 2008] (the usual misleading statements about
Galileo are repeated here, but it is instructive to see how early they originated and were
propagated by successive generations).
8. At least that is the case for the dozens of articles, as well as congratulatory telegrams that are on
file in the Saarinen Archives at YaleUniversity.
9. Copies of Grant’s letter and Saarinen’s response are in the Saarinen archive. It is the only place
I have found where Saarinen describes his winning design with as much specificity.
10. Quoted in [Crosbie 1983].
11. Whether this formula (or, for that matter, any other formula) is displayed inside the Arch as
claimed, is something I have not been able to determine. It is not impossible, since the precise

Nexus Network Journal – Vol.12, No. 2, 2010 187

 shape of the Arch was modified so many times during the planning process, but it is unlikely,
because the decision not to use an exact catenary was made very early – long before
construction began.
12. The precise conditions for the graph of a curve to be representable as a weighted catenary are
that it be convex, and that it makes a non-zero angle with the vertical at its endpoints. See
[Osserman 2010].
13. Johann Bernoulli was already aware of the connection between arches and a chain of variable
weight in 1698; see, for example, [Benvenuto 1991: Pt. II, 327].
14. There are 71 such sections on each leg of the arch, according to the specifications.
15. This is further complicated by a note on the blueprints that for the top 41 sections, the
“exterior skin to be curved into a continuous smooth surface.” This corresponds to roughly the
upper half of the Arch, where the curvature is the greatest. It is also above the 300-foot level
where there is another transition in the mode of construction, as we discuss more fully later on;
below that level the stainless steel outer skin is backed by a layer of concrete.
16. Both of these examples are described in fascinating detail in the book The Tower and the
Bridge [Billington 1985].
17. George B. Hartzog, Jr. was Director of the National Park Service at the time. See the chapter
on the Gateway Arch in [Hartzog 1988], especially pp. 52-56.
18. See, for example, [Coir 2006]: “Milles strongly advised Saarinen to drop his initial plans to
fashion the arch with a quadrilateral section in favor of a more sculpturally pleasing triangular
section. Eero’s failure publicly to recognize Milles’s contribution to the monument angered the
Swedish sculptor, who thereafter severed contact with Saarinen” [2006: p. 41 and footnote 40,
p. 43].
19. On the other hand, some of Libera’s sketches for the arch show a radical thickening near
ground level, which would seem to indicate at least an intuitive understanding of the need for a
suitable weighting to support the shape of a circular arc. From an esthetic point of view that
thickening may be seen as producing a more dramatic effect for Libera’s arch than for
Saarinen’s, or else as far less graceful. (Perhaps both.)
20. Benvenuto is referring to Valéry’s “Eupalinos ou l’architecte” [1960: 83].
21. Eupalinos was the engineer to whom is ascribed (by Herodotus) the construction of the famous
3400 foot long tunnel on Samos, excavated from both ends, meeting in the middle.
22. A number of experiments seem to confirm that we tend to consistently overestimate lengths in
the vertical direction over equal ones that are horizontal. See for example [Wolfe et al. 2005:
967-979] and further references given there.
Bibliography
BENVENUTO, Edoardo. 1991. An Introduction to the History of Structural Mechanics. Part II:
Vaulted Structures and Elastic Systems. New York, Springer-Verlag.
BILLINGTON, David. 1985. The Tower and the Bridge. Princeton: Princeton University Press.
BLOCK Philippe, Matt DEJONG, John OCHSENDORF. 2006. As Hangs the Flexible Line:
Equilibrium of Masonry Arches. Nexus Network Journal 8, 2: 13-24.
BUKOWSKI, John. 2008. Christiaan Huygens and the Problem of the Hanging Chain. College
Mathematics Journal 39, 1: 2-11.
COIR, Mark. 2006. The Cranbrook Factor. Pp. 29-43 in Eero Saarinen, Shaping the Future. New
Haven: Yale University Press.
CROSBIE, Michael J. 1983. Is It a Catenary? New questions about the shape of Saarinen’s St. Louis
Arch. AIA Journal (June 1983): 78-79.
Eero Saarinen: Shaping the Future. 2006. Eeva-Liisa Pelkonen and Donald Albrecht, eds. New
Haven: Yale University Press.
FIGUEIREDO, Vera L.X., Margarida P. MELLO, Sandra A. SANTOS. 2005. Cálculo com Aplicações:
Atividades Computacionais e Projetos. Coleção IMECC, Textos Didáticos.
GALILEI, Galileo. 1974. Two New Sciences. Stillman Drake, trans. Madison: University of
Wisconsin Press.
HARTZOG, George B., Jr. 1988. Battling for the National Parks. Kingston (Rhode Island): Moyer
Bell.

188 Robert Osserman – How the Gateway Arch Got its Shape
HEYMAN, Jacques. 1999. The Science of Structural Engineering. London: Imperial College Press.
HEYMAN, Jacques. 1982. The Masonry Arch. Chichester: Ellis Horward Limited.
HUERTA, Santiago. 2006. Galileo was Wrong! The Geometrical Design of Masonry Arches. Nexus
Network Journal 8, 2: 25-51.
JOHNS, Jasper. 2005. Catenary. New York: Steidel Publishing.
KING, Ross. 2000. Brunelleschi’s Dome: How a Renaissance Genius Reinvented Architecture.
Penguin Books.
MERKEL, Jayne. 2005. Eero Saarinen. London: Phaidon Press.
OSSERMAN, Robert. 2010. Mathematics of the Gateway Arch. Notices of the American
Mathematical Society 57, 2: 220-229.
THAYER William V. 1984. The St. Louis Arch Problem. Module 638 of UMAP: Modules in
Undergraduate Mathematics and its Applications.
VALÉRY, Paul. 1960. Eupalinos ou l’architecte. In Oeuvres, Vol. 2, Paris: Bibliotèque de la
Pléiade/Éditions Gallimard.
The Visual Dictionary of Architecture. 2008. Lausanne: Ava Publishing.
WOLFE, Uta, Laurence T. MALONEY, and Mimi TAM. 2005. Distortion of perceived length in the
frontoparallel plane: Tests of perspective theories. Perception and Psychophysics 67, 6: 967-
979.

About the author

Robert Osserman is professor emeritus of mathematics at Stanford University and Special Projects
Director at the Mathematical Sciences Research Institute in Berkeley. His research interests within
mathematics have centered on geometric questions in a variety of fields, including complex
function theory, partial differential equations, Riemann surfaces, minimal surfaces, isoperimetric
inequalities, and ergodic theory. He has also written about interactions between mathematics and
areas of the arts and humanities, including music, visual arts, theatre, and film, as well as a book on
geometry and cosmology, Poetry of the Universe: a Mathematical Exploration of the Cosmos
(Anchor Books, 1996).

Nexus Network Journal – Vol.12, No. 2, 2010 189

 David M Foxe Research
125 Warren Street Saarinen’s Shell Game: Tensions,
Newton Centre, MA 02459,
USA Structures, and Sounds at MIT
[email protected]
Abstract. For all of the criticism of Eero Saarinen’s Kresge
Keywords: Eero Saarinen, Kresge Auditorium and MIT Chapel, they exist as a highly focused
Auditorium, MIT Chapel, moment of deliberate experimentation with geometric form
music, acoustics, religious in materials old and new which both contrasted the typical
buildings, auditorium, forms of rational modernism and resonated deeply with the
shell/membrane design, modernist quest for the incorporation of novel structures.
structural design, graphic statics, This paper explores the metaphorical and literal tensions
interdisciplinary, student life, through three dichotomies: geometrical ones with
Massachusetts Institute of implications for acoustics, programmatic ones with
Technology implications for use, and structural ones between the
appearances and actual structural actions of the architecture. I
seek to illuminate how the geometric issues of both buildings
relate to structural optimization. I also approach the
Auditorium and Chapel from the roles of having been a
performer and composer of instrumental and vocal music for
both spaces while earning degrees at MIT in architecture.
The simple act of listening defies one’s typical expectations in
both spaces, and the dichotomies of geometry, use, and
structure illuminate the relationship of sound to place in
these architectural spaces.

Prelude: Sites and Sounds

Nestled in a small stand of London Plane trees planted west of Massachusetts Avenue
stands the MIT Chapel designed by Eero Saarinen. Designed in tandem with the grassy
oval and the spherical Kresge Auditorium, Saarinen’s chapel creates a small, closed,
cylindrical form in the midst of its domed and rectilinear neighbors on the MIT campus.
The exterior form (modified in recent decades by causeway-like ramps moderating the
level change at the entry) encloses a space that is purposefully hidden. Across Kresge
Oval, the auditorium seeks a less contemplative and more forthright appearance from
Massachusetts Avenue; its curtain wall broadcasting views of the lobby since its entry is at
the grade of the broadly sloping turf.
While the chapel is relatively recent, merely a half-century old, it hearkens back to
more primal architecture concepts, focusing on simple elements of water and light.
While sited within viewing distance of a major river, Saarinen’s design includes a shallow
pool of water that functions as a reflecting pool to the exterior as well as reflecting light to
the interior. While the exterior remains a quiet, closed cylinder surrounded by still
water, the interior of the chapel is a ring of undulating brick ripples.. Provided it has not
been seasonally drained from the concrete “moat,” the water reflects light through small
horizontal curved windows up onto the brick; these ever-changing patterns of light are
the primary contact the interior space has with the outside world, the light having been
mediated by the water’s surface. For many traditions, the single overhead oculus piercing
the curved, black, night-like ceiling is an appropriate connection to divine connotations
(fig. 1).

Nexus Network Journal 12 (2010) 191–211 Nexus Network Journal – Vol.12, No. 2, 2010 191
DOI 10.1007/s00004-010-0027-3; published online 4 May 2010
© 2010 Kim Williams Books, Turin
 Fig. 1. MIT Chapel, main aisle (view toward Bertoia screen). Photograph by the author

Fig. 2. MIT Chapel, exterior approach. Photograph by the author

192 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
 While the water reflects light upward, the oculus admits a stream of light downward
onto the altar and circular marble slabs, bringing the “light of the world” to a
surrounding environment of darkness.1 While the quality of light from the oculus is
diffused by the grate-like structure therein, the altar screen is a powerful connection from
the light source through the space to the ground. Descending from the heavens to the
realm of humanity, the deceptively simple yet intricate structure uses wire and metal
plates to reflect light (see fig. 7). Designed by Harry Bertoia (another alumnus of the
Cranbrook School in Bloomfield Hills, MI, which Eliel Saarinen designed and his son
Eero attended), this structure uses the individuality of the rough metal plates to be
beautiful not only in surface detail, but also as a whole when perceived from further away
in the space. Since it is supported at its top and bottom by the building, it does not tend
to read as “sculpture” separate from the architecture, but rather as an emphatic focal
point within the aesthetic whole. The sculpture mediates the transition from the light of
above to the illuminated marble below, reflecting the mediation between God and man
at times of worship.
The use of light and water for elegant architectural principles is not limited to the
round interior of the chapel space. In contrast to the open sunlight of Kresge Oval, the
trees around the chapel provide a filtered mosaic of light and shadow on the brick
cylinder as well as the ground (fig. 2). The water surrounding the chapel creates a
physical and visual separation penetrated only by the bridge-like passage of the entryway.
Rather than having the spatial feel of a traditional narthex, this low, translucent corridor
provides a transitional, intermediate space that sets the viewer on a linear course down
the chapel’s main axis to the altar and oculus (fig. 3).

Fig. 3. Chapel, entry lobby. Photograph by the author

Nexus Network Journal – Vol.12, No. 2, 2010 193

 The chapel also benefits from Saarinen’s careful treatment of exterior space. The
aforementioned axis is aligned, not with the center of the auditorium, but with the off-
center placement of the south auditorium entrance (fig. 4). Furthermore, the whole
chapel-auditorium combination has been placed to the south of the main axis of MIT.

Fig. 4. Detail of a map of the MIT campus, showing the placement of Kresge Auditorium and the
MIT chapel
By shifting the axial emphasis, a bold and unusual concept for compositions of
circular forms, Saarinen creates asymmetrical outdoor spaces amongst the trees, giving
each building its own sense of place and identity without competing for prominence.
Even the use of brick in the chapel and the exterior terrace of the auditorium helps to
build a visual relation to the curved brick forms of Aalto’s Baker House visible beyond
(see fig. 10). Prior to the existence of Kresge Auditorium, the MIT Chapel, and the
raised oval of landscaped space, this area behind the apartment buildings (now Bexley
Hall) on Massachusetts Avenue was a flat, open space.2 While the chapel may now
appear more “central” to the overall campus, it predates all of the subsequent dormitories
on Amherst Street as well as the student center. Therefore, it was slightly more on the
outskirts of campus at the time of construction, but foreshadowed the growth of West
Campus as a residential area. Along with the formal development of the humanities as a
school and of students’ living opportunities on campus, Saarinen’s buildings constituted
a concurrent attempt to develop the opportunities for spiritual space in an urban, secular
environment.
In the highly crafted exterior space between the chapel and auditorium, the
surrounding forms are mainly horizontal, except for the single element atop the chapel.
Theodor Roszak’s design for the Bell Tower cannot be appreciated at the same proximity
or level of detail as the interior sculpture due to its physical location, but it contributes a
bold upward gesture that relieves the staid, earthbound forms of the chapel and
auditorium.3 Although one MIT description calls it a symbol of “the history and
authority of three major religious persuasions,” [MIT, Art and Architecture 1988: 45]
this seems to contradict and limit the more broad definition of an inherently
‘multidenominational’ or ‘nondenominational’ chapel, and is not supported by
Saarinen’s own statements. The bell tower can also be perceived in campus lore as an
abstracted ‘rocket to the heavens,’ or a gestural steeple, but its essence seems to exist not
in its overt symbolism, but rather in its verticality as the one element that aspires above

194 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
and beyond the surrounding tree canopy, extending the worshipful interior space into the
exterior environment and the atmosphere above.
The auditorium has no such pinnacle; it is often recognized and published for its
iconic copper-paneled roof , but it did not initially have such a covering; at the time of its
opening the stark whiteness of the membrane prompted critical reception.4 Even though
the initial membrane failed and reroofing was necessary less than a decade after initial
construction,5 the copper has persisted in a role as the metallic analogue of brick in
having a surface which becomes richer with age.
Kresge Auditorium, MIT’s performance and rehearsal hall, is the chapel’s
fraternal – and far from identical – twin. Kresge’s graceful roof, sheathed
in copper, its triangular plan and its glass-and-steel windows contain a hive
of activity. A little theater, a concert hall and rehearsal rooms within are
used for everything from drama to dance to music performances, as well as
symposia and science, technology and engineering conferences [Wright
2005].
Moreover, these program spaces designed for sound are column-free: the main hall
seating over 1200 and the theatre for 200, and the remainder of subterranean spaces for
rehearsal, storage, and service purposes are supported from below. Early site schemes
foresaw the demolition of Bexley Hall along Massachusetts Avenue and a more formal
relationship to both the main entrance at 77 Massachusetts Avenue at street level or via
other overpasses and underpasses, none of which were built. Saarinen’s vision of a
consistent northern edge was compromised by the looming brutalist symmetries of
Eduardo Catalano’s Student Center in 1968, but have been improved by the elegant
glazed scrim of the Zesiger Fitness Center completed by Roche/Dinkeloo (the successor
firm to Saarinen’s) in 2002 (fig. 5). The character of the auditorium is less baroque and
fine-grained in the scale of its detailing than the chapel, its curtain wall mullions less
delicate and its glazing more conventional, but even the exterior components of its
support structure at the corners define the lobby’s interior geometries within the
enclosure.

Fig. 5. Kresge Auditorium, exterior view, approach from Massachusetts Avenue (Zesiger Center at
right). Photograph by the author

Nexus Network Journal – Vol.12, No. 2, 2010 195

1 Geometries: Formal and Informal
First, the techniques of form introduce
geometrical complexities into the
aforementioned first-level simplicity. Both
the chapel and the auditorium are
constructed on a circular plan and a solid
of revolution: a cylinder and a sphere. The
solid is shown plainly in the case of the
chapel, while the sphere (112-ft radius) is
truncated into an eighth, i.e., one rotated
quadrant of three-dimensional space such
that its center (origin) is far underground
(fig. 6). The chapel is embellished by
surrounding the cylinder with its moat,
while the auditorium ripples with further
repeated concentric brick plinths, stepping
down to the lower land elevation to the
west.
Yet the circles are presented as both a
solution and as a problem to be solved:
Saarinen speaks of the cylindrical plan as a
result of site explorations to create a form
that maximizes its similarity as seen from Fig 6. Kresge Auditorium, sketch of geometry
many vantage points, and as a function of (sketch by the author,
minimizing the maximum interior after [Saarinen 1962: 127])
distance [Saarinen 1962: 34-6]:
We believed that what was required was a contrasting silhouette, a form
which started from the ground and went up carrying the eye around its
sweeping shape. Thus, a domed structure seemed right. There were other
reasons, too, that influenced us toward a dome. There was the large dome
of Welles Bosworth’s central building at MIT. A dome is an economical
way of covering an area with concrete. It is a shape which provides a
pleasant atmosphere for an auditorium. And a thin-shell concrete structure
seemed an appropriate form to express the spirit of this advanced school of
technology. […]
After many experiments, exploring different shapes in the site plan, the
round cylindrical form seemed right. The circular shape also seemed right
in plan – for this was basically a chapel where the individual could come
and pray and he would be in intimate contact with the altar. […]
The interior wall was curved, both for acoustical reasons, and to give the
space a lack of sharp definition and an increased sense of turning inward.
The circular geometries that circumscribe both buildings are only a point of
departure Saarinen seeks to resolve with a few modifications, to make them less – rather
than more – “sharply defined.” He therefore juxtaposes elemental6 or formal objects with
informal spaces within – not spaces that are casual, but that are intentionally irregular
such that they cannot be summarized in a single Platonic solid or form. The circular

196 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
geometries are contradicted to create dynamic spatial tension and reduce acoustical
deadness – the irregularly sinusoidal brick waves, the eccentricities of floor steps, the
inverted conical ‘bowl’ of the chapel ceiling that tapers asymmetrically, the contrasting
curvatures beneath the auditorium ceiling, and so forth. In doing so, Saarinen provided
clues of the external structure to the internal space, yet did not let the geometrical
structures dictate the interior spatial configurations.
Furthermore, within their enclosure, the subordinate spatial divisions emphasize the
difference in scale between the public realm in the auditorium and the corresponding
public realm of the narthex in the chapel. Both admit entry from two sides with
symmetric pairs of unsheltered doors,7 but whereas the Auditorium beckons the eye
upward to dynamically sloping planes beneath the seating above, the chapel presents a
striking contrast of a tightly compressed, rectilinear narthex as a “decompression
chamber” in Saarinen’s terminology. His evocation of the parlance of modern travel
emphasizes the tubular room’s bridgelike connection, leading directly across the
rhomboid grain of the marble floor tiles toward the eccentric internal steps.8
In both interiors, wood heightens the “informal” (or even “aformal”) qualities not
only through its material properties but as a material placed to curve alongside areas of
human interaction, with varying and increasingly variable radii of curvature. Their
variations increase or form a visual crescendo toward the focal areas. Saarinen shows a
hierarchy between arc segments with fixed curvature and these splines with variable
curvature: their curves appear to be chosen and expressive, which implies that they stand
in implicit contrast to the exterior fixed geometry, which we shall revisit in considerations
of the structure.
Therefore, at the auditorium and at the chapel (as in the TWA Flight Center and
other works) Saarinen’s apparent formal virtuosity with curves is not arbitrary and plastic
[Saarinen 1962: 11], but is instead grounded by the hierarchy of axial relationships. That
this is accomplished entirely without radial subdivisions is notable: Conventionally for
the time, many other American modern architects incorporated centralized and circular
geometries into contemporary religious and meeting structures. For example, in the
Midwest, Frank Lloyd Wright’s Annunciation Greek Orthodox Church (Wauwatosa,
WI; 1956), Alden Dow’s St. John’s Lutheran Church (Midland, MI; 1953) [Robinson
1983], and John Randal McDonald’s St. Therese Convent (Kenosha, WI; 1953) [Beno
1997] all use centralized forms and directed natural light to achieve intimacy within a
larger congregational space. In contrast to Saarinen’s consistent use of curvilinear
geometry, Dow and McDonald introduce various polygonal geometries to rationalize the
construction of the form. Wright and Dow place the altar in the geometric center, with
seating in a circumferential pattern. These other structures are between the chapel and
the auditorium in scale, so their degree of intimacy is somewhat diminished, but they
make Saarinen’s use of axiality rather than centrality that much more startling: since both
the Chapel and Auditorium have no object or emphasis at the geometric center, one
discovers their centers through acoustical reflections or deadening (where not to sit or
play) rather than through material or spatial paths that heighten the center’s importance.
2 Uses: Program and Experience
Along with the spatial reciprocity between the interior and exterior forms, there is a
temporal tension between the use of the spaces as originally programmed and as
experienced. These relate to differences of illumination quality and sound quality. While
a full discussion of the programmatic goals of the MIT administration and possible

Nexus Network Journal – Vol.12, No. 2, 2010 197

unexecuted alternative schemes for both buildings would be somewhat conjectural, it is
evident in the dedication that there was a burgeoning ambiguity as to how both spaces
would be used primarily for intended programs of spoken lectures and ‘non-prescribed’
multidenominational worship.
First, both spaces seem to have been commissioned and envisioned as paired meeting
places for an “academic community” that would also accommodate secondary
performance uses. Just as MIT’s application to the Kresge Foundation for a $1.5M grant
noted that the auditorium should be “less distinctly religious in atmosphere” but “could
be used for larger religious gatherings”,9 MIT President Killian’s dedication stated that
worship is first (for the auditorium, no less) and that music and drama linger further on:
In visualizing an auditorium for MIT, we sought a building to provide a
similar nucleus for our academic community, especially for our student
body. We felt it would be proper and possible to design a building which
would be appropriate for worship, for academic ceremonials, for
educational meetings and conferences, for music and drama, and for the
maintenance of the civil life of our academic community.10

Fig. 7. Kresge Auditorium, interior (view toward stage and organ loft).
Photograph by the author
The wood strips over plastic cloth that shield the acoustically absorptive glass-fiber
blankets at the rear of the hall are the most visible acoustic measures, and they form an
architectural transition between the reflective wood panels to the open wood screens at
the transverse aisle (fig. 7). Saarinen’s sectional drawings show the front movable portion
of the stage configured lower as a seating or small theatrical ‘orchestral pit’ area, rather
than having the movable portion elevated for an enlarged ensemble, since at design
orchestra performance was intended to comprise perhaps 5% of the use.11 Reporting in
Progressive Architecture, George Sanderson observed the Holtkamp organs included in
both the auditorium and chapel, but these permanent instruments, ancillary to the
spaces,12 were only the first indications of the role of music:
[sidebar] Though the large 1238-seat auditorium was designed primarily
for voice projection, it is also used for such musical events as organ recitals
and symphony concerts.

198 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
 [main text] On Dedication Day, we experienced the hearing environment
under a wide variety of speeches, fanfares and processional marches by a
Brass Choir; and the MIT symphony Orchestra, Glee Club, and Choral
Society performed Aaron Copland’s “Canticle of Freedom,” especially
composed for the occasion, and later in the program, a noble Bach cantata
[Sanderson 1955].
The Copland piece, revived and performed by the MIT Wind Ensemble and
Chamber Choir in 2000,13 is not one of the composer’s most known or lauded works,
but the precedent had been set at its premiere: large orchestral performances would
define much of the auditorium’s role on campus, and MIT has still not erected a
purpose-built music performance facility. As a result, without significant wing space and
only the freight elevator connecting to the Little Theatre to move its large instruments,
the space serves for rehearsals and performances of acoustic and amplified music from
student and professional chamber quartets to international dance festivals.14 Saarinen’s
landmark pair of structures serve as cultural sites within a campus possessing strong
traditions of music performance in ways that have therefore become highly particular to
both spaces.
The tension between the initial secondary role of music and its expanded role in the
use of the Auditorium is mirrored in the chapel, where the initial primary intentions
focused upon Saarinen’s aforementioned “bilateral lighting” and its “unique” role in
accommodating multiple religious traditions. Saarinen stated:
The challenge of the interior was to create an atmosphere conducive to
individual prayer. Since this is, uniquely, a non-denominational chapel, it
was essential to create an atmosphere which was not derived from a
particular religion, but from basic spiritual feelings. A dark interior seemed
right – an interior completely separated from the outside world (to which
the narthex passage would serve as a sort of decompression chamber). I
have always remembered one night on my travels as a student when I sat
in a mountain village in Sparta. There was bright moonlight over head and
then there was a soft, hushed secondary light around the horizon. That
sort of bilateral lighting seemed best to achieve this other-worldly sense.
Thus the central light would come from the altar – dramatized by the
shimmering golden screen by Harry Bertoia – and the secondary light
would be light reflected up from the surrounding moat through the arches
[Saarinen 1962: 36].
In this and other accounts, the role of the space for performance and for being a
haven for long reverberant sounds are generally absent from the architectural accounts,
save for the shaping of the interior wall for factual acoustic reasons rather than the
expressive axial reasons stated above (figs. 8 and 9).
The actual impact of this powerful lighting scheme described earlier therefore varies
with the actual worship practices in the chapel. As the definition of such a space as
“multidenominational” has expanded from merely connoting Jewish, Protestant and
Catholic in the 1950s, many of MIT’s religious groups use the chapel for their respective
worship times during the day and at night, and even more use the spaces in the adjacent
Building W11 with its offices, meeting areas, tradition-specific food preparation areas,
and other spaces absent from it.15

Nexus Network Journal – Vol.12, No. 2, 2010 199

 Fig. 8. MIT Chapel, screen detail (view upward toward ceiling detail at top of brick).
Photograph by the author

Fig. 9. MIT Chapel, interior lighting (showing light bouncing up through perimeter slot).
Photograph by the author

200 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
 Some religious events do not occur within the chapel because its axial orientation is
not compatible with their traditions, and others meet in spaces which are more “open”
and less solemn than the chapel. In actual worship services, however, the incandescent
lighting from the ceiling is almost always employed; natural light is almost always too
dim to read comfortably in the space. Although the electric lighting is providing the
actual usable light, the reflected light from the water and the overhead oculus remain lit
in such a manner that the daylighting diagram remains evident while acknowledging the
artificial light as secondary.
As the space is used for more than conventional religious observances and memorial
services, its acoustics, which make a single voice able to dominate even when a hundred
people are in the space, have also played a major role in defining it as a nonreligious
performance venue. Its interior soundscape, with its acoustical properties similar to
ecclesiastical buildings orders of magnitude larger, make it a compelling space for
performances as well as for other student-initiated ceremonies. While its visual size makes
chamber music of recent centuries seem appropriate, the persistence of sound and the
delay between when multiple musicians hear each other indicate otherwise. The present
Thursday Noon concerts offer the MIT community a range of musical genres, with a
particular emphasis on early music suited for long reverberation times. As in the
auditorium, the fittingly modernist pipe organ is a visually dominant but rarely used
harbinger of the musical activities: the harpsichord or piano are used far more regularly,
and a single tiny note can fill the volume. Thus, while performing in the space, qualities
of liveness and of warmth to the resonance of a wide range of frequencies transform
otherwise ordinary instruments’ and voices’ sounds into highly enriched and sustained
echoes.

Fig. 10. Kresge Auditorium, exterior (view with temporary tents and Aalto’s Baker House in rear).
Photograph by the author

Nexus Network Journal – Vol.12, No. 2, 2010 201

 Therefore, while the design and the initial reception focused on the properties of
vocal acoustics, in the decades since the acoustics have supported a broader range of
instrumental uses; the importance of musical performance has illuminated how these two
spaces have become small oases of natural materials, providing vibrant settings for
student life within MIT’s campus. The performance spaces may be within, but the
periphery of the Auditorium plinth and the concrete edge of the moat serve as larger
performative settings for campus social life: the brick steps form a rich outdoor setting
with the curtain wall as a backdrop and its turf a stage for a variety of temporary pavilions
such as a tensile fabric tents (fig. 10). These tents, stretched over the oval, take ideal
shapes in their tensile material, their forms an inverted counterpoint to the compressive
arches and domed shapes nearby. But how do Saarinen’s iconic structural shapes
communicate their structural logic?
3 Structure: Appearance and Reality
The dichotomies of human use and experience inform the third and most striking
tension, between the appearances of purely compressive forms (masonry walls with
arches, concrete shells), and the actualities of why and how they do not act in pure
compression.
Saarinen states that the dome was chosen not only for its resonance with the other
domes on the campus, but also for its efficiency as a modern form. Given uniform
loading on the roof, the shape of a doubly-curved vaulted (rather than necessarily
hemispherical) structure which is funicular and resists gravity loads in pure compression
would approximate a catenary; the minimum radius of a catenary is less than that of a
circle that passes through the same points, but the difference would be relatively slight
given the shallowness of the vault. Saarinen recognized the impulse of structural integrity,
but as with the initial catenary shape for the Gateway Arch in St Louis, his architectural
articulation of the Auditorium form complicates its visible lucidity. The complications
are not with the circular profile but with the manner in which it is supported and
truncated: Domes and spherical shells can be easier geometries for engineers to calculate
and builders to construct than those based on parabolas or other conic sections, and for a
complete dome, the circumferential hoop action of the shallow dome-like shell
contribute to its inherent stability and efficiency.
But since the one-eighth sphere is placed such that its outer perimeter is curved and
ventures beyond the triangle inscribed by the three points of support these areas at the
perimeter (shaded in fig. 6) are not acting in pure compression – transmitting loads
directly to the three points of support – but are instead in tension. This means the shaded
areas require substantial thickening of the shell overall and specific tensile reinforcement
that contradicts the pure structural logic in order to maintain the external appearance.
The dome is an equilateral, spherical triangle whose apexes reach down to
the ground. There, they rest in pintle-like supports that are free to rotate.
The structure is complex and, for a thin shell, quite thick. The structural,
reinforced concrete part of the shell varies in thickness from 3.5 in. at the
crown to 11 in. at the edges and over 20 in. at the points of the shell that
sweep down to the supports [Engineering News Report 1963].
An ideal shell supported on the three points could be found experimentally using a
hanging model or analytically using structural computations, but its form would be
irregular rather than comprising conventional conic sections. Unlike other masterful

202 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
compressive concrete shells by Felix Candela or those built more recently by Heinz Isler,
designers have recognized that when Saarinen chopped the dome into a tricorner form,
he destroyed the inherent balance he sought:
The underside of the construction is a geometrically exact part of the
surface of a sphere. Construction and form coincide. The hemisphere, out
of which this form is cut, is statically in equilibrium under vertical loading.
But, as soon as a piece is cut out of the hemisphere, the balance is
destroyed. Additional forces have to operate at the edges to restore the
shell’s equilibrium. The distribution of these forces is not externally
apparent in the construction of the auditorium. The form reflects the pure
geometry of the sphere. As in other instances, the problem lay in deciding
whether a shell ought to assume a form chosen beforehand, or whether the
form should be dictated by the distribution of forces [MIT buildings
bibliography: 126].

Fig. 11. Auditorium, Lobby (view south toward support condition). Photograph by the author
For Saarinen, the sculptural ideal of the spherical segmentation dominated over the
tectonic ideal of other doubly-curved shell forms optimized to the points of support (fig.
11). It is now notable in retrospect that this prefigures the same challenges relative to the
Sydney Opera House (which Saarinen championed as a competition juror), a design
whose sketched shells were ‘simplified’ and modified into spherical geometries (long after
Saarinen’s death). These changes, imposing a single radius on each spherical segment,
facilitated constructability and analysis by the team at Arup [Roman 2003; Jones 2006]
(including the young engineer Peter Rice), but the result is vastly thickened and the shell
functions structurally due to its thickened qualities rather than its shape. In a sense, all of
the sculpturally shaped shells from Kresge onward through the Sydney example have
demonstrated that even the prospect of a ‘thin’ shape creates numerous challenges in
realizing the project:

Nexus Network Journal – Vol.12, No. 2, 2010 203

 ‘What’s the philosophy behind a building like this?’ I asked one who had
to do with the auditorium. ‘To enclose space at the least possible cost.’ He
replied. ‘Did it?’ ‘Well, not perhaps this time, but from what we’ve learned
if we had three to do each would be cheaper than its predecessor and the
third would be considerably less expensive than the traditional form’
[Weeks 1955].16
…[T]echnical problems …arose as work proceeded, in spite of the most
careful and expert analysis by Amman and Whitney, the engineers.
…[N]ot only has the structural analysis been thoroughly justified in
practice but I have found the architect very willing to discuss his problems
in most frank terms. Amendments to the design have been of a minor
nature and were introduced as a result of contingencies which could not
have been foreseen in the design stage of such a revolutionary structure,
and they in no way detract from the appearance of the building [Scott
1955].
Yet the auditorium’s shell is not a question of mathematical analysis, as these and
other commenters seem to imply. Rather, it is a choice of conscious design
manipulations: at MIT Saarinen set forth the spherical shell as an inevitable and efficient
structural solution of supreme priority, placing all interior acoustical and construction
modifications as inherently secondary to the structural motivation. Again, to continue
the above assessment from 1955 by architect, planner, and founder of the BDP music
society N. Keith Scott:
There can be little doubt that the architect’s departure from the logical
approach to design, as we currently understand that term, has created
many structural and mechanical problems, and in many instances he has
relied upon the sheer ingenuity of modern technology to get him out of
difficulties…In his choice of structural form Saarinen flouts every precept
of basic acoustical design, for the concave ceiling and the curved rear wall
combine to prohibit good hearing conditions unless there is a vigorous
appliqué design to counter the tendency to focus sound [Scott 1955].
Nonetheless, Saarinen remains steadfast in his logic:
One concept underlying auditorium design is to let the functional
requirements – acoustics and sightlines – determine the form. But there is
no one ideal acoustical shape. Though function has to be respected, it
seemed equally justifiable to let the basic form come from structure. Thus,
in developing the design of this building, we felt very strongly guided by
Mies’ principles of architecture – of a consistent structure and a forthright
expression of that structure. As there are many ways of doing equally
functional things, we built dozens of models. As an auditorium requires a
triangular shape, we tried spanning this one-room building with a dome
supported at three points – the shape of one-eighth of an orange. At first it
seemed strange, but gradually it became the loved one [Saarinen 1962:
34].
Therein lies Saarinen’s shell game: the sphere persisted in the guise of a purely
efficient solution, even though this purity was no longer an ideal form for the truncated
plan shape or for other interior aspects.

204 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
 Nevertheless, the descriptive elegance of the one-eighth orange, melon, or other
handy illustrative fruit from the apocryphal story has persisted in historical discourse. In
the years since shells have no longer been economically efficient since expenses of labor
and formwork have outpaced material conservation in many Western economies.
Practitioners and students who study the auditorium rarely anticipate or comprehend the
structural gymnastics rendered invisible from within and without.
At a much more subtle scale, the brick arches of the MIT chapel have similar hidden
inner workings: While arches effectively support planar brick walls, transmitting their
distributed loads to discrete points of support, the curvature in plan of the cylindrical
face means that in perimeter the brick swerves beyond the line of action of the arch.17
This eccentricity invites viewers to speculate as to how ‘true’ the brick is. Saarinen may
have stated that this is undeniably a masonry structure –
It seemed right to use a traditional material, such as brick, for the chapel –
for brick would be a contrast to the auditorium and yet the same material
as the surrounding dormitories. But we felt that brick should be used with
the same principles of integrity to material as concrete or steel. This is
forthrightly a brick structure [Saarinen 1962: 36] –
but in truth this is an expression of structure rather than a diagram of its performance, as
the visible brick surfaces enclose the layered assembly within.
The arches on the outside occur where the exterior wall and the
undulating interior wall meet. In retrospect, especially having looked again
at the archivolts of Romanesque churches, I wish that we had given these
arches a richer, stronger three-dimensional quality. And I am aware that
the connection between narthex and chapel is clumsy. However, I am
happy with the interior of the chapel. I think we managed to make it a
place where an individual can contemplate things larger than himself
[Saarinen 1962: 36].
While no other permanent shells have been built at MIT since 1955, issues of shell
geometry have been a recurrent leitmotif of design exploration: architectural faculty
members such as Eduardo Catalano were famous for their shells, such as the hyperbolic
paraboloid for a house in North Carolina. Longtime faculty member Waclaw Zalewski
built dozens of highly efficient thin-shell roofs for factories and other spaces in his native
Poland and in Venezuela before teaching at MIT from 1966-1988.18 Several generations
of designers in the MIT community have collaborated on recent educational exhibits and
publications that seek to communicate the usefulness of graphical approaches for
designing vaults, domes, and even more conventional trusses, walls, and columns. These
educational aims, used in teaching at MIT and by associated faculty members practicing
and teaching far abroad are captured in Form and Forces [Boston Structures Group
2009], a book that provides instruction in the processes of graphic statics as well as form
optimization for all major structural systems, and uses works such as Saarinen’s main
terminal at Washington Dulles International Airport to illustrate applications of
structural design strategies that use the shape of structures for strength and efficiency
rather than for arbitrary purposes. Saarinen’s works do in fact demonstrate effective
structural logic. Saarinen himself was aware of these dichotomies. In an address given at
Dickinson College on December 1, 1959, he said:
The principle of structure has moved in a curious way over this century
from being ‘structural honesty’ to ‘expression of structure’ and finally to

Nexus Network Journal – Vol.12, No. 2, 2010 205

 ‘structural expressionism.’ Structural integrity is a potent and lasting
principle and I would never want to get far away from it. To express
structure, however, is not an end in itself. It is only when structure can
contribute to the total and to the other principles that it is important
[Saarinen 1962: 6].
And in a letter to a friend dated June 3, 1953:
Esthetically, we have an urge to soar great distances with our new materials
and to reach upward and outward. In a way, this is man’s desire to
conquer gravity. All the time one works, one concerns oneself with the
fight against gravity. Everything tends to be too heavy and downward
pressing unless one really works at it [Saarinen 1962: 5].
Postlude: Performing and Hearing
Finally, the acoustic qualities that result from and within Saarinen’s shells, vaults, and
walls are among the most iconic of American modernism. While scholars such as Emily
Thompson [2002: 319-320 ff] have articulated the rich evolution of the “modern sound”
for controlled acoustics in the earlier decades of the twentieth century, Thompson rightly
foreshadows how changing tastes and preferences motivated consultants such as Bolt,
Beranek, and Newman to reach beyond reverberation time toward a more qualitative
assessment of parameters such as the “initial time-delay gap” that linked data and
perceptive opinions. Leo Beranek’s book Music, Acoustics, & Architecture focused on a
litany of eighteen qualities that support excellence in concert hall acoustics, and
reverberation time is not one of them although it plays a constituent part in half
[Beranek 1962].

Fig. 12. Chapel, Interior (Detail at perimeter). Photograph by the author

206 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
 Fig. 13. Auditorium, Lobby Ceiling and curtain wall. Photograph by the author
The single most important quality, according to Beranek, is intimacy, followed by (in
combination) liveness, warmth, and the loudness of direct sound and loudness of the
reverberant sound, and then (in combination) diffusion, balance, and blend.19 These
qualities relate specific acoustic phenomena to a range of geometrical and material
connections, and translate them into the language of musicians. For example, the relative
audibility of bass-range sounds improves warmth, while intimacy is measured by the
shortness of time (the duration or “gap” mentioned above) between the initial sounds
and the first reflected sounds as determined by the proportions and physical geometry of
the surfaces that bound a space. That these experts collaborated with Saarinen for the
auditorium indicates that its acoustics should be viewed as a transitional experiment
leading up to the major works Beranek and his colleagues collaborated upon, and an
example of a building notable in that the reverberation time is not what characterizes the
success at the auditorium for various performing forces.
When the Kresge building was in its next-to-final stage members of the
Institute faculty used to wander into the main auditorium and standing in
the pit clap their hands for the joy of hearing the sound bounce from the
cement floor to the cement dome and back again; then they would seek
out Mr. Newman and tease him. But the acoustical engineers had the last
laugh; with delicate wooden gratings, with unobtrusive backstops of plastic
fabric hung over fiberglass pads, with sound-absorbent seats, they have
controlled and clarified the voices that come from the stage…[who] spoke
well and were heard with ease. The single voice and the solo instrument
are beautifully accentuated, but it is still a question of how true a blend we
shall get from the full orchestra [Weeks 1955].
Just as the roof membrane and metal panels over the shell were substantially replaced
not long after construction, the interior of the Auditorium has also had ongoing

Nexus Network Journal – Vol.12, No. 2, 2010 207

acoustical tweaks as visible in the configuration of the “clouds” to become more
undulating and more extensive. In its ‘deliberate reach for shape, beyond the definition of
function’ that made it “a pivotal one in US architecture today” in 1955 and a half-
century later, Saarinen created structures for creating community and introspection, but
the new faithful that come to venerate are the architects themselves:
Fifty years after the construction of Kresge and the chapel, architects still
make pilgrimages with their sketchbooks and cameras, and try to figure
out how this master of mid-century modernism did so much with
seemingly so little.20
The acoustical tensions and dichotomies between the spaces as perceived visually and
aurally serve to reconcile the internal complexities with the overarching gestures which
makes Saarinen’s works such resonant contributions to American modernisms. That they
have not only iconic imagery but iconic sounds is further evidence of how they transcend
the limitations of forms that could be static. Saarinen’s work here and elsewhere
constitutes not a series of built diagrams – simple forms reduced to their least
complicated – but rather a group of richly particular humanistic experiments that
negotiate the architectural pursuits of external clarity and internal discovery. Saarinen
places these in dynamic tension such that the act of listening within the spaces – however
unexpected and atypical – is just as particular and unreplicated as the forms and lighting
effects. Echoing the common theme in religious traditions continuing in the chapel, the
architecture and the soundscape of the Auditorium are in their world but not of it; they
recognize their role in the urban campus to transcend their context and create
independent environments for listening. Therefore, the visible architecture is not an end
but rather a means towards creating a more total sensory environment for spiritual and
temporal meaning. In Saarinen’s words:
I look for the day when our spiritual qualities catch up with our physical
advances. Then our architecture will take an important place in history...
[Peter 1994: 193].
Notes
1. The chapel design at MIT contrasts Saarinen’s chapel at the Lutheran Seminary in Ft. Wayne,
IN. In this instance, Saarinen designed not only the chapel and its environs, but the residences
and other surrounding campus buildings. That chapel, however, is a large, pointed form based
on an abstracted stave church, relating to the Lutheran church’s Scandinavian heritage. Unlike
the brick chapel at MIT, the example in Ft. Wayne is constructed mostly of concrete, yet it
shares the strip of horizontal windows that admit light into the nave from a shallow outdoor
pool. Due to the lighter tone of the concrete and the angled walls, the author observes that
light is reflected somewhat more readily into the space than at MIT, but without the textural
surface articulation achieved with brick.
2. The wartime growth of MIT had brought in many commuter students, and there was a large
parking lot for a time. After World War II, the development of student life and humanities
programs as described in the 1949-50 MIT Lewis Report coincided with the construction of
MIT’s first dorm west of Massachusetts Avenue.
3. “The chapel’s spire and bell tower, designed by sculptor Theodore Roszak, were added in
1956. The bell, also designed by Roszak, was cast at the MIT foundry, which was then located
on the top floor of Building 35” [Wright 2005].
4. “Seen by night, with its lights within, it is an opal; by day it suggest one of those curved white
hats the ladies have been affecting this spring, or, less elegantly and to the critics, it suggests a
diaper. Since the married students are housed directly behind the Auditorium it might be that
Mr. Saarinen had this symbol in mind, though I doubt it” [Weeks 1955].

208 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
5. “Kresge Auditorium at MIT has been completely reroofed just eight years after it was opened.
The surface of the auditorium’s remarkable three-point supported, concrete dome roof was in
trouble almost from the beginning. MIT’s maintenance engineers, who incidentally
collaborated on the original design, watched in dismay as violent and unexpected thermal
stresses weakened and destroyed the outer layer of the shell’s three layer system” [Engineering
News Report 1963].
6. “In designing a chapel suitable for use by worshipers of all faiths at the Massachusetts Institute
of Technology, Eero Saarinen chose as his starting point the elemental cylindrical form. The
way he made the form live without destroying its essential simplicity reflects his careful
consideration of every part” [Grubiak 2005].
7. See comments in [Weeks 1955].
8. One could compare the form of these in plan to conventional depictions of moving sound
sources and Doppler waves, but that would be highly conjectural and is not supported by
Saarinen’s documentation.
9. See [Grubiak 2007], note 33, “Submitted application to establish at MIT a Kresge School of
Human Relations,” 11 April 1950, AC 4 MIT Office of the President, 1930-1958, box 131,
folder 12, MIT Archives.
10. See also [Grubiak 2007], note 32: “Full text of an address prepared by Dr. James R. Killian, Jr.,
President of the Massachusetts Institute of Technology, for delivery at the Dedication of the
Kresge Auditorium and the MIT Chapel at 3:30 o’clock on Sunday afternoon, May 8” with
annotations, AC 4 MIT Office of the President, 1930-1958, box 131, folder 8, MIT Archives.
11. “50% for lectures and drama, 15% for organ, 5% for orchestra, 18% for chamber music and
operetta, 12% for soloists and glee club” [Beranek 1962: 109]. In the past decade, the relative
proportion of orchestral and large chamber rehearsals and performances has exceeded all other
uses, perhaps even large but infrequent conference lectures.
12. “Most organists point out that the reverberation time is too low...” [Beranek 1962: 108].
13. December 2000, Kresge Auditorium. The author played in the MIT Wind Ensemble for this
performance.
14. MIT has a music faculty larger than most other non-conservatory institutions in Boston,
serving over a thousand students who participate in music courses and ensembles each year; the
minor course of study in music is among the largest in the Institute, and MIT music majors,
such as Alan Pierson (Alarm Will Sound chamber orchestra), Andrew McPherson
(Tanglewood composition fellow), and others, have become nationally recognized composers,
conductors and performers.
15. The chapel was envisioned in some Saarinen studies to have a stronger rear bar with library and
office support functions, a version published but not executed; the rear wall simply lines the
alleyway with a single mute egress door; the masonry itself had deteriorated over the length of
the wall and its southern end was rebuilt in 2007-8.
16. Weeks also muses: “If in the intermission, there should be a choice between continuing with a
difficult pay or concert or going home to an open fire, I wonder whether the fire would not
win.”
17. Special thanks to Edward Allen FAIA for his illumination of this related point and for his
helpful commentary on structural design issues more broadly
18. In retirement, Zalewski has collaborated and co-taught with former faculty such as Edward
Allen FAIA, who along with MIT’s architecture, building technology, and engineering faculty
and students have explored the analysis and design of compressive structures with newfound
computational tools as well as centuries-old practices of graphic statics. See also
http://www.shapingstructure.com and http://web.mit.edu/masonry.
19. Cf. [Nuzum 2009]. Beranek’s eighteen qualities are: [intimacy (presence), liveness, warmth,
loudness of the direct sound, loudness of the reverberant sound, definition (clarity), brilliance,
diffusion, balance, blend, ensemble, immediacy of response (attack), texture, freedom from
echo, freedom from noise, dynamic range, tonal quality, and uniformity. Even though Beranek
developed an interrelated way to use these qualities to ‘score’ concert halls, Nuzum notes that
the manner by which the eighteen qualities are inextricably linked to each other means that at
least nine of them relate directly to reverberation, further evidence of the qualitative specificity

Nexus Network Journal – Vol.12, No. 2, 2010 209

 of Beranek’s descriptions: “Just as the Inuit people of the northern reaches of North America
are said to have hundreds of words for all the different kinds of snow, in the world of acoustics,
reverberation is known by many names” [Nuzum 2009: 10].
20. William J. Mitchell, professor of architecture and media arts and sciences, quoted in [Wright
2005].
References
ALBRECHT, Donald, ed. 2006. Eero Saarinen: Shaping the Future. New Haven: Yale University
Press, in association with the Finnish Cultural Institute in New York.
BENO, Fr. Brian. 1997. John Randal McDonald. Milwaukee, WI: Private Printing.
BERANEK, Leo. 1962. Music, Acoustics, & Architecture. New York: Wiley
BOSTON STRUCTURES GROUP (Allen, Zalewski, Foxe, Anderson, Ochsendorf, Ramage, et al.).
2009. Form and Forces: Designing Efficient, Expressive Structures. New York: Wiley.
DELONG, David G. 2008. Eero Saarinen. Buildings from the Balthazar Korab Archive. Library of
Congress visual sourcebooks in architecture, design, and engineering. New York: W. W.
Norton in association with the Library of Congress.
Engineering News Record. 1954. Tripod dome built on tricky formwork. Engineering News
Report, 27 May 1954.
———. 1963. Roof Repair: More than Skin Deep. Engineering News Report, August 8, 1963.
FOXE, David M. 1999. Finland: Architecture and History – The quest for national identity from
traditional vernacular architecture through Saarinen and Aalto. Unpublished.
GRUBIAK, Margaret M. 2007. Educating the Moral Scientist – the chapels at IIT and MIT. ARRIS
The Journal of the Southeast Chapter of the Society of Architectural Historians 18: 1-14.
JONES, Peter. 2006. Ove Arup: Master Builder of the Twentieth Century. New York: Yale
University Press.
MIT buildings bibliography. Cambridge, MA: Rotch Library, Massachusetts Institute of
Technology, 198-?]-T171.M423.R68 1980
MIT Committee on the Visual Arts. 1988. Art and Architecture at MIT: A Walking Tour of the
Campus. Cambridge, MA: MIT Press.
MITCHELL, William J. 2007. Imagining MIT: Designing a Campus for the Twenty-First Century.
Cambridge, MA: MIT Press.
MERKEL, Jayne. 2005. Eero Saarinen. London; New York: Phaidon.
NUZUM, James M. 2009. How and Why Architects Should Claim Acoustics. Unpublished research
conducted for a 2009 course “Variations in Sound and Architecture” at the Boston
Architectural College taught by the author of this paper.
PETER, John. 1994. The Oral History of Modern Architecture. New York: Abrams Publishing.
ROBINSON, Sidney K. 1983. The Architecture of Alden B. Dow. Detroit: Wayne State University
Press.
ROMÁN, Antonio. 2003. Eero Saarinen: An Architecture of Multiplicity. New York: Princeton
Architectural Press.
SAARINEN, Eero. 1962. Eero Saarinen on his Work: A Selection of Buildings Dating from 1947 to
1964 with Statements by the Architect. New Haven: Yale University Press.
SANDERSON, George. 1955. New MIT Buildings Open. Progressive Architecture 36, 10 (July
1955): 74-75.
TEMKO, Allan. 1962. Eero Saarinen. New York: G. Braziller.
SCOTT, N. Keith. 1955. MIT auditorium: an English view. Architectural Record 117 (July 1955):
138.
THOMPSON, Emily. 2002. The Soundscape of Modernity: Architectural Acoustics and the Culture
of Listening in America, 1900-1933. Cambridge, MA: MIT Press, 2002.
WEEKS, Edward. 1955. The Opal on the Charles. Architectural Record 117 (July 1955): 131-137.
WRIGHT, Sarah H. 2005. Architectural wonders, chapel, Kresge turn 50. MIT News Office,
October 19, 2005. http://web.mit.edu/newsoffice/2005/chapel-1019.html.

210 David M Foxe – Saarinen’s Shell Game: Tensions, Structures, and Sounds at MIT
About the author
David M. Foxe holds degrees in architecture and in music from MIT, and was a Marshall Scholar
to Clare College, Cambridge (UK). A former architectural tour guide of the two Saarinen
structures as well as Aalto’s Baker House at MIT, he draws upon experience as a music composer
and performer who has written and played vocal and instrumental music in both spaces since 1999.
While continuing to teach courses and publish works on design history, structural techniques, and
the relationship of architecture and music, he currently practices architecture in Boston and
practices piano in Newton.

Nexus Network Journal – Vol.12, No. 2, 2010 211

 Ozayr Saloojee Research
University of Minneapolis The Next Largest Thing:
College of Design
School of Architecture The Spatial Dimensions of Liturgy in
UMN Twin Cities
89 Church St S E
Eliel and Eero Saarinen’s Christ
Minneapolis, MN 55455 1 Church Lutheran, Minneapolis
[email protected]
Abstract. Christ Church Lutheran in Minneapolis,
Keywords: Eliel Saarinen, Eero Minnesota, was designed by Eliel Saarinen, then 75, and
Saarinen, Christ Church added to by his son Eero Saarinen 10 years later. Deeply
Lutheran, modern architecture loved by its community, it also serves as a touching example
of the relationship between the father and the son. This
present examination looks at the building on various scales,
underscoring the finesse and material elegance of the building
complex, the spatial genius and expertise of Eliel Saarinen,
and the deferential addition by Eero.

Written into the National Register of Historic Places, Eliel and Eero Saarinen’s
Christ Church Lutheran was named as a National Historic Landmark by the U.S.
Department of the Interior in early 2009. Widely considered to be Eliel Saarinen’s
masterwork, the church has been hailed as the singular building example that heralded a
new direction for ecclesiastic architecture in the United States. Completed in 1949,
Eliel’s sanctuary sits – in the words of his grand-daughter Susan Saarinen – “quietly
there”.1 It is an unprepossessing building from the exterior, a massing of simple forms –
largely rectangular seeming solids, faced with Chicago brick and Mankato Stone,
occupying the corner of 34th Avenue South and East 33rd Street in Minneapolis’s
Longfellow neighborhood. Its exterior belies its interior, which demonstrates Eliel
Saarinen’s consummate skill as an architect capable of understanding the scale of
experience as an essential part of liturgy and as an evocative catalyst for a deep and
personal sense of spirit. Approached by the local congregation shortly after World War
II, Pastor William Beuge wrote a challenging letter to Saarinen, then head of the
Cranbrook Academy of Art in Ann Arbor: “I asked him if it were possible in a
materialistic age like ours to do something truly spiritual.” The young pastor observed
that, “He soon showed me”.2
An important question during this ongoing research into Christ Church Lutheran has
regarded how to approach a reading of this building. This research – and this writing – is
not the work of an historian, so the question of how to uncover or tease out readings of
this building has been a preoccupying concern; it has perhaps resulted in an initial
reading that departs from an expected analysis or interpretation. Compounding that is
the focus on a building that has been the subject of numerous, if not hundreds, of
interpretations over its almost sixty year history. Christ Church Lutheran has been
studied and drawn by students and architects from all over the world, recently hosting
part of an international symposium paralleled with the exhibit “Eero Saarinen: Shaping
the Future”, mounted in Minneapolis at the end of 2009. Curiously, however, the
building is not as firmly embodied in the architectural radar of those who visit the area;
building aficionados are more likely to know Jean Nouvel’s (new) Guthrie theatre (and
sadly, be less familiar with Ralph Rapson’s original – now torn down), or Jacque
Herzog’s and Pierre de Meuron’s Walker Art Center, or the Gehry project – the
Weisman Museum –more than the Saarinen project.

Nexus Network Journal 12 (2010) 213–237 Nexus Network Journal – Vol.12, No. 2, 2010 213
DOI 10.1007/s00004-010-0032-6; published online 11 May 2010
© 2010 Kim Williams Books, Turin
 Fig. 1. Eliel Saarinen helping lay the cornerstone of Christ Church Lutheran in 1949. Image ©
Christ Church Lutheran Archives. Still from archival footage digitized by author

Fig. 2. Eliel and Loja Saarinen at the building opening. Image © Christ Church Lutheran Archives.
Still from archival footage digitized by author

214 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 That Christ Church Lutheran was designed by Eliel Saarinen at the age of 75, and
added to by Eero Saarinen 10 years later, adds to the knotty question of how to approach
an interpretation of this particular building. Eero was asked by the congregation in the
early 1950s to design the addition to his father’s building. Two very distinct and
idiosyncratic masters shaped this project over a period of more than a decade. Apart from
all of this, perhaps the most important aspect that has been brought to the fore in the
study of this building is knowing that it is, and has been, the object of a deeply felt love
by its community – well before the cornerstone was laid in 1949 (figs. 1, 2) – and that it
serves as a touching example of the relationship between the father and the son.
If architects design for effect and to affect, then a mark of the good architect is to do
so without compromising either the technical or the experiential. This dual capacity of
the good architect – the ability and nuance to negotiate the scale of experience between
the mechanical and experiential – was one of Eliel Saarinen’s great skills. In considering
that the building was authored both in the name of the father and the son, this notion of
“effect” (and in a way, of “cause”) has a larger resonance; the hand of two designers with
markedly different attitudes about form and space underpin this building.
The evidence of this is embodied at Christ Church Lutheran in two particular ways:
in the subtle deftness of Eero’s addition to his father’s building, and in Eliel’s designed
experience of the sanctuary. Eliel Saarinen was a master of scale: his work encompasses
the broad scope from the local to the global. He designed and proposed city plans all over
the world: in Estonia, Finland and Australia, even consulting on projects in South
America. He designed buildings – museums, apartment, manors and campuses – and he
designed furniture and fixtures – from chairs to tables to teapots and teacups.
If the normative way to look at a building would be from the outside in, this
investigation will start instead from the inside out – from object to space to room to
building. By looking at these scales, and through these lenses, an attempt is made to
underscore the finesse and material elegance of this building and to touch on the spatial
genius and expertise of Eliel Saarinen, and a surprisingly soft and deferential project by
Eero.
Eliel Saarinen was never satisfied with acceptable, or even good responses; he
attempted to conceive of projects in their entirety, at all scales and, perhaps most
importantly, to know what the appropriate use of these scales are. The range of scales is
embedded here at Christ Church Lutheran, and it is reflected in many ways: from the
material and tectonic to the immaterial and the ineffable; to the way that choral music
surrounds worshippers and visitors; to how a monumental experience of space is made
personal and intimate; to the subtle and deeply poetic way in which light is treated, used
and shaped. Eliel understood how light becomes a material principal of this building. In
the Lutheran tradition, God’s Grace comes by faith alone, through Christ alone – Sola
Gratia, Sola Fide, Solus Christus. Light, of course is one of the mechanisms with which
architects and designers try to express a sense of the sacred. At Christ Church Lutheran,
for Eliel Saarinen, light revolves around liturgy connotatively and denotatively.
Eliel Saarinen would have been very familiar with this liturgy. His father, Juho
Saarinen, was himself a Lutheran minister, tending to congregations in Rantasalmi,
Finland and in Lisisila in Ingermanland, Russia. Juho would later leave his congregation
in St. Petersburg and move to the outskirts of Helskini, to join his son and family at their
Hvitträsk estate. The Lutheran liturgy is organized by four particular movements:

Nexus Network Journal – Vol.12, No. 2, 2010 215

  Gathering;
 Word;
 Meal;
 Sending.
These movements are embedded at varying scales in Christ Church Lutheran, and
overlap through both the material and immaterial qualities of the building. They serve as
a kind of conceptual touchstone for the reading of this building and as a narrative glue.
This interpretation will, as noted earlier, start with an inversion of sorts, moving from the
micro and shifting out towards the macro, a kind of Eamesian “Powers of 10,” looking at
both Christ Church Lutheran and its designers, Eliel and Eero Saarinen.

Fig. 3. Title-page, ‘Furnishings Book’ for Christ Church Lutheran.

Image © Christ Church Lutheran Archives
Appropriately, for a church, we can begin with a book. Not THE book, but a book
(fig. 3), and certainly an important one; The Book of Details, if you will, and a gospel for
an architectural understanding of Christ Church Lutheran. Provided for the church as
part of the final set of construction drawings for the building, this book contains original
drawings from Saarinen’s office showing the furnishings for the church. If we start at the
scale of the liturgical objects depicted within, we can extract aspects from them that allow
us to posit a manner in which the architect understood his work, not just in the form of
the building, but also the way in which he communicated these ideas to us and the
community at large. The drawings in this book parallel the way in which the represented
objects themselves exist in the physical space of the church, how they occupy it and are
highlighted by it. This is a key relationship. British artist David Hockney observed that
how we depict space indicates how we behave in space;3 there is a relationship between
the image of the object and our experience of it. Here, in the good Book of Details, we
see these representations – these object-drawings – as artifacts displayed against a surface,
organized against the backdrop of the page. Their drama is played out not only in the

216 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
usual convention of the architectural drafting board, through plan, section or elevation,
but also in perspective, with cast shadows. These objects have been rendered as
experienced things. The images in the book can therefore be understood as an expression
of architectural thinking through drawing; before the objects were made, they were
imagined as matter, thought of as matter. These liturgical artifacts themselves can then be
understood as the manifestation of this thinking through making/drawing. At Christ
Church Lutheran, the immaterial was understood through the material realization of
ideation.

Fig. 4. Main Cross, ‘Furnishings Book’ for Christ Church Lutheran.

Image © Christ Church Lutheran Archives
The cross (fig. 4) is the focal point of the chancel and for the congregants as they
gather, in fulfillment of the first movement of the Lutheran liturgy. The volume of the
sanctuary itself and the subtle geometry of the nave walls direct one’s vision forward. The
cross – the ultimate symbol of Christianity, representing at once Christ’s saving death
and His resurrection – casts an ever-lengthening shadow along the rear of the chancel
wall. Through the architecture, the artifact sits against a backdrop of light falling upon
the Chancel wall: a material scrim for the divine.
The cross is 22 ft. tall, constructed of stainless steel and mounted to the wall with
aluminum brackets. Its section is diamond shaped, with clear, hard edges. Recollections
from Nick Hayes, the son of Mark Hayes of Hills, Gilbertson and Hayes, the local office
associated with the Saarinens, tells the story of one of Eliel Saarinen’s periodic site visits.
Laborers had painted the chancel wall with two coats of white, oil-based paint, and
Saarinen, upon seeing it, instructed them to grind back the surface of the newly painted
Chancel wall, knowing that the softened backdrop would yield a more appropriate
experience of the cross and its shadow. Indeed, the drawing of the cross in Saarinen’s
book of details articulates the dramatic intentionality of the liturgical object against an
ephemeral backdrop; the shadows are rendered against an implied surface. In
constructing the effect of the cross, Saarinen simultaneously understood its affect (fig. 5).

Nexus Network Journal – Vol.12, No. 2, 2010 217

 Fig. 5. Time-lapse imagery of light across the Chancel. Image © P. Sieger, AIA and Tom Dolan,
Minneapolis, Minnesota
From the cross, our view moves down to the altar and clockwise to the piscina, the
bowl used during liturgy. Both the chancel and chapel altars were part of the original
design of the church, and are critical components of the experience and nature of the
sanctuary. Saarinen highlights both altars against a curved wall which wraps the objects
with the diffuse light that enters through translucent windows. The chancel wall curves
forward from the sanctuary altar and leads to the piscina, which is located near the altar
used for Communion and is mounted on the same wall as the cross. This curve leads to
the pulpit, which has its own curved geometries that align with the side-aisle to the
north. This suggests a line towards the baptismal font – completing a kind of liturgical
loop – one that highlights the space of the Chancel, strongly evident in the plan of the
building. The pulpit (fig. 6) is the station of the utterance of the Word – the second
movement of the Lutheran liturgical movement – and is in a way a centralizing moment
within the space itself.
Through these objects, Saarinen observes and celebrates the integrity of a set of
individual moments within the space, but this isn’t a hermetic strategy because each of
these spaces is impacted by the broader program of the building. These object-moments
exist as individual events, but are connected, both conceptually and experientially to the
larger scale of space around them.

218 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 Fig. 6. Pulpit, ‘Furnishings Book’ for Christ Church Lutheran.
Image © Christ Church Lutheran Archives
The altars and their candles are also objects against the page (as drawn in the book)
but are also projected against the walls of the church, or in the case of the baptismal font,
against a volume defined by its location: situated in a curve in the floor, in a space of its
own, with a flooring pattern that differs from the surfaces around it. The Baptistery is
bounded by a short, low wall and one steps down into it. The floor is a glossy tile,
distinct from the other material surfaces of the sanctuary. Stepped down, it recalls
Christ’s baptism in the River Jordan by John the Baptist.
These liturgical artifacts – the cross, the altars, the piscina and the font – have
deliberately considered and architecturally nuanced spatial and experiential conditions.
As we expand our scale beyond the object as a single moment, we can observe that they
are always inserted in a larger, careful material and luminous context. Although the space
of the liturgy occupies the entire spatial field of the sanctuary – indeed, the whole
building – Saarinen draws out a set of careful architectural conditions that highlight these
liturgical and (to coin a term) artifactual moments. The curve of the chancel wall (fig. 7),
and the wooden baffle or screen carefully modulate the light that falls onto the surface of
the whitewashed brick. All the glass used in the sanctuary (save the entry and bell tower
windows) is translucent, which further tempers the eastern and southern light; Eero
Saarinen would later use this exact strategy in his design for the Kresge Chapel on the
M.I.T campus in Cambridge, Massachusetts. Specific walls are terminated in such a way
that the visible cut edge of the walls themselves define program and space within the
building. For example, the two flanking walls from the narthex to the sanctuary are faced
with brick headers only, as is the short wall of the baptistery. These edges help articulate
particular zones or programs of the church in a less explicit way. The chancel wall edge,
perpendicular to the window, has its own pattern of headers and stretchers; further
emphasizing this particular volume and its bounding by the symbol of the cross, the
station of the Word, the font and altar assist in expressing it as a space of elemental
spirituality (fig. 8).

Nexus Network Journal – Vol.12, No. 2, 2010 219

 Fig. 7. Sanctuary, view of the chancel and cross. Image © P. Sieger, AIA and Tom Dolan,
Minneapolis, Minnesota

220 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 Fig. 8. Sanctuary. Image © P. Sieger, AIA and Tom Dolan, Minneapolis, Minnesota
Saarinen is deliberate and methodical about the walls in Christ Church, the way they
extend, connect, imply or are cut and exposed. The west wall that bounds the baptistery
extends outward, well past where we would expect it to merge with the southern aisle
wall. In doing so, it mimics a similar strategy used in the Chapel’s East wall. The material
and formal expression of these moves point to a considered hierarchy of the spaces of the
church; Saarinen’s attitude toward the building is clear, in that his architecture
underscores spaces of profound spiritual experience within the building, but he does so
with quiet poise. The side aisles, too, use this strategy: the west aisle links to the heavy
volume of the bell-tower, the east connects to the narthex and lobby. Both connect the
glazed end of an aisle wall to an opaque masonry volume.
For those coming for the first time, the reaction to the building is often one of casual
indifference. Certainly from the outside, the Saarinen complex maintains a remarkably
unassuming and unprepossessing presence. Upon entering the building however, that
casual indifference changes quickly. The church is raised up a few steps on a plinth above
the level of the sidewalk, and one enters an arcade before turning to the main doors of
the building. The arcade was not part of Eliel’s early design, but was instead added
through Eero’s addition. Eliel Saarinen, with the entryway, demonstrates not only his
careful attention to detail with a carefully and intricately crafted door, but also in framing
the entry with a mullion-less glazing, clearly separating two formal elements within the
entry spaces of the church, again suggesting the integrity of these moments within the
overall structure and experience of the architecture. He then does this in several places
within the building and the result is anything but fragmentary; he achieves a kind of

Nexus Network Journal – Vol.12, No. 2, 2010 221

architectural unity through multiplicity of scale. He also does this at the campanile, an
object linked by glass on both sides, and tied into the structure of the building through
roof, floor and stair. It reads simultaneously as an artifact in its own right, against the
backdrop of the volume of the building of the church, but the stairs that lead to the choir
loft are pulled off the side walls by a small gap of a few inches, which allows light to
penetrate between wall and stair. This scale of separation – at the scale of both the detail
and the volume of the building – would lead Eero Saarinen to employ a similar strategy
in the double stair of the addition that leads down into the basement level of the
education wing.

Fig. 9. Side-aisles. Image © P. Sieger, AIA and Tom Dolan, Minneapolis, Minnesota

222 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 The movements of the Lutheran liturgy are ancient, and go back to the very
beginnings of the Christian Church. The entry of the congregant into the space is
directed forward to the altar and the cross and the chancel wall; this is made more
emphatic by the volume of the sanctuary and smaller coves of the side-aisles. The side-
aisle windows are angled forward, with a smaller edge profile that reduces the amount of
visible lit surface as the congregants gather. The second movement of the liturgy is the
Word – Sola Scriptura – which takes the pulpit as the seat of the book, and shifts the
focus of the congregants to the pulpit from the cross. The Meal directs worshippers
forward again to the altar. The manner in which the third movement is performed
underscores how Saarinen’s architecture augments the experience of the Lutheran liturgy
and how this building can be understood as a facilitator for worship. Once complete,
congregants split into two lines and move back to the pews, along the side-aisles. This
time, however, they are exposed to the broader face of the side-aisle windows, with their
edges more prominently in light (fig. 9). The view back, towards the east, is again, a brick
surface, softly lit from either side through translucent glass.
This particular reading of the side aisles is particularly striking as a carefully
considered and beautiful effect: following the Meal – the central act of Christian
Worship, the Holy Communion – one’s walk back is in a field of light, brighter than
when one entered. The worshipper approaches the altar in darkness, and leaves in light.
The geometry of the building is more than a conceptual framework for the liturgical
act or a way to facilitate the spiritual drama of this worship: it is also a demonstration of a
carefully considered set of technical solutions. Christ Church has been called one of the
finest examples of acoustic finesse in American architecture. Eliel collaborated with the
acoustic consultants Bolt, Beranek Newman (BBN) in the design of Christ Church
Lutheran’s acoustic qualities. BBN was founded by Leo Beranek and Richard Bolt (two
professors at MIT), who partnered with a former student, Robert Newman, to form the
company. Their first major consulting commission was the acoustic design for the
United Nations General Assembly Hall in 1949,4 but archival material from Christ
Church Lutheran notes their involvement with Eliel Saarinen on Christ Church.
The geometries of the building carefully attenuate any acoustic resonance that might
result in echoes or unwanted reverberation. No surfaces are truly parallel (fig. 10). The
north wall of the Sanctuary, with its variegated surface, is angled slightly off the
orthogonal. The east wall is also turned in, the chancel wall is curved, and the canted
ceiling, both in the sanctuary, and over the side-aisles, have been carefully considered.
The choir loft of the balcony edge is angled forward too. The perforated ceiling has been
designed with a pattern of sound absorption panels as well as void spaces, to allow for a
richer, truer auditory experience. No electronic augmentation or microphones are
necessary within Christ Church; one can comfortably hear a sermon or presentation
without amplification. The sensory experience is remarkable and the choral effect of the
music is outstanding. One is, in a way, truly surrounded by music and by the Word. The
sanctuary is a receiver both of light and of sound.

Nexus Network Journal – Vol.12, No. 2, 2010 223

 Fig. 10. Building geometry, composite image by author,
with photographs from P. Sieger, AIA and Tom Dolan
The building is never lopsided. Given its sometimes subtle and apparent shifts, it
manages to achieve, in a special way, a balance and a very human scale through its
asymmetries. This appropriateness of scale continues when we return to the exterior; the
side aisle on the south side immediately presents a comfortable pedestrian scale to the
street edge; as opposed to, for example, the massiveness of the sanctuary wall coming
straight down. This creates a kind of cove, a pedestrian-friendly volume pulled from the
sanctuary extending out in a gesture to the street, in recognition of the quiet residential
quality of the Longfellow neighborhood. Even with a five-story campanile, the church
fits very nicely in this part of town. This attention to human scale is again reiterated on
the north side of the sanctuary, with the courtyard faced by the east side aisle, giving a
more comfortable and humanizing scale to the garden. Eliel Saarinen continues his
nuanced manipulation of scale, further breaking up the mass of the sanctuary through a
simple line on the north and south walls – a single brick header extrusion – suggesting
the volume of the chancel within. Instances of how the scale architecturally “speaks”
about space are found on the outside of the church as well (fig. 11).
Eliel Saarinen was unlike the architectural crop working at the time. In 1949 the year
the Eames House was built and Phillip Johnson’s Glass House was completed, Wallace
Harrison was collaborating, with, among others, Le Corbuser, Ernest Cormier and Oscar
Niemeyer on the United Nations Building in New York, and Mies van der Rohe was
working on the Lakeshore Drive Apartments in Chicago and the Farnsworth house in
Plano. Mies and Gropius came to the US in 1937; by the time they arrived, Eliel
Saarinen had been in the US for almost fifteen years.

224 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 Fig. 11. Building exterior. Image © P. Sieger, AIA and Tom Dolan, Minneapolis, Minnesota
Eliel, on the heels of his second place award in the Chicago Tribune competition in
1923, was working in the United States a few years after the first World War and during
the Second. He was fully immersed in a quiet, but conscious pursuit of redefining the
post-war architectural landscape. If the modern movement at the time emphasized form,
structure, and the “coolly” rational, Saarinen proved with Christ Church that he was
quite unlike the modernists of the time. When Frank Lloyd Wright called Eliel Saarinen,
“The best of the Eclectics,” Saarinen labeled him “Frank Lloyd Wrong” [Art: The
Maturing Modern 1956].

Nexus Network Journal – Vol.12, No. 2, 2010 225

 Saarinen was a master of context and place, establishing an appropriateness of
architectural expression for a buildings site, distinguishing him from the Zeitgeist.
Today, none of us would argue about the importance of an architecture of context and
Eliel Saarinen was ahead of his time with regard to this as well. This is reflected in the
early work he did in Finland. The apartments at 17 Fabianinkatu or the projects in
Katajanokka, the Railway Station and the National Museum, for example, are considered
subtle and nuanced departures from the prevailing National Romanticism that
preoccupied many of the architects working in Helsinki and Finland at the time.
Preoccupied with this tricky question of style, his desire for his own architectural
expression was expressed in the book he completed in 1948, The Search for Form in Art
and Architecture. He believed in collaboration – as evidenced by how his office
functioned – and not in the role of the architect as a heroic figure working alone.
Saarinen said:
Architecture embraces the whole form-world of man’s physical
accommodations, from the intimacy of his room to the comprehensive
labyrinth of the large metropolis. Within this broad field of creative
activities, the architect’s ambition must be to develop a form language
expressing the best aims of his time – and of no other time – and to
cement the various features of his expressive forms into a good
interrelation, and ultimately into the rhythmic coherence of the multi-
formed organism of the city [Christ-Janer 1979: xvii]
This particular observation is the heart of his accomplishment at Christ Church
Lutheran because the building privileges an awareness of spirit through a careful
relationship between effect and cause, through a clear and deft touch with material and
light – the creation ultimately, of sonorous space – both as a crafted architectural strategy
and as a rich and lasting experience. The building demonstrates – and this is one of its
great effects – the manner in which the masonry is held up with light, almost dissolving
through the experience on the inside.
Completed at the same time as the Johnson’s Glass house, Saarinen ultimately
eschews the overt character of the form itself, as the structure – that physical causal reality
of all architecture and in the case of Christ Church Lutheran, steel – is completely hidden
through the material and surface elegance of the brick itself (figs. 12, 13). If architecture
is achieved by the thoughtful resolution of opposites, then Eliel Saarinen, here at Christ
Church Lutheran does it in spades. Light and the surfaces that are carefully designed to
receive and hold it, its conceptual experience versus its programmatic reality, its mass
versus its form, its effect versus its cause, all serve the desire of the architect to create a
space of manifest spirituality. If architecture is a battle plan against gravity, then Eliel
Saarinen, with the quiet and restrained poise of an architectural acrobat gives, us a space
where light trumps steel and brick.
In 1948, Saarinen’s second book, Search for Form: A Fundamental Approach to
Architecture was published. 1948 was also the year that Albert Christ-Janer’s Eliel
Saarinen: Finnish-American Architect and Educator was published, “...the book that he
(Eliel Saarinen) and his wife Loja both approved” [Christ-Janer 1979: xv]. Coincident
with this – the publication of Christ-Janer’s biography and The Search for Form – Eliel
Saarinen would begin to design, with Saarinen, Saarinen and Associates, and the
Minneapolis firm of Hills, Gilbertson and Hayes, Christ Church Lutheran.

226 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 Fig. 12. Building construction, view of side aisle and sanctuary from street.
Image © Christ Church Lutheran Archives

Fig. 13. Building construction, view of side aisle and sanctuary from campanile.
Image © Christ Church Lutheran Archives

Nexus Network Journal – Vol.12, No. 2, 2010 227

 Saarinen begins his book with the observation that the search for form, if “sincere and
honest” [Saarinen 1985: v] is a simultaneous process of creation and analysis; that the
creative process is the immediacy of design married to a developing system of personal,
critical reflection. Saarinen’s ideas of form, their genesis and evolutionary paths, logics
and theories to the transitional imagination of man (the dogmatic to the mechanized to
the creative mind) are all detailed here. This book is the amalgam of over a half century’s
experiences, reflections and musings, all directed at the principal meditation of how and
why we make the things we do and essentially, of art as a total experience, rooted in the
primacy of nature.
That book is not an historical overview of this thinking, nor it is a scholarly treatise
on the subject. It is an attempt by Saarinen to impart a broad sense of what constitutes
form-making to the young artist, this after more than quarter of a century of teaching,
and more than fifty years of architectural practice, begun in 1894 in Finland. Saarinen’s
primer on the search for form which, as both the title and content of his book, is a
combination of a reflection of his life as an architect and as an educator.
His early projects in Finland – those done first with Armas Lindgren and Hermann
Geselius and later on his own – evidence the sensibility of forward thinking that has
characterized much of Eliel Saarinen’s work. While his contemporaries were still deeply
connected to ideas rooted in National Romanticism, Art Deco and Jugendstil, Eliel
Saarinen was quietly, but very consciously engaged in the search for an appropriate
architecture through a deliberate dialogue with material and context. Christ-Janer
references Johan Sirén’s (head of the Architecture Department in Helsinki) 1955 opening
address to Saarinen’s memorial exhibition in Helsinki:
The early growth of Eliel Saarinen reveals the inner relationship between
personality and style. It is also convincing proof, as it always is when an
artist is born, that stylistic expression is only a surface phenomenon; the
impulses of the soul of the architect go deeper. The quality of expression is
the constituent that decides the artistic, the ultimate value. This quality
was evident during his early years [Christ-Janer 1979: 7].
Christ-Janer further notes that although the firm of Geselius Lindgren Saarinen were
resistant to the embellishment of surface so characteristic to Art Nouveau, the three
young architects embraced the “inventive spirit of the trend” [Christ-Janer 1979: 9].
This search for an appropriate context of architectural expression is the theme of
Saarinen’s life as an architect. He said:
Most urgently something had to be done to build an art form out of the
eternal fundamental principles and to bring architecture and design, in
general, out of its humiliating state. In order to achieve this goal, it was
necessary to free design from the its style grip and to let it develop in full
freedom according to the nature and character of the time” [Christ-Janer
1979: 25].
Christ Church presents a unique opportunity to evaluate the work of two architects
over a period of time, and explore their work critically and creatively in relation to the
other. Even though Eero Saarinen was partnered with his father, the sanctuary for Christ
Church is predominantly Eliel’s hand. In 1953, in a New York Times article titled,
“Now, Saarinen the Son” published on April 26th and written by Aline Bernsetin
Loucheim, then an associate art editor, whom Eero would marry a year later, Eero
Saarinen observed of his father that,

228 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 I often contributed technical solutions and plans, but only within the
concept that he created. A better name for architect is ‘form-giver’ and
until his death in 1950, when I started to create my own form, I worked
within the form of my father [Loucheim 1953].
The competition for the St. Louis Arch was perhaps the inflection point for this
realization. Initially awarded to Eliel Saarinen (the award notification was sent to “E.
Saarinen), the correction was noted shortly after and Eliel celebrated his son’s coming
into his own with a grand party at Cranbrook. Designed by Eero in 1947, a year before
the contract for Christ Church was issued, Eero Saarinen would not live to see the Arch
built; construction was only begun in 1963 and finished five year later. The Jefferson
Memorial competition coincides with Eero’s burgeoning assumption of increased
responsibility in his and his father’s shared architectural practice. Robert Clark and
Andrea Belloli write that:
Eero Saarinen returned to Cranbrook during the summer of 1936 and
entered into practice with his father. During the next three years, they
produced a series of designs that marked the end of the elder Saarinen’s
transitional phase, established the manner of expression that characterized
his last decade of practice, and served as a point of departure for Eero’s
independent work [Clark and Belloli 1984: 65].
Clark and Belloli also observe that Eliel’s genius lay in “craft and synthesis rather than
innovation” [Clark and Belloli 1984: 68], compared to Eero’s characteristic ingenuity,
inventiveness and his aggressive pursuit of new measures in form and technology in
architecture.
When clarifying some issues for Saarinen’s biographer some years after Eliel’s death,
Joe Lacy noted that:
As you probably know, the last three or four years before Eliel’s death were
more or less transitional. Eero assumed more and more responsibility until
eventually Eliel seemed content to let Eero lead the way. However, Eero
had great respect for his father’s ability and I doubt if Eero ever ignored
Eliel’s ideas.5
In a letter two months later, Lacy wrote that:
The very last building that was designed by Eliel and which was built was
the Christ Lutheran Church in Minneapolis That was entirely Eliel and
Eero had very little to do with it. The pastor of the congregation, Rev.
William Buege, is a brilliant man. He made an exhaustive search for the
right architect and was finally convinced that Eliel was the right one.6
If Eero ever felt shadowed by the form of his father, he quickly demonstrated the
power of his own architectural trajectory following his father’s death, distinguished even
in the way their practices operated; Eliel embodied the very description of a master in his
atelier, sanguine, grand, dapper, while contrasted with Eero, rumpled and intense in his
office (fig. 14), often working 18-20 hours a day, and relying on the collaborative spirit
of his now famous employees: Pelli, Parker, Paulsen, Lacy, Roche, Dinkeloo, among
many others. Although Eero had achieved a notable amount of fame with his early
furniture projects, the Jefferson Arch competition was an important defining moment in
the relationship between father and son. Their studio was “divided with a wall of

Nexus Network Journal – Vol.12, No. 2, 2010 229

blankets, pillows and sheets,” so as not to reveal the developing designs of father and son
to each other.7 Indeed, Eero would significantly revise the design and detailing of the
General Motors Technical Center in Warren, Michigan following his father’s death –
clearly demonstrating a commitment to his own ideas compared to the original scheme
of his father’s. Originally a shared project between father and son, the GM Technical
Center became one of Eero’s career-defining projects, eventually leading to many other
corporate commissions.

Fig. 14. Eero Saarinen’s office. Image © Leonard Parker, AIA

From his remarkable entry into the St. Louis Memorial Arch competition,8 Eero
Saarinen was one of the pioneering spirits in redefining the American Architectural
landscape (fig. 15). Out of the approximately forty “25 Year Awards” given by the
American Institute of Architects, Eero won five during his career. Growing up under the
gaze and desk of his father, his career – although tragically cut short at the age of 51 –
and contributions to the modern architectural canon cannot be overstated.
Saarinen, Saarinen and Associates continued until Eliel Saarinen’s death in July 1950,
shortly after the dedication of Christ Church Lutheran. That ultimately saw the full
flowering of Eero Saarinen’s great capacity as an architect, as well as his discipline and
superhuman commitment to work, on one project committing approximately
$12,000.00 towards a competition whose first prize award was $4000.00. It was noted
that:
In a single evening he will run through 170 ft. of tracing paper” and he
made more than 2,000 drawings in revising his plan for the London
embassy. A woman in his office, whose desk Saarinen sometimes uses late
at night, inevitably knows when he has been there. Says she: “It's like
slicing down through the excavations at Troy—tracing paper, tobacco,

230 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 paper, paper, matches, more paper, a cigar stub, paper, paper, paper [Art:
The Maturing Modern 1956].
This work ethic would yield some of the most celebrated architectural works during
the period between his father’s passing and his own in 1961.

Fig. 15. The St. Louis Arch. Image © Leonard Parker, AIA

Fig. 16. The TWA International Flight Center, model. Image © Leonard Parker, AIA

Nexus Network Journal – Vol.12, No. 2, 2010 231

 Perhaps best known for his iconic designs for the Gateway Arch in St. Louis and the
TWA International Flight Center at JFK Airport (fig. 16), Eero’s oeuvre is impressive.
He designed the John Deere Headquarters in Moline, Illinois, but died a week before
construction commenced on the almost 700-acre campus. He designed the GM
Technical Center in Warren Michigan, the CBS Building – the Black Rock – in New
York, the IBM Plant in Rochester, the Bell Labs Complex in New Jersey, the Miller
House in Columbus, Ezra Stiles College and Ingalls Rink at Yale, and Washington
Dulles Airport. The echo of his father’s influence can be best seen in the Kresge chapel
on the MIT campus. Quiet, restrained and singularly nuanced, much of the careful
quality of Christ Church is evidenced in this gem – in particular, the nuanced use of
transparent and translucent views in the entry – akin to Christ Church’s windows – and
in the careful attention to brick as a surface. From the TWA International Flight Center
to the Tulip Chair, like his father, Eero was adept at many scales.
Through all of this emerging body of work, and seemingly embodying the paradox
and complications of being at once a mainstream architect, a multi-vocal designer and an
innovative stylist, Eero Saarinen would return to the site of his father’s last built work,
and would build on his father’s capacity for understanding effect and cause and the
appropriateness of scale. He would design and supervise – at least the early stages of this
project, because he died before this building was completed – this addition to Christ
Church Lutheran (fig. 17). One can imagine that, at what was really the height of his
career, when Eero Saarinen received a request to add to his father’s building, the architect
of the Jefferson Arch and the TWA Terminal, perhaps, straightened his tie, smoothed
down his hair and honored the building and the man that ultimately transformed the
direction of ecclesiastical architecture in the United States.

Fig. 17. Eero and Paulsen’s addition, with the sanctuary beyond.
Image © P. Sieger, AIA and Tom Dolan, Minneapolis, Minnesota

232 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 The relationship between Eliel and Eero Saarinen is a timeless evocation of the
enduring bond between father and son. In their own ways, they represented broadly
different temperaments in period, style, attitude and time. Eliel was the genteel, cultured
aristocrat, educator and writer, whose deft drawing hand and ideas of architecture as a
synthetic craft quietly suggested a new direction for American modernism, while Eero
was the intense, focused and inward-looking innovator who came to symbolize the
trailblazing enthusiasm of post-war architecture in the United States. Eliel’s oeuvre was
identifiably consistent; it is easy to trace the formal and architectural lineage from one
projects to another; Eero’s body of work is vastly different in form, resulting in early
scathing criticisms from the likes of Vincent Scully (cf. [Romàn 2003: 2]) who accused
him of having no consistent language and said that his adaptability (now acknowledged)
signified a lack of discipline!
Eero’s addition to his father’s work, with the significant contributions of Joseph Lacy,
Glen Paulsen, Leonard Parker and Cesar Pelli,9 is – to use a current of the theme of this
writing – perhaps the most appropriate scale of effect to Christ Church Lutheran. It is
modest and utilitarian, in the best senses of those words, and playful too, with a veritable
Knoll and Eames showroom fronting 34th Avenue (fig. 18), but it is understated and
deferential to the form of his father. Glen Paulsen, interviewed for the developing
monograph on this building, noted that Saarinen’s driving concern was the
appropriateness of the Education Wing as a quiet response to the sanctuary. Paulsen’s
office worked with Eero Saarinen on this project, and Glen Paulsen10 served as the
primary designer, with Eero as the deciding voice on the project.

Fig. 18. The Education Wing’s “Knoll Showroom.”

Image © P. Sieger, AIA and Tom Dolan, Minneapolis, Minnesota
The addition, although programmatically different from the sanctuary, is connected
to it through materiality and detailing. The great success of Eero’s and Paulsen’s addition
lies in the apparent unity of the Education Wing with Eliel’s sanctuary, but also shows an
almost, I believe, tongue-in-cheek response to the work of “The Great Man,” as Eliel was
known, or as he was more affectionately called by his students and employees, “Papi.”

Nexus Network Journal – Vol.12, No. 2, 2010 233

Fig. 19. The sanctuary seen through an atrium window in the Education Wing. Image © P. Sieger,
AIA and Tom Dolan, Minneapolis, Minnesota

234 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
 Fig. 20. Christ Church Lutheran, exterior view of the sanctuary.
Image © P. Sieger, AIA and Tom Dolan, Minneapolis, Minnesota
While Eliel Saarinen’s sanctuary is a profoundly inward-looking building, the
Education Wing, particularly on the main floor, opens up to its context.
The floor-to-ceiling windows of the Luther Lounge (fig. 19), running the full length
of the east facade of the Education Wing, open to the residential life of the street. Eero
Saarinen and Paulsen turns the interior hallway that separates the classrooms into a
windowed and sky-lit atrium; the main hallway that links the new entry around the small
courtyard, past the church offices and into the sanctuary, is a glazed cloister. The
basement is admittedly more utilitarian, but the spaces are still very social, with two

Nexus Network Journal – Vol.12, No. 2, 2010 235

lounge areas, connected to an industrial (and elegant) kitchen, linked to huge
gymnasium, whose bulk and mass surpass that of the sanctuary, but is ultimately not part
of the experience of the building’s exterior.
While the sanctuary is a quiet volume whose experience is wholly interior and is an
expression of a poetic understanding of light and liturgy, the education wing is more
extroverted – in an introverted sort of way. As a tribute to his father (and in a way, as
Paulsen’s tribute to his friendship with Eero Saarinen), with the Education Wing, Eero
has managed to reflect his own attitude about architecture in a way that honors Eliel, but
also echoes his own vision for architecture. There is a unity at Christ Church Lutheran
that is not only the unity of Eliel Saarinen’s Cranbrook, of views and changes of scale
and, as Paul Goldberger noted, of “creating a constant sense of surprise” [Goldberger
1981: 311]. Certainly the sanctuary embodies these notions, from its exterior simplicity,
to its interior views and its experiential complexity. The unity at Christ Church Lutheran
is of the balance of opaque and transparent, of interiority and exteriority, of the
restrained craft of Eliel, and in a wonderful surprise – an innovation of sorts – of the
subtle and quiet hand of Eero.
When asked what his favorite building was, Eliel Saarinen replied, “The next one”,11
anticipating that the work of the architect is never truly complete, that the search for
form is unending. Christ Church Lutheran sits, in Susan Saarinen’s words “quietly
there,” serving as a sensitive tribute to the broad and unique genius of both father and
son (fig. 20). Eliel Saarinen sought specificity and appropriateness through scale, but
with Christ Church Lutheran he achieved a kind of spiritual universality, and the
addition by Eero Saarinen and Glen Paulsen continues that essential quality. The
education wing is not flamboyant by any stretch of the imagination, but extends in a
generous arc from the work of Eliel’s Sanctuary, bounding the courtyard and deftly
borrowing and interpreting the father’s work – a curved wall, a slight reveal between stair
and edge, a carefully scaled facade – while maintaining, very discreetly and informally,
the authorship of the son.
Born on the same day thirty-seven years apart, August 20th, and collaborators on the
same projects for some time, both Eliel and Eero Saarinen died of similar causes. They
were both awarded the AIA Gold medal – Eliel Saarinen in 1947, Eero posthumously in
1962. Eero would accept the RIBA Gold Model for his father’s contribution to
architecture in September of 1950 just two months after the death of Eliel on July 1st of
that year. Eero died in 1961. But for this text and these reflections and importantly, for
the experience of the building itself, the church complex constitutes, in the buildings that
embrace a courtyard, the lasting continuity of the relationship between father and son.
Notes
1. Susan Saarinen made this comment in September of 2008 at a lecture she delivered at the
Minneapolis Institute of Art. Named as one of the Finlandia Foundations co-lecturers of the
year (with Marc Coir), Ms. Saarinen observed – it was her first visit to the complex designed
by her father and grandfather – that it was one of her favorite buildings in their oeuvre.
2. Pastor William Buege, in a letter dated November 2005 to Rolf Anderson. Mr. Anderson, a
local Minnesota historian, generously shared this correspondence with me.
3. Paraphrased from Hockney’s comments in the excellent documentary film, “A Day on the
Grand Canal with the Emperor of China or: Surface is Illusion but so is Depth” [Haas and
Hockney 1989].
4. See the BBN Technologies website at http://www.bbn.com/about/timeline/ (accessed 28
March 2010).

236 Ozayr Saloojee – The Next Largest Thing: The Spatial Dimensions of Liturgy in Eliel and Eero Saarinen’s Christ...
5. Letter from Joseph Lacy to Albert Christ-Janer, dated January 2nd, 1971, Box 2, Folder 2.1
Saarinen Family Archives.
6. Letter from Joseph Lacy to Albert Christ-Janer, dated March 18th, 1971, Box 2, Folder 2.1
Saarinen Family Archives.
7. Susan Saarinen, conversation with the author, Minneapolis 2008.
8. According to Marc Coir, Eero never gave credit to Carl Milles for suggesting that the section of
the arch be triangular when Eero had requested his advice concerning his entry into the
competition. Coir noted this in his lecture at the Minneapolis Institute of Art in September
2008.
9. The author has conducted several interviews with Leonard Parker, who, as the site supervisor
on the addition, noted Pelli’s contributions to the developing design in Glen Paulen’s office.
10. It’s interesting, albeit incidental, to note here that Eero recruited Paulsen personally to join the
Saarinen office. Paulsen’s first projects included working with Eliel Saarinen, but later
transitioned to joining Eero on his work. Paulsen, notably, became the third president of
Cranbrook, after Eliel Saarinen and Zoltan Sepeshy. He was interviewed in November 2008
for a developing book on Saarinen (of which this paper will form a part) in November 2009.
11. Marc Coir made this observation in his presentation with Susan Saarinen at the Minneapolis
Institute of Art in September 2008.
References
Art: The Maturing Modern. 1956. Time Magazine, July 2, 1956.
CHRIST-JANER, Albert. 1979. Eliel Saarinen: Finnish-American Architect and Educator. Chicago:
University of Chicago Press.
CLARK, Robert Judson, and Andrea P.A. BELLOLI. 1984. Design in America: The Cranbrook
Vision, 1925-1950. New York: Harry N. Abrams Publishers.
HAAS, Philip and David HOCKNEY. 1989. A Day on the Grand Canal with the Emperor of China
or: Surface is Illusion but so is Depth. Directed by Phillip Haas and written by David
Hockney. Documentary film.
GOLDBERGER, Paul. 1981. Eliel and Eero Saarinen. In Three Centuries of Notable American
Architects, Joseph J. Thorndike, Jr., ed. New York: American Heritage Publishing Company.
LOUCHEIM, Aline. 1953. Now Saarinen The Son. New York Times, April 26, 1953.
ROMÀN, Antonio. 2003. Eero Saarinen: An Architecture of Multiplicity. New York: Princeton
Architectural Press, 2003.
SAARINEN, Eliel. 1948a. The Search for Form in Art and Architecture. Rpt. 1985, New York:
Dover Publications.
———. 1948b. The Search for Form: A Fundamental Approach to Architecture. New York:
Reinhold Publishing.
About the author
Ozayr Saloojee is an Assistant Professor of Architecture at the University of Minnesota’s College of
Design. He joined the faculty in 2005, after teaching and practicing architecture in Ottawa,
Canada, where he completed his B.Arch and M.Arch degrees at Carleton University’s School of
Architecture. Born and raised in Johannesburg, South Africa and in Canada, Professor Saloojee’s
current academic areas include questions of tradition and modernity in Islamic Architecture, and
research interests in contested landscapes, and the political agency of the architect in these conflict
terrains. He participated with the Walker Art Center (Minneapolis) and Minneapolis Institute of
Art (Minneapolis) with the hosting of the Eero Saarinen retrospective “Shaping the Future” in
2008 (as a co-organizer and invited presenter of a related symposium). He also collaborated on the
exhibit “Christ Church Lutheran: Three Photographic Visions” with VJAA (Minneapolis) Site
Assembly (St. Paul). The exhibit design was awarded an American Institute of Architects Honor
Award in 2009. Professor Saloojee is currently completing a monograph on Eliel and Eero
Saarinen’s Christ Church Lutheran.

Nexus Network Journal – Vol.12, No. 2, 2010 237

 Luisa Consiglieri Research
Painel da Lameira Cx 8 Morphocontinuity in the work of Eero
7300-405 Portalegre
PORTUGAL Saarinen
[email protected]
Abstract. Continuity as the mathematical tool in the creation
of architectural forms is known as morphocontinuity. In the
Victor Consiglieri present work, we explain how morphocontinuity appears on
Faculty of Architecture the work of Eero Saarinen and discuss its correspondence
Technical University with its environmental (physical, social and cultural) contexts.
Lisbon, PORTUGAL Keywords: Eero Saarinen, mappings, transformations

Introduction
The buildings designed by Eero Saarinen are constituted by continuous forms that
have the particular characteristic of unifying the internal and external spaces. The
equilibrium of the compositions is achieved by the relationship between the visual spaces
and the existent movement between them.
Here we discuss Saarinen’s Kresge Auditorium at MIT, the Ingalls Rink at Yale
University, and the TWA Flight Center at JFK International Airport in terms of
morphocontinuity. The thin shell structures are interpreted as continuous forms and
these forms appear mathematically as the membranes of cells [Consiglieri and Consiglieri
2009]. These new forms supersede discrete processes such as breaking up and scattering.
Rather, they are characterised by fluidity and unity. Moreover, the architectural design
flows into vaster urban areas than mere buildings. This development coincides with our
concept of morphocontinuity. The movement and complexity transmitted by the
surfaces have a new meaning. The main achievement is then that the building is no
longer a closed object but is now an open one. The symbiosis between cities and nature is
achieved when no boundaries exist between private and public domains, and between
interior and exterior spaces. The form can give rise to a multitude of effects which offer
light and sound behaviors in a wide variety of potential directions. The inner spaces,
which flow one into one another, are becoming a concern in the reconfiguration of cities.
The process of production of the buildings is the result of the interaction between
organization and material.
Morphocontinuity is correlated with the morphologic rhythm. It is the four-
dimensional framework due to the self-organized life process of a cell [Consiglieri and
Consiglieri 2006]. The mathematical cell is a 4D element reflecting cyberspaces.
Actually, when we move about in a city, whether inside or outside the buildings, the act
of moving brings the fourth dimension into the behavior. This new organization provides
the input for the ongoing existence of cities. With the morphologic rhythm, the cell and
its correspondent form provide the dynamism for the human living in the different
environments.
The present study uses mathematical mappings as a means of analysis. Moreover,
their graphical representation shows the visual contrasts of fullness/emptiness and
brightness/darkness. Although not all of Saarinen’s projects include Euclidean forms, the
architectural forms have such differentiability properties that equilibrium is reached.
Finally, the aesthetic value of the object is determined by the continuity of the forms and
the relationship between inside and outside.

Nexus Network Journal 12 (2010) 239–247 Nexus Network Journal – Vol.12, No. 2, 2010 239
DOI 10.1007/s00004-010-0026-4; published online 5 May 2010
© 2010 Kim Williams Books, Turin
The mathematical framework
In this section, we introduce some definitions for a fixed positive time parameter,
T>0.
Definition 1. We say that a set C is a cell if it is defined by

C : {(x, y, z)  IR 3 : f(x, y, z; t) d 1} , (1)

with x, y, z representing (as usual) the spatial Cartesian coordinates and f denoting a
continuous mapping defined in the three-dimensional Euclidean space and dependent on
a time parameter t  [0,f[ (for details, see [Consiglieri and Consiglieri 2006]).
Fig. 1 displays four 2D-cells, namely the graphical representation of circles of radius
1:

(a) at instants of time t = 0 and t = 2 when f(x, y; t) x 2  (y - t) 2 .

(b) at instants of time t = 0 and t = 1 when f(x, y; t) x 2  (y - t) 2 .

(c) at instants of time t = 0 and t = 1 when f(x, y; t) | x |2-t  | y |2-t .

(d) at instants of time t = 0, t = 1 and t = 2 when

f(x, y; t) |x | t 2 /2-3t/2  2
 | (2t  1)y - 5t 2 |t
2
/2-3t/2 2
/(1 t(t -1)) .

Fig. 1. (a) Euclidean rhythm: translation and juxtaposition; (b) interpenetration of congruent
objects; (c) self-organized life cell; (d) movement and shape change

240 Luisa Consiglieri – Morphocontinuity in the work of Eero Saarinen

Definition 2. We say that a cell C has a morphologic rhythm if
R ! 0, C(t) Ž B R (0), t  [0, T ] ,

where BR (0) denotes an open ball centered at the origin (0,0,0) with radius R>0.
The above examples take one ball in 2D as a motif (see [Consiglieri and Consiglieri
2006] for 3D examples) under a rigid body motion where the transformation is such that
the two balls act as the unique set (1) that has a free parameter denoting the time. In
order to represent the juxtaposition (fig. 1a), choosing t=0 we get the initial ball and
choosing t=2 it represents the translated ball. If we choose smaller instant of time, for
instance t=1, we obtain an interpenetration of the congruent ball (fig. 1b).
The self-life of a cell may be clarified as follows. If we take the circle of radius one, its
shape can freely change maintaining the quality of its boundary, the circumference, in
separating its interior from its exterior, e.g., becoming a square (cf. fig. 1c). Moreover,
the morphologic rhythm can have the following interpretation: the ball can even perform
motion and shape changes simultaneously (cf. fig. 1d).
Definition 3. We define a form as the subpart of the boundary of the cell given by

F : {( x, y, z)  A u IR Ž IR 3 : z h( x, y; t )} ,
where h is a continuous mapping defined in A depending on a time parameter
t  [0,f] (for details, see [Consiglieri and Consiglieri 2009]).
The mathematical space is the infinite set of points given in reference to the three
dimensional Cartesian coordinate system. However, the architectonic space can be
understood as an urban space given by the continuum 3D-space plus the time which are
embedded in the 4D (space-time) coordinate system (see fig. 2).

Fig. 2. The graphical representation of a form using A [2,2]2 \ ^ (0,0)`, for the time parameter
t>0 and h(x, y; t) 
x 4 y4 : (a) t==2; (b) t=3.
(x 2  y 4 ) t

Nexus Network Journal – Vol.12, No. 2, 2010 241

 The equilibrium should be understood as the aesthetic harmony of the visual and
physical movements of the curves of the surfaces, i.e., of the properties of continuity and
differentiability of the forms. At the moment, there is no a unique solution to a perfect
equilibrium if the equilibrium is interpreted in terms of the dynamical point of view just
described. In order to explain the present interpretation, a brief look back at the past is
helpful. The Greeks aimed to obtain unifying effects, with the objective of abolishing and
balancing the ascendant and descendant forces, without destroying the perception of the
volumes. Using the same principles, the Roman edifices were more complex and
appeared to be lighter, notwithstanding the fact that they were still massive buildings.
During the Gothic period, the structures are dominated by arches and buttressing
elements. These systems of equilibrium were structurally stable due to the application of
proportions. With the dome of Santa Maria del Fiore cathedral in Florence by
Brunelleschi [Argan 1955], the fusion between constructive technique and aesthetic
emotion is managed by load methods. The movement, dynamism and complexity
transmitted by the surfaces in the Baroque period were given by the successive repetitions
of fragmented pieces, but plastic expression was an entity disassociated from the
functionalities of the building. However, even in this era the essential characteristic of
exuberance in architecture could be achieved by infinitesimal analysis [Frankl 1981:
177]. As long as the structure of a building was conditioned by the weight of the
volumetric masses and the distribution of the loads, the form does not express its
structures. The form is subordinated to the balance of its volumetrics. Only in the
modern era does the structure become the form itself. It stands up to the plasticity. The
plasticity introduced in architecture affords the possibility of realizing new
configurations.
The Kresge auditorium
Extracted from the Euclidean object, the sphere, the concrete roof is the one-eighth
of a sphere anchored at hidden abutments at three points [L’architecture d’aujourd’hui
1956; Wright 1989] (fig. 3).

Fig. 3. Eero Saarinen, Kresge Auditorium, MIT. Drawing by Teotónio Agostinho

Under the morphocontinuity approach, the concrete shell can be understood as a
form and the building as a cell (see Definitions 3 and 1, respectively). This simple form is
one solution for the problem of roofing of large spaces. At the same time, it engages both

242 Luisa Consiglieri – Morphocontinuity in the work of Eero Saarinen

the senses and the intellect. It achieves an aesthetic value, although some critics argue that
the visual form is not in correlation with its function (for instance, Bruno Zevi, quoted in
[L’architecture d’aujourd’hui : 50-51]).
Mathematically, 1/8 of a sphere can be represented as the form defined by:

F {( x, y, z)  IR 3 : x, y t 0, z R2  x 2  y2 } ,

where the constant R>0 denotes the radius. The cell correspondent to the auditorium is
not 1/8 of a sphere, but is rather the set

C {(x, y, z)  IR 3 : x, y, z t 0, x  y  z t R, x 2  y 2  z 2 d R 2 } .
This cell represents the closed set constituted by the inner space surrounded by the
(1/8 of the) sphere and the planar section (fig. 4) and its boundary, namely the portions
of the sphere and the plane. In order to identify the set C with the auditorium in
accordance with Definition 1, the set C is rotated such that the triangle becomes the
ground floor, i.e., xy-plane (z=0), under the rotation over the axis located at (x,R-x,0), for
all real x. The projection of the boundary of the spherical surface is no longer the triangle
after rotation, but is rather the union of three arcs of ellipses.

Fig. 4. (a) 1/8 of a sphere; (b) planar section

Moreover, the building is characterized by its volume, which is calculated as the
difference between the volume integral under spherical coordinates that represents the
volume of 1/8 of the ball and the one that represents the volume of the tetrahedron
S /2 R S R R -x R -x - y S 1 3
V ³ ³ ³S
0 0 /2
U 2 sin M dM dU dT  ³ ³ ³
0 0 0
1 dz dy dx
6
R3 
6
R .

At a first interpretation, the spherical structure separates the inner from the outer
spaces. However, the internal environment changes in accordance with different
situations of the two systems: doors and windows. The sensation of space-time depends
on the location of the doors, and the communication with exterior depends on the
dimensions of the windows. Besides, the interconnection of the interior with the exterior

Nexus Network Journal – Vol.12, No. 2, 2010 243

is quite different if the openings are located at a lower or a higher level. At the superior
level, the relationship with the exterior is panoramic. If it is located at the ground level, it
creates movement toward the different directions and human activities. It reflects that the
outside is always another inside, as argued by Le Corbusier [Consiglieri 1994: 95].
Here, the transparency of the glass walls dematerializes the boundary between the
interior and the exterior spaces, dissolving the traditional dichotomy between inside and
outside.
The Ingalls rink

Fig. 5. Eero Saarinen, Ingalls Rink, Yale University. Drawing by Teotónio Agostinho
Given the couple composed of the rectangular rink and the seats that surround it,
which forms the basis for the architectural object, the principal facade is surprisingly
constituted by an awning with the boundary edge measuring 73.0 m (fig. 5). The main
purpose of this shell is to funnel the audience from outside into inside through the
mediating point, which is the entrance. This entrance, which faces onto and
simultaneously ends at the lateral walls, flows into the walls of the parking and the
external arrangements. The axis of symmetry of the suspended vault coincides with the
axis of the rink and the corresponding lateral seats, establishing the unity of the inside
with the outside. Thus we can say that the building is constituted by two twin cells, each
of which is highlighted by a form under the continuous mapping:

244 Luisa Consiglieri – Morphocontinuity in the work of Eero Saarinen

 h : A [1,1] u [1,1] o IR
defined as:
h( x, y) sin (a x ) cos(b y)

for some constants a, b  IR (compare to the FresH2O pavilion, Netherlands, 1998, by

Nox team and Kas Oosterhuis [Consiglieri and Consiglieri 2009]). The intertwining of
functional spaces and facades recreates a cyclical and shared feeling around the building.
The TWA Flight Center
The roof of the TWA terminal is constituted by four free-flowing shell structures
suggesting flight and having a thickness that varies from 20 cm to 1.10 m (fig. 6). This
variation reveals the development of new structures and formal systems. The flying shells
are continuous forms of the kind of non-poids in architecture theorised by modernist
European architects [Robichon 1965]. The various interior levels are linked by stairs and
sculptural galleries. This fluid space transforms our visual perception. Indeed, the created
cells are provided by an emotional movement because the images change under our visual
perception. This movement can be mathematically interpreted by the time parameter in
eq. (1). The organization is commonly compared to the cyclical nature of the Möbius
strip folding back onto itself. The application of topological deformation of a surface can
lead to the intersection of external and internal planes in a continuous morphological
change and insert differential fields of space and time into an otherwise static structure
[Emmer 2004: 65].

Fig. 6. Eero Saarinen, TWA Flight Center, JFK International Airport.

Drawing by Teotónio Agostinho
In addition to the glass in this building acting to dissolve the boundary between
interior and exterior, as it did in the other two buildings considered earlier, here the

Nexus Network Journal – Vol.12, No. 2, 2010 245

different sized surface units are distributed such that the modulation of the interior and
exterior spaces becomes irrelevant. Moreover, they organize the internal environment and
maximize the penetration of sunlight. The final object is constituted not only by different
forms but also by different cells, which include the interior of such forms.
Conclusions
In order to interpret the chosen works conveniently, we consider two types of
interpretations for the structure:
1. the structure is understood in terms of its physical characteristics and
mathematical methods. In this point of view, the structure is defined by
the characteristics of the materials and the loads;
2. the structure loses its importance in relation to the final object and is
driven by mathematical precision and clarity.
The works studied here have roofs of different forms. Although they are all limited by
the use of concrete, the three buildings show that the dynamic search is possible. We can
conclude that the study of eurhythmy is based on the concepts of continuity and
differentiability of the object itself and continuous movement between interior/exterior.
As far as contemporary interpretations of architectural space (namely utopian space
[Roisecco 1970: 52-53, 364-445], existential space [Norberg-Schulz 1977: 430-433],
emergent, biomorphic or organic spaces [Architectural Design 2008]) attempt to imitate
the biological behaviour of self-organization, along with the simultaneous, ongoing
search for new materials that attends these new concepts, these concepts are still included
in our definition of morphologic rhythm and consequently in morphocontinuity.
Saarinen’s oeuvre shows that designing morphogenetic buildings inspired by fractals
is not the only way to be create a perfect symbiosis between nature and functional
behaviour. At a macroscopic level, the concrete structures seem to be modeled with the
same plasticity as contemporary architectural objects using the new materials inspired by
nanotechnology. Structurally, concrete is relatively economical, in that it affords strength
through curvature and form. The goal is to enhance the surface’s contact with air and sun
through characteristics such as a maximisation of the surface-to-volume ratio. Although
these three buildings were built before the invention of today’s biomaterials, they still
demonstrate the capacity to affect the flow of rainwater and wind as well as to absorb
solar energy. Moreover, because the buildings cannot successfully exist in isolation, the
aesthetic value of the objects discussed here is still influential today, since the objects are
situated in the dynamic centre of even more dynamic surroundings.
The equilibrium of the composition of the object is achieved through the continuity
and differentiability of its elements. The main difference between the equilibrium of the
auditorium and that of the other buildings is that the auditorium is based on an
Euclidean form, while the cells of other buildings are non-Euclidean. However, the
equilibrium of their spatial elements of all three buildings is based on the fluidity
between the inside and the outside.
Acknowledgement
The authors gratefully thank Teotónio Agostinho for figs. 4-6.

246 Luisa Consiglieri – Morphocontinuity in the work of Eero Saarinen

References
L’architecture d’aujourd’hui. 1956. Auditorium du M.I.T., Cambridge, U.S.A., Eero Saarinen et
Associés, Architectes. L’architecture d’aujourd’hui 64, 27 (Mars 1956): 50-53.
Architectural Design . 2008. Neoplasmatic design. Architectural Design 78, 6.
ARGAN, G.C. 1955. Brunelleschi. Arnoldo Mondadori Editore (BMM 415), Veronesi.
CONSIGLIERI, L. and CONSIGLIERI, V. 2006. Structure of phenomenological forms: morphologic
rhythm, Nexus Network Journal 8, 2 (October 2006): 101-110.
———. 2009. Continuity versus Discretization. Nexus Network Journal 11, 2 (2009): 151-162.
CONSIGLIERI, V. 1994. A morfologia da arquitectura 1920-1970, vol. 2. Ed. Estampa, Lisbon.
EMMER, M. 2004. Mathland: from flatland to hypersurfaces. Birkhäuser, Basel.
FRANKL, P. 1981. Principios fundamentales de la Historia de la Arquitectura. El desarrollo de la
arquitectura europea: 1420-1900. Ed. Gustavo Gili, S.A., Barcelona 1981. Translated by Die
Entwicklungsphasen der Neueren Baukenst. Ed. B.G. Teubner, Stuttgart 1914.
NORBERG-SCHULZ, C. 1977. La signification dans l'architecture occidentale. Pierre Mardaga
editeur, Bruxelles. Translated by Significato nell'architettura occidentale. Electra Editrice,
Milano 1974.
ROBICHON, R. 1965. Vers le non-poids en architecture. In Structures nouvelles en architecture.
Conservatoire National des Arts et Métiers, Paris (Avril 1965), 10-11.
ROISECCO, G. 1970. Spazio, evoluzione del concetto in architettura. Mario Bulzoni Editore,
Roma.
WRIGHT, S.H. 1989. Sourcebook of Contemporary North American Architecture: From Postwar
to Postmodern, Van Nostrand Reinhold, New York.

About the authors

Luisa Consiglieri received her B.S., M.S. and Ph.D. degrees in mathematics from University of
Lisbon in 1988, 1992 and 2000, respectively. She taught at the Faculty of Sciences of University of
Lisbon from 1987 until her retirement in April 2009. Her research interests span a variety of areas
in differential equations such as existence, uniqueness and regularity of solutions, fluid mechanics,
heat transfer, electromagnetism, biomechanics, and applications to biomedical problems. At the
present her principal interest is interdisciplinarianism.

Victor Consiglieri received a B.S. degree in architecture from Escola Superior das Belas-Artes de
Lisboa (ESBAL) in 1956, and a Ph.D. degree in morphology of architecture from Faculty of
Architecture of Technical University of Lisbon in 1993. He was in Paris 1964-65 on a scholarship
from Centre Scientifique et Technique du Bâtiment (CSTB). He was in Camâra Municipal de
Lisboa 1956-62, Ministério do Ultramar 1962-66, Caixa da Previdência 1966-76, Faculty of
Architecture of Technical University of Lisbon 1976-97 and as invited professor in University of
Évora 2004-05. He realized many projects for public buildings such as kindergartens, elementary
schools, institutions for youth, and centres for the elderly. He was member of Associação dos
Arquitectos and of Ordem dos Arquitectos 1956-2004. His current interest is the contemporary
aesthetic.

Nexus Network Journal – Vol.12, No. 2, 2010 247

 Tyler Sprague Research
University of Washington Eero Saarinen, Eduardo Catalano and
College of Built Environments
Seattle, WA 98103 USA the Influence of Matthew Nowicki:
[email protected] A Challenge to Form and Function
Keywords: Eero Saarinen, Abstract. Matthew Nowicki befriended Eero Saarinen at the
Eduardo Catalano, Matthew Cranbrook Academy and was succeeded as Chair of the
Nowicki, modern architecture, School of Design at North Carolina College of Design by
hyperbolic paraboloids, saddle Eduardo Catalano. Nowicki’s influence is evident in
shapes subsequent work of these two architects. Themes of function,
structure and humanism resonated differently in each. All
three of these interconnected individuals were engaged in the
same intellectual milieu, each manifesting his own
architecture in a unique yet contextual way. Taken as a
whole, their endeavors stand as evidence of the shifting
understanding of what modern architecture was about.

Introduction
In the years following World War II, modern architecture was a facing a cross-roads.1
Many architects were skeptical of the ability of the dominant “International Style”2 to
respond to social, humanistic demands – a sentiment magnified by the loss of life during
the war – and were searching for a new direction forward. The self-justified rational
architecture did not engage the richness of human experience. Multiple calls for
“expression” and “monumentality” in architecture revealed a need for more human
engagement than the International Style was providing [Barr 1948].
Thus post-war modernist architects had to make a choice: maintain the rigid
obedience to “functionalist” architecture, or search for a new means of expression. They
were consciously engaged in re-thinking what architecture should be, interested in
recasting modern architecture to suit an altered social landscape; they wanted to make
architecture more appealing, but didn’t want to abandon industrial efficiency. They
sought to re-place architecture in the minds of citizens, but felt the “immaturity of
modern architecture”, and the need to “grow up”.3 Modernism had not failed, but it
needed revision.
One architect who engaged in this discussion and would prove influential in altering
its course was Matthew Nowicki. His untimely death in an airplane crash in the fall of
1950 cut his budding career short, but not before he had written a few influential articles
and designed one seminal work, the Livestock Pavilion in Raleigh, North Carolina.
Nowicki’s work suggested a new direction for architecture in the post-war world – ideas
that would fatefully remain unrealized by him, but would be picked up and extended
other like-minded architects.
During his life, he befriended the architect Eero Saarinen at the Cranbrook Academy
and was succeeded as Chair of the School of Design at North Carolina College of Design
by Eduardo Catalano.4 Through the subsequent work of these two men, Nowicki’s
influence – his personality, his inventiveness, his ideas about architecture – can be seen.
Themes of function, structure and humanism resonated differently in each man, striking
different chords in their work. Together, these three interconnected individuals, with

Nexus Network Journal 12 (2010) 249–258 Nexus Network Journal – Vol.12, No. 2, 2010 249
DOI 10.1007/s00004-010-0033-5; published online 6 May 2010
© 2010 Kim Williams Books, Turin
great similarities and vast differences, were engaged in the same intellectual milieu, each
manifesting his own architecture in a unique yet contextual way. Their combined
endeavors, physical and intellectual, stand as evidence of the shifting understanding of
what modern architecture was about.
Matthew Nowicki
Matthew Nowicki was born in 1910, into a Polish noble family, the son of a
politically active Consul to the Polish State. He traveled extensively as a child, spending
much time in Chicago, and learning English before enrolling in architectural studies at
the Warsaw Polytechnic in 1928. Here he quickly demonstrated an incredible drawing
ability, a talent that would provide his most lasting legacy in his striking sketches that
remain. Instruction at the Polytechnic also strove to “teach him to see things as
structures. To this end a drawing was built, with skeletons of structural lines exposed”
[Mumford 1954a: 142]. Drawing for Nowicki was not just symbolic representation, but
an intellectual process of architectural and structural synthesis. Engineering and geometry
would always go hand-in-hand with his formal, architectural explorations. Nowicki also
met his wife, Sasha, while in school; a fellow architecture student, she was by all measures
his equal in drawing and design.
Inspired by Le Corbusier and Wright, Nowicki began his own professional practice in
Warsaw after graduation, also accepting a position as associate professor at the
Polytechnic. He built a number of churches, homes and sports arenas in Poland before
the German invasion in September 1939. The darkness of the war descended on Warsaw
as the mass destruction increased, and Nowicki and his family were forced to escape to
the distant mountain regions. After the war, Nowicki was involved in the planning to
rebuild Warsaw, but when the Polish government was taken over by Communist powers,
he decided to come to the United States [Brook 2005: 37]. He served as Cultural Attache
to the Polish Consulate in Chicago, was a visiting critic at the Pratt Institute, and served
a formative role as the Polish representative to the committee for the United Nations
Building.5
The wartime experience drastically effected Nowicki’s outlook on life, and revised his
approach toward architecture. After the war he wrote:
The study of the well-being of contemporary man, which has been
introduced into the language of architecture, continues to be the
inspiration for our work but this time the quality is differently analyzed. It
is no longer ‘the machine to live in’ that stirs our imagination. It is the
eternal feeling of a shelter to which we subordinate our creative ideas
[Mumford 1954: 148].
Nowicki invokes a more humanist approach to architecture, one that is sensitive to
emotional feelings as well as function.
His technical education at the Warsaw Polytechnic also rooted his architectural
theories in structural realities. Advancing technologies, as they changed over time, were a
crucial part of creating effective architecture. He stated:
...we now rely in our expression of the potentialities of materials and
structures. This interest in structure and material may find within the
building medium decorative qualities of ornament that are much too
involved for the purist of yesterday. The symbolic meaning of a support

250 Tyler Sprague – Eero Saarinen, Eduardo Catalano and the Influence of Matthew Nowicki: A Challenge ...
 has been rediscovered, and a steel column is frankly used as a symbol of
structure, even when it is not part of the structure itself [Nowicki 1951:
279].
For Nowicki, structure had become expression and he embraced the new potential
that emerging technologies suggest.6 A new type of expression was emerging from an
awareness of material and structure, one that was not rooted in a singular, formal
prescription but rather encouraged multiple investigations. He states, “Art may be one,
but it has a thousand aspects. We must face the dangers of the crystalizing style... trying
to enrich its scope by opening new roads for investigation and future refinement”
[Nowicki 1951: 279].
In an article titled “Origins and Trends in Modern Architecture,” Nowicki clearly
stated what many architects had begun to feel regarding functionalist architecture:
In the growing maturity and self-consciousness of our century, we can not
avoid the recognition ... that the overwhelming majority of modern design
form follows form and not function. And even when a form results from a
functional analysis, this analysis follows a pattern that leads to a discover of
the same function, whether in a factory or a museum [Nowicki 1951:
273].
Striking to the heart of the proclaimed objectivity of pre-war modern architecture,
Nowicki’s statements resonated with many architects of his generation. From Paul
Rudolph [1986: 153] to Colin Rowe [1976: 130], Nowicki’s statements provided a
springboard for a new line of architectural thinking, causing, in 1962, the critic Allan
Temko to call Nowicki the “spokesman for young Modernists” [Temko 1963: 43]. For
the emerging generation of architects, function influenced but did not dictate form. For
Nowicki, Catalano and Saarinen, post-war architecture would engage the mutual
dependence of function and form.
Nowicki’s theories can be seen in his sketches for the State Fair Livestock Judging
Pavilion (later the Dorton Arena) in Raleigh, North Carolina. His death during the
design process left architect William Henley Deitrick, working with engineers Severud-
Elsted-Kreuger, to complete the project. Despite the intent to do everything “as Matthew
would have wanted it” [Parabolic Pavilion 1952: 137], substantial design changes due to
budget and construction issues altered the building to the point where some questioned if
Nowicki would have been pleased with it (North Carolina Dean Henry Kamphoefner
quoted in [A radical settles down in Raleigh, NC 1980]. But Nowicki’s intent is
evidenced through his sketches, and is indicated in the built work.
The sketches (and the building itself; fig. 1) are structurally bold, consisting of two
intersecting parabolic arches. The sweeping, mathematically-driven forms are angled to
the horizon, and on this account, trace the plan of the arena in the area between them.
The roof consists of draped cables strung between the two arches. The cables’
dependence on the geometry of the arches, coupled with their own catenary behavior,
creates a warped roof surface displaying a curvature in both lateral and longitudinal
directions. This is clearly a geometrical investigation of shapes in space, a mixing of
elevation and plan, but it is also underlined by a structural logic.
The two parabolic arches were made of concrete, and act in pure compression, with
the roof cables hung in pure tension. The force-imbalance of the canted arches was to be
ideally counteracted by the thrust of the roof cables, freeing up the curtain wall below to
provide only stability. In the final building, construction methods had not advanced

Nexus Network Journal – Vol.12, No. 2, 2010 251

enough to support these intentions, and the exterior curtain wall became load bearing.
These structural innovations have been widely cited by people such as Frei Otto [1954]
as innovative and suggesting a new field of architectural form. The Livestock Pavilion
displayed a material logic employed through the mathematics of complex geometry.

Fig. 1. Matthew Nowicki’s State Fair Livestock Judging Pavilion (later the Dorton
Arena) in Raleigh, North Carolina. Photograph by Yoshito Isono, reproduced
courtesy of Nicholas Janberg and Structurae
These structural forms also provided an expression of the function of the building
and shaped the space within. As a “single great room” the livestock show floor and
surrounding grandstands mirror the structure above, providing axiality and a center of
focus (Paul Rudolph quoted in [The great Livestock Pavilion complete 1954]).
Architectural Forum stated:
Nowicki was seeking first of all not for a unique structure but for a unique
space. The remarkable warping of the space upward, the exact reverse of a
dome, would guarantee maximum daylight admitted from the two sides to
the central arena. This labile kind of curvature of enclosed space marks a
new epoch in architecture [Parabolic Pavilion 1952].
The innovative, three-dimensional aspects of the Pavilion indicated a different means
of enclosing human experience, a new spatial relationship within architecture. The roof
provided shelter and also engaged the questioning mind.
Nowicki’s theories of humanist expression through structure are in play here. He is
demonstrating a navigated position between the human experience, structural rationality
and the functional demands of the building. Although not without shortcomings,7 in this
single work he has embodied many different strands of the architectural discourse of the
time. Paul Rudolph described a sublime experience, and stated that this new space
“helped man forget something of his troubles.” He also claimed that it satisfied “our need
for more expressiveness to emphasize our places of worship, meeting places of governing
bodies, and centers of recreation” [The Great Livestock Pavilion Complete 1954: 132]. It
embodied a “basic soundness and high spirited boldness” that indicated a new direction
for modern architecture [The Great Livestock Pavilion Complete 1954: 134].

252 Tyler Sprague – Eero Saarinen, Eduardo Catalano and the Influence of Matthew Nowicki: A Challenge ...
Eduardo Catalano
After Nowicki’s death, his position as chair of the North Carolina State College,
School of Design was soon filled by Dean Henry Kamphoefner, with Eduardo Catalano.
Catalano, an Argentinian-born architect who trained at both the University of
Pennsylvania and under Walter Gropius at the Harvard Graduate School of Design, was
keenly interested in advanced geometrical forms in architecture. Prior to coming to
North Carolina, he had co-written a book on the mathematics of geometrical forms and
the use of perspective [Crivelli, Nery and Catalano 1940]. His un-built auditorium
project in Buenos Aires used a thin shell structure to enhance acoustics, and he entered
several competitions for pre-fabricated housing solutions, utilizing a modular approach to
building [Arts and Architecture’s Second Annual Competition 1945].
But it all appears to come together for Catalano once he arrives in Raleigh. Along
with the continuing construction of the Livestock Pavilion, Nowicki’s presence was still
felt through the curriculum and education system he had initiated. His pedagogical
approach to teaching architecture emphasized a merge of the technical requirements with
a humanistic awareness. Influenced by Le Corbusier’s “Modulor”, Nowicki emphasized
designing around “Man” as the “unchanging module of of scale and proportion” and the
role of technology and structure as a means to satisfy changing human demands
[Mumford 1954b]. This legacy is digested and synthesized by Catalano, filtered through
his own experiences and disposition, in a unique way.
Catalano’s work focused on geometrically advanced forms, investigating new means
of spanning space. He worked primarily with hyperbolic paraboloids, investigating
geometrical surfaces with curvature in two longitudinal directions (e.g., saddle shapes) –
forms like Nowicki’s Livestock Pavilion roof. But Catalano extended this form, coming
up with a modular system to create these shapes using individual, linear elements.
Drawing on his experience with advanced geometry, Catalano developed a system where
new shapes could be generated by simply modifying parameters of the design. His project
Structures of Warped Surfaces explored various combinations of hyperbolic paraboloid
forms with a variety of supports, searching for new ways to provide an over-head surface
[Catalano Gubitosi Izzo 1978: 55-70]. In 1953, three years after Nowicki’s death,
Catalano was quoted stating:
Following the research begun by Nowicki, our work at North Carolina has
gone far beyond the Raleigh Pavilion in the study of both space structures
and repetitive spatial structural systems ... these have led to interesting
structures (comment by Eduardo Catalano, in [Is this Tomorrow’s
Structure? 1953: 160]).
Catalano also utilized war-related technologies of aluminum and plywood in
experiments with these forms, stressing innovation tied to industrial production.
His work at North Carolina State College culminated in the construction of his own
house in the woods outside Raleigh. Known simply as the “Catalano House” (1953-55,
fig. 2) it consisted of a single hyperbolic paraboloid, made up of three layers of laminated
timber. Spanning 90 feet between supports with a total thickness just over two inches,
this house was celebrated as “a structure that is all skin and no bones,” reflecting “the
most advanced engineering know-how” of the time “[Why are People Talking about this
House? 1955]. The functions of house take place then beneath the shell, with full height
glass curtain walls defining the interior space.

Nexus Network Journal – Vol.12, No. 2, 2010 253

 Fig. 2. Eduardo Catalano’s Catalano House
But Catalano’s house was recognized as an experiment, an attempt to utilize
advancing “skin technology in architecture” “[Why are People Talking about this House?
1955]. Given his approach to design, he was faced with the awkward challenge of fitting
an architectural function to the forms he had developed through mathematics. The
majority of his “warped surface” sketches contain no indicator of relation to the human
scale, nor to which functions or building types they would be most suited. Nowicki’s
work clearly influenced his explorations, but Catalano has placed new emphasis on
geometrical “purity” and advancement in structural engineering without direct ties to a
specific architectural program [Catalano 2009]. His “warped plane roof” became famous
with both “avant-garde aestheticians and building technicians ” “[A New Way to Span
Space 1955]. It is a shape like a potato chip but is also mathematical, geometrical, and
analytical: a universal shape for covering.
Eero Saarinen
The parallel work of Eero Saarinen reveals a different aspect of Nowicki’s influence.
Saarinen and Nowicki met at a symposium in February 1948, and in the summer of
1949, Nowicki was appointed Visiting Professor at the Cranbrook Academy.8 United by
a propensity for drawing, the two were fast friends and shared a fruitful summer
designing the campus plan for Brandeis University [Merkel 2005: 105]. Their sketches
show heavily sculptural forms, with undulating walls and domed spaces, which, though
not without precedent, mark a formalistic departure from Saarinen’s previous work, like
the GM Technical Center. With Nowicki by his side, Saarinen is able to merge his
sculptural interests developed in furniture with the architectural humanism of Nowicki.
Though their plan for Brandeis went unrealized, their sketches reveal many hints of
Saarinen’s future work.
Saarinen would later acknowledge the influence of their brief time together, which
lasted only a few months. In a letter, he declared Matthew Nowicki to be the third most
significant influence on him after his father, Eliel, and his life-long collaborator Charles
Eames [Saarinen Pelkonen and Albrecht 2006: 332]. In Nowicki’s obituary in
Architectural Forum, Saarinen stated: “If time had allowed his genius to spread its wings
in full, this poet-philosopher of form would have influenced the whole course of

254 Tyler Sprague – Eero Saarinen, Eduardo Catalano and the Influence of Matthew Nowicki: A Challenge ...
architecture as profoundly as he inspired his friends” [Mumford 1950]. Through these
statements, we can see that Nowicki had a profoundly different effect on Saarinen than
on Catalano.
After his experience with Nowicki, Saarinen’s work is defined by his formal
expression, in such work as the MIT Chapel and Kresge Auditorium, and the TWA
Flight Center and Dulles Airport Terminals [Serraino Saarinen and Gössel 2005]. The
structural logic of his architecture is often complicated by his attempts to create a
meaningful form, one that communicates as well as functions [Pelkonen 2006]. Inspired
by seeing Nowicki as a poet-philosopher of form, Saarinen was not interested in pure
structure or engineering, like Catalano, but in linking architectural form to context and
broader, humanist theories. Geometry alone was not enough; it needed energizing if it
“was to serve the spatial-structural-spiritual totality” that he wished to express [Temko
1962: 43]. This is architecture that engages Nowicki’s ideas of the multiplicity of art.

Fig. 3. Eero Saarinen’s David S. Ingalls Ice Rink at Yale University, New Haven, CT. Photograph
by Yoshito Isono, reproduced courtesy of Nicholas Janberg and Structurae
Nowicki’s influence is most clearly seen at Saarinen’s Ingalls Ice Rink (fig. 3). With a
program similar to the Livestock Pavilion, Saarinen gives expression to the Ice Rink as a
“single room” with a central rink surrounded by grandstand seating. A single, long
concrete arch swoops over the long axis of the rink, with catenary cables running
perpendicular to either side, forming the roof surface. But in an artistic and functional
move, Saarinen does not terminate the main arch at its support, but reverses its curvature
and cantilevers an additional portion to serve as an awning over the building entrance.
This extended curve gives the building an undulating quality, making it a graceful active
presence on the north side of the Yale University campus [Yale’s Hockey Rink 1958].
Exposed on the interior, the smooth concrete arch contrasts with the wood plank finish
of the roof, coming to its peak over the hockey rink below. Commonly called the
“dinosaur” or likened to a Viking ship [Yale’s Viking Vessel 1958], the building has
“personality” and engages people spatially and intellectually.

Nexus Network Journal – Vol.12, No. 2, 2010 255

 Not only did Saarinen use a structural system similar to that of the Livestock Pavilion
(as well as hiring the same engineer) but he captured a similar dynamic movement that
can be seen in Nowicki’s sketches. His structure dominates the form, serving the purpose
of providing enclosure and supporting function, but clearly standing as a unique, artistic
creation. The Ice Rink was criticized as not “sensible” because it spanned the arch over
the long dimension (rather than the short axis) but it highlights the fact that Saarinen
was not interested in a prescribed, “logical” approach to structure; the central spine
served more than just a structural purpose.9 He was not interested in the geometrical
warping of overhead space as an isolated experiment, but in its inclusion, modification
and distortion in the overall architectural experience.
Conclusion
The work of Eduardo Catalano and Eero Saarinen reveals different themes in post-
war architecture. Ranging from systematic, geometrical form-finding to expressive,
artistic architecture, these architects display different directions following their experience
with Matthew Nowicki. The sketches, writings, and built work of one man were
simultaneously described as the investigation of spatial structures, and the work of a poet-
philosopher, spawning different yet related directions in post-war architecture. This study
indicates the nature of modern architectural discourse - as the action of neither isolated,
autonomous individuals nor a unified group. Nowicki’s influence does not at all indicate
a lack of originality or innovation on the part of either architect, but serves to show how
architectural discovery is often contextual and interrelated to broader discourses. By
discussing the milieu of theories and shapes, functions and forms, this comparative study
reveals the changing landscape of ideas about how to design modern architecture in the
post-war period.
Lewis Mumford’s comments following Nowicki’s death in 1950 appear prophetic yet
misleading. A friend and proponent, he stated that Nowicki
bore within him the seed of a new age. In his designs, spontaneity and
discipline, power and love, form and function, mechanical structure and
symbol, were united. What he left undone through his death must now
call forth the creative efforts of a whole generation [Mumford 1950: 201,
206-207].
Not simply completing Nowicki’s work, Catalano and Saarinen forged their own
ways forward, inevitably leaning on their own experiences and influences, on the paths to
creating their own unique architecture.
Notes
1. For a much longer discussion of the post-war architectural scene, see [Goldhagen and Legault
2000].
2. The title of a 1932 Exhibition at the Museum of Modern Art, followed by the publication
[Hitchcock and Johnson 1932].
3. Lewis Mumford used the terms “immaturity” and “grow-up” during the 1948 symposium
“What is Happening to Modern Architecture” [Barr 1948].
4. Nowicki was returning from a trip to Chandigarh, India where he was designing a new capital
complex with Albert Meyer. This project would be later taken up by Le Corbusier.
5. Nowicki is credited, along with Le Corbusier and W. K. Harrison, as having a significant
impact on the UN design. See [UN General Assembly 1952: 141].
6. These ideas are very similar to Louis Kahn’s claims in his article “Monumentality” [1944].
7. Problems with the acoustics and waterproofing of the Pavilion proved difficult to solve.

256 Tyler Sprague – Eero Saarinen, Eduardo Catalano and the Influence of Matthew Nowicki: A Challenge ...
8. They were both invited to the “What Is Happening to Modern Architecture” symposium at
the Museum of Modern Art, organized by Alfred H. Barr and Henry-Russell Hitchcock.
Cranbrook position cited in [Nowicki and Schafer 1973: xi].
9. Quote from Robert Venturi in “Appreciations by Former Collaborators Panel Discussion” in
[Saarinen, Pelkonen, Albrecht 2006: 361].

References
A New Way to Span Space. 1955. Architectural Forum 103 (November 1955): 170-177.
A radical settles down in Raleigh, NC. 1980. AIA Journal 69, 11 (September 1980): 54-61.
Arts and Architecture’s Second Annual Competition for the Design of a Small House. 1945. Arts
and Architecture 62 (Feb. 1945): 28-41.
BARR, Alfred H. 1948. What Is Happening to Modern Architecture?: A Symposium at the
Museum of Modern Art. MOMA Bulletin XV, 3. New York: Museum of Modern Art.
BROOK, David Louis Sterrett. 2005. Henry Leveke Kamphoefner, the Modernist: Dean of the
North Carolina State University School of Design, 1948-1972. Master’s Thesis, Norrth
Carolina State University. http://www.lib.ncsu.edu/theses/available/etd-07252005-
164332/unrestricted/etd.pdf.
CATALANO, Eduardo, Camillo GUBITOSI, and Alberto IZZO. 1978. Eduardo Catalano: buildings
and projects. Rome: Officina.
CATALANO, Eduardo. 2009. Interview by the author. 9 December 2009.
CRIVELLI, Oscar F., Rene NERY, and Eduardo Fernando CATALANO. 1940. Teoria de las sombras y
trazados de perspectiva. Buenos Aires: Francisco A. Colombo.
GOLDHAGEN, Sarah Williams and Réjean LEGAULT. 2000. Anxious Modernisms: Experimentation
in Postwar Architectural Culture. Montréal: Canadian Centre for Architecture.
HITCHCOCK, Henry-Russell, and Philip JOHNSON. 1932. The International Style: Architecture
since 1922. New York: W. W. Norton & Company. Rpt. 1995, New York: Norton.
KAHN, Louis. 1944. Monumentality. Pp. 570-579 in The New Architecture and City Planning,
Paul Zucker, ed. New York: New York Philosophical Library.
MERKEL, Jayne. 2005. Eero Saarinen. London: Phaidon.
Is this Tomorrow’s Structure? Architectural Forum 90 (February 1953): 150-160.
MUMFORD, Lewis. 1950. From the legacy of Matthew Nowicki. Architectural Forum October
1950: 200-201.
———. 1954a. The Life, Teaching and Architecture of Matthew Nowicki. Architectural Record
115 (June 1954): 129-139.
———. 1954b. Matthew Nowicki as an Educator. Architectural Record 116 (July 1954): 128-
135.
NOWICKI, Matthew. 1951. Origins and Trends in Modern Architecture. Magazine of Art 44
(November 1951): 273-79.
NOWICKI, Matthew, and Bruce Harold SCHAFER. 1973. The Writings and Sketches of Matthew
Nowicki. Charlottesville: University Press of Virginia.
OTTO, Frei. 1954. Das hängende Dach (The Hanging Roof). Berlin: Bauwelt Verlag der Ullstein.
Parabolic Pavilion. 1952. Architectural Forum 97 (October 1952): 134-139.
PELKONEN, Eeva-Liisa 2006. The Search for Communicative Form. Pp. 83-07 in Eero Saarinen:
Shaping the Future. Eero Saarinen, Eeva-Liisa Pelkonen, and Donald Albrecht. New Haven:
Yale University Press.
ROWE, Colin. 1976. Mathematics of An Ideal Villa and Other Essays. Cambridge MA: MIT Press.
RUDOLPH, Paul. 1956. The Six Determinates of Architectural Form. Architectural Record 120
(October 1956): 183.
SAARINEN, Eero, Eeva-Liisa PELKONEN, and Donald ALBRECHT. 2006. Eero Saarinen: Shaping the
Future. New Haven: Yale University Press.
SERRAINO, Pierluigi, Eero SAARINEN, and Peter GÖSSEL. 2005. Eero Saarinen, 1910-1961: A
Structural Expressionist. Köln: Taschen.
TEMKO, Allan. 1962. Eero Saarinen. Makers of contemporary architecture. New York: G. Braziller.
The Great Livestock Pavilion Complete. 1954. Architectural Forum 100 (April 1954): 130-134.

Nexus Network Journal – Vol.12, No. 2, 2010 257

UN General Assembly. 1952. Architectural Forum 97 (October 1952): 140-149.
Yale’s Hockey Rink. 1958. Architectural Record 124 (October 1958): 151-158.
Yale’s Viking Vessel. 1958. Architectural Forum 109 (December 1958): 106-111.
Why are People Talking about this House? 1955. House and Home, August 1955: 94-100.

About the author

Tyler Sprague is a doctoral student at the University of Washington, Seattle. He has a background
in structural engineering and continues to study architecture and engineering in the post-war
period. He would like to thank Meredith Clausen and Alex Anderson for their guidance, and
acknowledge support from the University of Washington Victoria N. Reed Endowed Student
Support Fund.

258 Tyler Sprague – Eero Saarinen, Eduardo Catalano and the Influence of Matthew Nowicki: A Challenge ...
 Rachel Fletcher Research
113 Division St. Eero Saarinen’s North Christian
Great Barrington, MA
01230 USA Church in Columbus, Indiana
[email protected]
Abstract. Eero Saarinen’s North Christian Church, an
Keywords: Eero Saarinen, North important contribution to post-war liturgical church
Christian Church, descriptive architecture, serves a community of Disciples of Christ in
geometry Columbus, Indiana. Early design sketches illustrate
elementary geometric shapes and symbols – triangle, square,
cross, hexagon, and octagon – whose proportions appear in
the plan and section of the completed structure.

Introduction
…they should feel they are all in unity and harmony in a special and
appropriate spiritual atmosphere.
Eero Saarinen, on North Christian Church (1960) [Saarinen 1962: 88]
An important contribution to postwar liturgical church architecture, Eero Saarinen’s
North Christian Church serves a northern residential suburb of Columbus, Indiana.
Designed between 1959 and 1961 and completed in 1964, the central plan church
accommodates greater participation and more intimate and democratic congregational
involvement.
Saarinen desired a simple structure that would support the liturgical needs of its
members and “clearly and logically express the form and character of the church” (on
North Christian Church (1960) [Saarinen 1962: 88]). Its open structure, facilitated by
new advances in steel construction, dissolves the traditional distinction between officiate
and congregant, while recognizing communion and baptism as important liturgical
sacraments.1
Exterior
Saarinen designed the church at the request of J. Irwin Miller, following the
architect’s design for Miller’s private residence (1953-57).2 In 1955, forty-three members
of First Christian Church broke away to organize a more liberal institution affiliated with
the Disciples of Christ. In 1958, with Miller’s assistance, the group purchased a five and
a half acre tract of land and met in various locations until Saarinen’s church was
completed.3
The whole thing, all the planes, would grow up organically into the spire
– on North Christian Church (1960) [Saarinen 1962: 90]
The main level of the two story structure rests on a massive concrete base. The lower
level nestles into a landscaped berm held back to expose a moat-like light well. The
building is hexagonal in plan, elongated on the east-west axis. Six steel legs clad in lead-
coated copper secure a pyramidal slate roof, then converge at an apex, rise to a tapered
192-foot central spire and terminate at a five-foot gold-leaf cross (fig. 1) [Knight 2008:
163; Thayer 1999: 4].

Nexus Network Journal 12 (2010) 259–270 Nexus Network Journal – Vol.12, No. 2, 2010 259
DOI 10.1007/s00004-010-0034-4; published online 18 May 2010
© 2010 Kim Williams Books, Turin
 Fig. 1. North Christian Church, Columbus, Indiana. Transverse section. Image:
Eero Saarinen, Collection Manuscripts & Archives, Yale University
(Image No. 9733)
…put only the sanctuary above ground and make it the significant visual
and architectural thing.
– on North Christian Church (1960) [Saarinen 1962: 88]
Only the main sanctuary and a small baptismal chapel, both surrounded by an
ambulatory, remain above ground. Support facilities consolidate within a single unit
hollowed out in the basement below. The surrounding six-foot high berm conceals from
view everything but the roof and entrance. The result is a simple and distinctly singular
structure that appears to hover lightly above the ground, an illusion enhanced by recessed
exterior glass walls set back some twelve feet from the roof’s edge [Thayer 1999: 4].
Interior
...you should have to work for it and it should be a special thing.
– on North Christian Church (1960) [Saarinen 1962: 88]
Acts of entry and passage simulate a spiritual journey to another world. From the
parking area, one ascends, and then descends before reaching the building at ground
level. Once inside, steeper steps lead to the sanctuary.
…communion is a very important act and the congregation participates in it.
– on North Christian Church (1960) [Saarinen 1962: 88]
Each building level follows the same hexagonal plan. Above, the main east entrance
leads through a spacious vestibule to a large bowl-shaped sanctuary placed at the center
and elevated to emphasize its liturgical importance. At the very center, beneath an oculus
and raised on a dais, a communion table, containing twelve places arranged in two rows

260 Rachel Fletcher – Eero Saarinen’s North Christian Church in Columbus, Indiana
with a taller place at one end, represents Christ and His twelve disciples. The dais can be
rotated and repositioned on other occasions.
…everyone feels equal and joined together.
– on North Christian Church (1960) [Saarinen 1962: 88]
Surrounding the communion table on five sides, mahogany benches bring
congregants close to the service. Organ, choir benches and pulpit complete the round on
the sixth side.
…immersion should be given more dignity…viewed only by family and
close friends.
– on North Christian Church (1960) [Saarinen 1962: 90]
Free-standing walls on diagonal axes frame a secondary baptistery chapel at the west
end, where immersion baptism is performed in a small hexagonal pool of white tile set
into the floor. Seating faces inward toward the center. On other occasions, the pool is
covered.
…put all that activity downstairs. Maybe underground, hidden away…
– on North Christian Church (1960) [Saarinen 1962: 88]
Underground perimeter classrooms, offices and other support facilities face the berm,
lit by ambulatory windows at the base of the church. In the center, an auditorium is
placed directly below the sanctuary.
The primary element to create the right spiritual atmosphere would, of
course, be light.
– on North Christian Church (1960) [Saarinen 1962: 90]
Theatrical light sources lend mystery to the sanctuary experience. A hexagonal
skylight at the base of the spire focuses natural light on the communion table, an effect
augmented by recessed can lights in the ceiling nearby. Along the sanctuary’s perimeter,
indirect natural light from ambulatory windows reflects angled ceiling surfaces, causing
the roof to appear to float [Merkel 2005: 160; Thayer 1999: 5; Miller 2006: 65].
Geometric symbolism
…the total concept is carried down to the smallest detail.
– On Interior Design (1960) [Saarinen 1962: 11]
More than decoration or ornament, abstract geometric symbols convey spiritual
experience and meaning. Hexagonal figures represent the Star of David, or Magen David,
recognized universally as the symbol of Judaism (cf. [Fletcher 2005: 142-145]). The spire
represents for the architect “a marvelous symbol of reaching upward to God,”
commanding the landscape in its function as axis mundi. The cross at its apex represents
Christianity’s emergence from Judaic origins through Christ’s sacrifice [Miller 2006: 65].
Early design sketches contain a variety of elementary geometric shapes and symbols—
triangle, cross, square, hexagon and octagon—that contribute to the finished church plan
and section (fig. 2).

Nexus Network Journal – Vol.12, No. 2, 2010 261

 Fig. 2. North Christian Church, Columbus, Indiana. Sketches. Eero Saarinen,
Collection Manuscripts & Archives, Yale University (Image No. 9736)

Fig. 3. North Christian Church. Transverse section with geometric overlay. Base
image as in fig. 1; geometric overlay: Rachel Fletcher

262 Rachel Fletcher – Eero Saarinen’s North Christian Church in Columbus, Indiana
 Fig. 4. North Christian Church. Transverse section with geometric overlay. Base
image as in fig. 1; geometric overlay: Rachel Fletcher
Geometric analysis
Section
In section, the spire and roof converge along three equally spaced axes. In fig. 3, a
regular hexagon, hexagram (or Star of David), and equilateral triangle are drawn on these
axes. Roof lines coincide with two equal sides of a 120o isosceles triangle.
In fig. 4, a 45° isosceles triangle is drawn on the same base as the 120° isosceles
triangle. Its apex coincides with the gold-leaf cross at the top of the spire. From the apex
is drawn a circle, as shown, that encloses a regular octagon and an eight-faceted figure
composed of two squares. The two squares intersect along the plane of the oculus.

Nexus Network Journal – Vol.12, No. 2, 2010 263

Plan
The main level of the church features an elongated hexagon that encompasses berm-
framed moats on the long sides, north and south, and stepped concrete approaches on
the short sides, east and west. Just beyond, parallel diagonal lines delineate landscape
features and trees. Inside, concentric hexagons locate the building proper and sanctuary
floor (fig. 5).

Fig. 5. North Christian Church, Columbus, Indiana. Plan of main level. Image:
Courtesy Cranbrook Archives, Maurice Allen papers. Also Eero Saarinen,
Collection Manuscripts & Archives, Yale University (Image No. 9734)
Forty-five degree isosceles triangles locate the short sides of the hexagon on the east
and west (fig. 6).
Three axes, as shown, situate the extent of the hexagon and suggest the three-
dimensional frame of the cube (fig. 7).
The outer hexagonal footprint derives from the root-two proportions of a regular
octagon and an inscribed eight-faceted figure composed of two squares. The hexagon’s
short east-west sides coincide with two edges of the octagon. Its long north-south sides
are equal in length to the radius of the octagon’s circumscribing circle and follow the
edges of the squares (fig. 8).
Saarinen’s early sketches include squares and octagonal figures of a similar nature
(fig. 9).

264 Rachel Fletcher – Eero Saarinen’s North Christian Church in Columbus, Indiana
Fig. 6. North Christian Church. Plan of main level with geometric overlay. Base
image as in fig. 5; geometric overlay: Rachel Fletcher

Fig. 7. North Christian Church. Plan of main level with geometric overlay. Base
image as in fig. 5; geometric overlay: Rachel Fletcher

Nexus Network Journal – Vol.12, No. 2, 2010 265

 Fig. 8. North Christian Church. Plan of main level with geometric overlay. Base
image as in fig. 5; geometric overlay: Rachel Fletcher

Fig. 9. Sketches of North Xian Church, 1950s / Eero Saarinen, artist, graphite; 20
x 32 cm. Courtesy of the Aline and Eero Saarinen papers, 1906-1977, Archives of
American Art, Smithsonian Institution

266 Rachel Fletcher – Eero Saarinen’s North Christian Church in Columbus, Indiana
 …in any design problem, one should seek the solution in terms of the next
largest thing…
– On Relationships in Design (1958) [Saarinen 1962: 11]
Within the plan, the outline of the building proper repeats and relates proportionally
to the outer hexagonal footprint. Its shape follows a smaller, eight-faceted figure whose
circumscribing circle inscribes the original eight-faceted figure (fig. 10).

Fig. 10. North Christian Church. Plan of main level with geometric overlay. Base
image as in fig. 5; geometric overlay: Rachel Fletcher
Landscape
…a building grows from its site…
– On Architecture (1959) [Saarinen 1962: 6]
Saarinen’s partnership with Dan Kiley, a leading landscape architect of the modern
era, ensured a seamless integration of building and setting. Kiley proposed to transform
the flat treeless lot by surrounding the church with dense groupings of trees. A plot plan
published in Architectural Record projects the building’s proportions beyond the walls of
the church (cf. [Saarinen’s Church 1964: 187]).
The landscape plan for North Christian Church developed over several years and
never materialized fully. But thickets of magnolia trees were planted along the north and
south, and dwarf crabapple trees at the east and west entrances [Olivarez 2006: 274;
Thayer 1999: 6, 10]. In Saarinen’s early sketches, geometric lines and proportions extend
beyond the church structure (cf. [Aline and Eero Saarinen Papers 1906-1977, (Image
No. 5) AAA_saaralin_284659]). In the finished plan, geometric lines that delineate

Nexus Network Journal – Vol.12, No. 2, 2010 267

landscape features on the north and south parallel the long sides of the hexagonal
footprint and are tangent to the circle that passes through the points of intersection of
the two major squares (fig. 11).

Fig. 11. North Christian Church. Plan of main level with geometric overlay. Base
image as in fig. 5; geometric overlay: Rachel Fletcher
Saarinen’s last building
Solving the total design of North Christian Church did not come easily to Saarinen
and after two years the client grew impatient with the length of his process.4 In April
1961, the architect responded:
We have finally to solve this church so that it can become a great
building…so that as an architect when I face St. Peter I am able to say that
out of the buildings I did during my lifetime, one of the best was this little
church, because it has in it a real spirit that speaks forth to all Christians as
a witness to their faith [Saarinen 1962: 90].
In July 1961 Saarinen wrote that “we have finally solved the Columbus church”
[Saarinen 1962: 90]. Weeks later, shortly after being diagnosed with a brain tumor, he
died on the first of September.

268 Rachel Fletcher – Eero Saarinen’s North Christian Church in Columbus, Indiana
Conclusion
…organic unity is the ideal.
– On Interior Design (1960) [Saarinen 1962: 11]
It is not certain that Saarinen composed the spaces of North Christian Church by
adopting these proportions and techniques. Nor do we know if he merged hexagon and
square to symbolize the Star of David united with the Cross.
Saarinen recognized the merits of geometric proportion, but in unpublished notes
warned against adopting a singe canon or allowing one’s sense of proportion to “to be
frozen by rules and regulations” or confined to “divine and indisputable laws.” Good
design, he maintained, is an “integrated package” that incorporates “space, use, structure,
material, texture, proportions, and last but not least, the spirit of our time” (Saarinen,
“Golden Proportions,” unpublished notes for a lecture, 1953 [Pelkonen 2006: 342-
343]). Intuition, he proposed, is the best vehicle for capturing the total picture, more
than any single element.
Intuition can be nurtured by a working knowledge of proportional techniques. These
should serve a program’s spiritual, social and functional requirements, and respond
appropriately and organically to the situation at hand. The sublime beauty of North
Christian Church demonstrates this ethic to perfection.
Notes
1. According to Jennifer Komar Olivarez, the liturgical revival begun in late nineteenth century
Europe evolved through the influences of the 1938 treatise The Church Incarnate by German
architect Rudolf Schwarz; postwar modern churches of Schwarz and German contemporary
Dominikus Böhm; the 1947 German Liturgical Commission, and in the Catholic Church,
Second Vatican (Vatican II) reforms adopted in 1965 [Olivarez 2006: 267-268]; see also [Ray
2005: 3].
2. Miller represented the new church, acting as chair of the search committee that selected the
architect among several nationally known candidates.
3. [Knight 2008: 45; Merkel 2005: 158-159]. Architect Eliel Saarinen, Eero’s father, designed
First Christian Church (1939) in collaboration with his son. The National Historic Landmark
Nomination locates North Christian Church near the west end of a 13.5 acre property [Thayer
1999: 4].
4. By late January 1961, Saarinen had neither resolved the lantern nor related the interior and
exterior to his satisfaction [Saarinen 1962: 90].
References
Aline and Eero Saarinen Papers, 1906-1977. Archives of American Art, Smithsonian Institution.
Box 2 Folder 3 Sketches of North Xian Church, Eero Saarinen 1950s.
http://www.aaa.si.edu/collectionsonline/saaralin/container37465.htm
Eero Saarinen Collection. 1880-2004 (inclusive), 1938-1962 (bulk). Collection Manuscripts &
Archives. Digital Images Database. New Haven: Yale University.
http://images.library.yale.edu/madid/
FLETCHER, RACHEL. 2005. Six + One. Nexus Network Journal 7, 1 (Spring 2005): 141–160.
KNIGHT, RICHARD. 2008. Saarinen’s Quest: A Memoir. San Francisco: William Stout Publishers.
MERKEL, JAYNE. 2005. Eero Saarinen. London: Phaidon Press Limited.
MILLER, Will. 2006. Eero and Irwin: Praiseworthy Competition with One’s Ancestors. Pp. 57-67
in Eero Saarinen: Shaping the Future, Eeva-Liisa Pelkonen and Donald Albrecht, eds. New
Haven: Yale University Press.

Nexus Network Journal – Vol.12, No. 2, 2010 269

OLIVAREZ, Jennifer Komar. 2006. Churches and Chapels: a New Kind of Worship Space. Pp. 266-
275 in Eero Saarinen: Shaping the Future, Eeva-Liisa Pelkonen and Donald Albrecht, eds.
New Haven: Yale University Press.
PELKONEN, Eeva-Liisa and Donald Albrecht, eds. 2006. New Haven: Yale University Press.
RAY, MAIA LEA. 2005. Eero Saarinen: Creating Sacred Space. Master Thesis, University of
Louisville.
SAARINEN, ALINE B. 1962. Eero Saarinen on His Work. New Haven: Yale University Press.
Saarinen’s Church. 1964. Architectural Record 136 (September 1964): 185-190.
THAYER, LAURA, Louis Joyner and Malcolm Cairns. 1999. “National Historic Landmark
Nomination: North Christian Church.” United States Department of the Interior, National
Park Service. http://www.nps.gov/history/nhl/designations/samples/in/nchrist.pdf

About the author

Rachel Fletcher is a geometer and teacher of geometry and proportion to design practitioners. With
degrees from Hofstra University, SUNY Albany and Humboldt State University, she was the
creator/curator of the museum exhibits “Infinite Measure,” “Design by Nature” and “Harmony by
Design: The Golden Mean” and author of the exhibit catalogs. She is an adjunct professor at the
New York School of Interior Design. She is founding director of the Housatonic River Walk in
Great Barrington, Massachusetts, co-director of the Upper Housatonic Valley African American
Heritage Trail, and a director of Friends of the W. E. B. Du Bois Boyhood Homesite. She has been
a contributing editor to the Nexus Network Journal since 2005.

270 Rachel Fletcher – Eero Saarinen’s North Christian Church in Columbus, Indiana
 R. Balasubramaniam Research
Department of Materials and On the Modular Design of Mughal
Metallurgical Engineering
Indian Institute of Technology Riverfront Funerary Gardens
Kanpur 208 016, INDIA
Abstract. The modular designs of two significant funerary
[email protected]
gardens of the Mughal period, the Humayun’s tomb and Taj
Keywords : Humayun’s tomb, Mahal complexes, have been analyzed. The inherent
Taj Mahal, modular design, symmetry in the designs is made evident through an analysis
Arthasastra, Indian architecture, of the dimensions in terms of units mentioned in the
Mughal period, metrology Arthasastra, in particular the dhanus (D) measuring 108
angulams and vitasti (V) measuring 12 angulams, with each
angulam taken as 1.763 cm. The low percentage of errors
between predicted and actual dimensions has confirmed, for
the first time, that the modular designs of these Mughal
funerary gardens were based on Arthasastra units. A novel
mathematical canon to analyze the dimensions of Mughal
architecture has been set forth.

Introduction
There is great interest in understanding technical aspects of Islamic architecture in
India, with particular emphasis on the mathematical systems used in the design and
planning of Islamic structures. The geometry of the multiple axes as well as the geometry
of ratios that were used in the design is of interest. Basic to all these, is the understanding
of the units of measure to which Islamic structures of the subcontinent were designed
and finally constructed. The Mughal period, extending for about 200 years from 1526
A.D., is a significant period of the Indian subcontinent. In this present paper attention
will be focused on Mughal architecture, because presumably the elements of earlier
Islamic architecture of India were well reflected in significant Mughal structures.
Most Mughal architectural designs of palace and tomb complexes follow certain
patterns [Asher 1992: 19-251; Koch 2006]. The plans and designs of several Mughal
structures are known [Nath 1982-2005].
The basic aim of this article is to analyze the modular architectural design of two
important Mughal structures, the Humayun’s tomb complex in New Delhi [Misra and
Misra 2003: 15-70] and the Taj Mahal complex in Agra [Koch 2006], in order to
understand the measurement units to which these structures were conceived and
constructed. Ideas about metrology of Mughal structures can be only gleaned from actual
structures because there are no technical manuals on architecture and building art form
in the Islamic literature of the Mughal period [Koch 2006].
The two complexes chosen for this present study are good examples of the classical
form of well-planned riverfront funerary gardens of the Mughal period. There is one
basic inherent difference in the overall designs of these complexes, especially the relative
placing of the tomb with respect to the garden. While the tomb structure is centrally
located in the garden of Humayun’s tomb complex, the Taj Mahal mausoleum is placed
on a riverfront terrace that is located to the north of the garden. A brief introduction to
the garden types of the Mughal period will provide the background about the
significance of these two complexes.

Nexus Network Journal 12 (2010) 271–285 Nexus Network Journal – Vol.12, No. 2, 2010 271
DOI 10.1007/s00004-010-0014-8; published online 9 February 2010
© 2010 Kim Williams Books, Turin
Mughal gardens
The first type of Mughal garden is the terrace garden, like the ones the Mughals
developed at Kabul and Kashmir. This was a Central Asian concept, wherein the garden
was laid out on a slope, blending with the landscape of the region. The main buildings
were arranged on ascending terraces, placed symmetrically with respect to the central axis.
This axis was usually defined by a channel sunk in a paved walkway through which
collected water from a spring flowed. These individual terraces sometimes contained the
canonical four-part garden, as in the case of Shalimar garden of Kashmir.
The second type of Mughal garden is the traditional charbag garden. The chahar
bagh (meaning, in Persian, “four garden”) or its abbreviated form charbag designated a
cross-axial four-part garden. This kind of garden was in vogue in South Asia from even
much earlier times. For example, the large royal gardens of Sigiriya in Sri Lanka were laid
out in the fifth century A.D. in a cross-axial pattern [Bandaranayake 2000: 1-36].
An important point is the location of the built structure within the garden complex.
The tombs were usually centrally located. This type of modular design with a centrally
located tomb in the charbag garden finds the greatest expression at the great imperial
mausoleums of Humauyn (1530-1539 A.D. and 1555-1556 A.D.) at New Delhi, Akbar
(1556-1605 A.D.) at Agra and Jehagir (1605-1628 A.D.) at Lahore. While all these
structures form part of funerary garden complexes, the Humayun’s tomb complex will be
taken up for detailed analysis because this is a riverfront complex. The plan of
Humayun’s tomb complex is shown in fig. 1.

Fig. 1. Dimensions of overall complex of Humayun’s tomb at New Delhi. The vitasti (V) equals 12
angulams and each angulam measures 1.763 cm.

272 R. Balasubramaniam – On the Modular Design of Mughal Riverfront Funerary Gardens

 There was a variation noted in some riverside Mughal constructions, which was
necessitated by the luxuriant rivers available in India. Here, the building was shifted to
one side of the garden such that it faced the river. This theme finds its grandest
expression in the terraced riverfront charbag funerary garden of the Taj Mahal complex
of Shah Jahan (1628-1656 A.D.). This development of riverfront gardens in Mughal
India has been explained in great detail elsewhere [Koch 1997]. The plan of the Taj
complex in fig. 2 shows that the main building was placed on a rectangular terrace
(labelled T) running along the riverfront.

Fig. 2. Dimensions of overall complex of Taj Mahal in Agra. The vitasti (V) equals 12 angulams
and each angulam measures 1.763 cm.

Nexus Network Journal – Vol.12, No. 2, 2010 273

Metrology during the Mughal period
It is now important to focus on the metrology (study of the science of measurement)
of the Mughal period because the analysis of modular design of Mughal funerary gardens
will be illuminating only if the original units of measurements to which the structures
were conceived and constructed are understood.
New insights have been obtained recently on the linear measurement units of the
Mughal period [Balasubramaniam 2008a, 2009a, 2009b]. The measures used by
different Mughal monarchs, their relationship with each other and, more importantly,
their relationship with a constant angulam unit (of 1.763 cm) are now known
[Balasubramaniam 2008a]. This is summarized in table 1.
Mughal Unit Relationship with angulam Measure
Sikandari gaz 28/24 x 39 x A (=45.5 A) 80.217 cm
Illahi gaz 28/24 x 288/7 x A (= 48A) 84.624 cm
Padshahi zira 45 A 79.335 cm
2 1/2
Shahjahani dira [ 2/3 x ((28/24) x (288/7) x A) ] (=39.192 A) 69.096 cm
Table 1. Linear measures of the common units used in the Mughal period expressed in terms of
angulam and modern centimeters. Each angulam (A) is 1.763 cm.
Name of measure No of angulams Centimeters
angulam 1 1.763
vitasti 12 21.256
pada 14 24.682
aratni 24 42.312
P-hasta 24 42.312
C-hasta 28 49.364
F-hasta 54 95.202
kishku 42 74.046
kamsa 32 56.416
danda 96 169.248
dhanus 108 190.404
Table 2. Units of measure mentioned in the Arthasastra in terms of number of angulams and
centimeters, using the conversion 1 angulam = 1.763 cm. This table is fundamental to the
understanding of metrology of the Indian subcontinent through the ages. The different types of
hastas are defined in [Balasubramaniam 2009a].
The angulam is the traditional unit of measure of the subcontinent. Recent studies
have shown the constancy of the angulam at 1.763 cm through the ages
[Balasubramaniam 2008a, 2008b, 2009a, 2009b, 2009c]. Interestingly, this value for the
angulam was obtained without any a priori assumptions from the designs of Harappan

274 R. Balasubramaniam – On the Modular Design of Mughal Riverfront Funerary Gardens

civilization sites [Danino 2005, 2009]. This is also directly confirmed by the appearance
of a similar unit in the Lothal ivory [Rao 1955-62: II, 689-690] and Kalibangan
terracotta [Balasubramaniam and Joshi 2008] scales of the Harappan civilization. In
particular, the units of measurement described in Kautilya’s Arthasastra [Shamasastry
1939; Kangle 1986: II, 138-140], dated to around 300 B.C., were used to understand
the engineering plans of most engineered structures of the Indian subcontinent through
the ages [Balasubramaniam 2008a, 2008b, 2009a, 2009b, 2009c], until the adoption of
British units in early twentieth century. The measurements defined by the Arthasastra in
terms of modern centimeters are listed in table 2. The significant measurements that are
emphasized in this present paper are vitasti (= 12 angulam), dhanus (=108 angulam) and
rajju (=10 dhanus).
Modular design of overall complexes
Before focusing on individual structures, we will first address the modular design of
the overall layout of the two complexes. The symmetric modular design is evident when
the known measurements of the Humayun’s tomb and Taj Mahal complexes are
converted to the traditional units mentioned in Kautilya’s Arthasastra. The plans of these
two complexes in terms of traditional units of measurement are presented in figs. 1 and
2. Notably, the measures of major dimensions appear as logical numbers when expressed
in terms of traditional units, in particular the vitasti (V). A brief discussion on the
modular plans of these two complexes in terms of traditional units follows.
Humayun’s tomb complex
The boundary wall of the Humayun’s tomb complex is missing on the eastern side.
The Yamuna flowed past this side in earlier times. The modular planning of the
Humayun’s tomb complex is clearly evident in the ordering of the small garden plots
around the central tomb (see fig. 1). Based on the dimensions of the complex reported by
Zafar Hasan [1997], the entire complex can be analyzed in terms of the units mentioned
in Arthasastra.
It can be seen that each side of the complex measures 1620V in length. Specifically,
the modular design can be understood in terms of grids of size 240V x 240V (see fig. 1).
Each square garden plot measures to this grid. The tomb structure has been designed so
that it is located in the central square grid of 540V x 540V (see fig. 1). The walkways are
another characteristic of the cross-axial garden. The central walkway is 60V in width,
while the smaller ones are 24V wide. The width of the lane bordering the wall is 12V.
In order to highlight the modular planning of the garden of Humayun’s tomb, one of
the quadrants of the complex has been analyzed (fig. 3). Each side of the garden plots is
240V in length. Therefore, considering one quadrant of the overall plan, the division of
the length of side would be 12V+240V+24V+240V+24V+240V+30V = 810V. The last
factor 30V is half of the central walkway of width 60V. One can also view the
dimensions of Humayun’s complex in terms of the larger Arthasastra unit of dhanus. The
overall length of each side of Humayun’s tomb complex in terms of dhanus units is
180D. The central square, where the platform is located, occupies an area 60D x 60D.
The dimensions proposed in figs. 1 and 3 and the actual measures [Hasan 1997: 117-
124] are quite close, as shown by error analysis, the results of which are presented in
Table 3. The error is defined as the deviation of the proposed measure from the actual
measure, expressed in percentage of the proposed measure. The low errors confirm that
the modular planning of Humayun’s tomb complex was based on Arthasastra units.

Nexus Network Journal – Vol.12, No. 2, 2010 275

 Fig. 3. Modular layout of one quadrant of Humayun’s tomb complex. The vitasti (V) equals 12
angulams and each angulam measures 1.763 cm.
No Location Proposed measure Actual %
measure (cm) error
vitasti cm
(V)
1. N-S length of complex 1620V 34853.88 -1.70
34272.72
2. E-W length of complex 1620V 34152.84 0.35
34272.72
3. Width of main causeway 60V 1264.92 0.35
1269.36
4. Width of minor causeway 24V 518.16 -2.05
507.744
5. E-W and N-S lengths of
central platform 540V 11424.24 11277.60 1.28
6. E-W and N-S lengths of
second platform 450V 9520.20 9326.88 2.03
7. E-W and N-S lengths of
mausoleum plinth 240V 5077.44 5029.20 0.95
Table 3. Comparison of proposed and actual dimensions [Hasan 1997] of some important sections
of Humayun’s tomb complex. The vitasti (V) equals 12 angulams and each angulam measures
1.763 cm.

276 R. Balasubramaniam – On the Modular Design of Mughal Riverfront Funerary Gardens

Taj Mahal Complex
The overall plan of the Taj Mahal complex is shown in fig. 2. The complex can be
considered to consist of four sections. The riverfront terrace (T) is the northernmost
section. The garden (C) is located beside the terrace. The forecourt (J) and caravanserai
(S) on the south complete the complex.
The dimensions of significant locations and features of the Taj complex are given in
[Barraud 2006a; 2006b]. Therefore, the dimensions of the complex can be determined in
Arthasastra units. The total length of the complex, including the forecourt and the
caravanserai is 4260 V. Each of the four sections of the Taj Mahal complex have been
designed based on grids. The modular planning of the four sections of the Taj Mahal
complex has been described in great detail elsewhere [Balasubramaniam 2009b]. The
terrace and charbag sections have been designed and planned to a larger grid size than
that used for the forecourt and caravenserai sections.
The modular planning of the terrace and garden section of the Taj complex are
highlighted in the schematic of fig. 4. The unit mentioned in this figure is the dhanus
(D), each being equal to 108 angulam. This can be converted to vitasti (V), which is
equal to 12 angulam. Therefore, 1D = 9V.
The total length of the riverfront terrace and the garden is 220D. The garden
occupies an area of square 160D x 160D. The length of the riverfront terrace is 60D.
Further, the garden has been planned by dividing the total garden area into smaller grids
of size 10D x 10D, as shown in fig. 4.
It is interesting to note that 10D is equal to 1 rajju, a measure mentioned in the
Arthasastra for measuring larger distances [Shamasastry 1939; Kangle 1986: II, 138-140].
(Incidentally, rajju in Sanskrit means “rope.”) Therefore the modular planning of the
garden and riverfront terrace sections of the Taj complex can be understood using a
modular grid pattern of 1 rajju x 1 rajju ( = 10D x 10D = 90V x 90V).
The errors between the proposed and actual dimensions [Barraud 2006a; 2006b] for
the Taj complex are low (see table 4). This confirms the use of Arthasastra units for the
modular planning of the Taj complex.
No Location Proposed measure Actual Percentage
vitasti cm measure error
(V) (cm)
1. E-W length of terrace 1440V 30464.64 30084 +1.25
2. N-S length of terrace 540V 11424.24 11189 +2.06
3. E-W and N-S lengths of marble 450V 9520.20 9569 -0.51
platform
4. E-W and N-S lengths of 270V 5712.12 5690 +0.39
mausoleum plinth
5. E-W and N-S length of charbag 1440V 30464.64 29631 +2.74
6. E-W length of jilaukhana 1440V 30464.64 30084 -1.25
7. E-W length of caravanserai 1440V 30464.64 30084 -1.25
8. N-S length of caravanserai 1560V 33003.36 33490 -1.47
Table 4. Comparison of proposed and actual dimensions [Barraud 2006a; 2006b] of some
important sections of Taj Mahal complex. The vitasti (V) equals 12 angulams and each angulam
measures 1.763 cm.

Nexus Network Journal – Vol.12, No. 2, 2010 277

 Fig. 4. Modular layout of riverfront terrace and charbag of Taj Mahal. The vitasti (V) equals 12
angulams and each angulam measures 1.763 cm.
The above analysis has highlighted the important fact that the overall design of the
Humayun’s tomb and Taj Mahal complexes can be rationalized in terms of the
traditional length units of the subcontinent, especially the ones defined in Arthasastra
with the angulam taken as 1.763 cm. This is an important original finding of this study.
This concept will be now applied in order to understand the modular design of the
platforms on which the two mausoleums are erected.

278 R. Balasubramaniam – On the Modular Design of Mughal Riverfront Funerary Gardens

Modular planning of the platforms
The positioning of the two mausoleums with respect to the platforms on which they
stand is of interest.
In the case of the Humayun’s tomb, the central feature of the garden is a platform
that is 540V square with corners slightly cut off. It is paved with large blocks of Delhi
quartzite and is about 4 feet tall from ground level. In the centre of this platform is
another platform, 450V square and chamfered at the corners. This stands about 22 feet
(approximately 30V) above the level of the lower platform. The relative positioning of
these two platforms and the mausoleum of Humayun is shown in fig. 5. Each face of the
second platform contains 17 arched recesses, 8 on each side of the central steps and one
recess in the splayed corners of the platform. The second platform serves as the base for
the mausoleum. Steps in the centre of each of the four sides lead to the actual tomb on
the second platform. The mausoleum stands on a low plinth in the center of the upper
platform and is entered from the south. Each side of the square of mausoleum plinth is
240V. All these dimensions have been marked in fig. 5. The match between proposed
and actual measurements [Hasan 1997: 117-124] for the length of the two platforms and
the mausoleum plinth is excellent (see table 3).

Fig. 5. Proposed modular plan and predicted dimensions of the central platform on which stands
the Humayun’s tomb. The vitasti (V) equals 12 angulams and each angulam measures 1.763 cm.
In a similar manner, the positioning of the Taj mausoleum in the riverfront terrace is
as shown in fig. 6. The relative dimensions of the platform and the plinth of the Taj, and
their relation to the N-S length of the riverfront terrace (540V) are noteworthy. The
length and breadth of the marble platform is 450V, while the length and breadth of the
plinth of the mausoleum equals 270V (see fig. 6). The match between predicted and

Nexus Network Journal – Vol.12, No. 2, 2010 279

actual measurements [Barraud 2006a; 2006b] for the marble platform (450V) and the
mausoleum plinth (270V) is excellent (see table 4). In this manner, the square
(representing the plinth of the mausoleum) of 270V per side was planned in the centre of
another square (representing the marble platform) of 450V per side. Incidentally, these
dimensions are symmetrically related to the N-S length of the terrace, which equals
540V. The overall symmetry of this design scheme can be appreciated in the plan shown
in fig. 6.

Fig. 6. Proposed modular plan and predicted dimensions of the riverfront terrace containing the
marble platform on which rests the plinth of Taj Mahal. The vitasti (V) equals 12 angulams and
each angulam measures 1.763 cm.
The low percentage of errors between the predicted and actual measures of platforms
of these complexes (tables 3 and 4) offers firm proof of the fact that the designs of these
complexes were based on Arthasastra units. It must be emphasized that logical numbers
do not result when the designs are expressed in Mughal length units listed in table 1. The
novel proposal of measurement units of Mughal structures, presented in this paper for
the first time, will now be used to analyze the design of the two tomb structures in these
two complexes
Modular design of mausoleums
The favorite Mughal form for mausoleums is a centrally planned building known in
Persian as hasht bihisht or eight paradises [Koch 2002: 45-50]. The hasht bihisht design
appears dramatically in the Humayun’s tomb (see fig. 7) and Taj Mahal mausoleum (see
fig. 8). In both the tombs, four radially planned hasth bihist elements can be noted. Both
the cross-axial pattern as well as a diagonal axis pattern is evident. The modular designs
can be better appreciated by discussing the metrology of these structures individually.

280 R. Balasubramaniam – On the Modular Design of Mughal Riverfront Funerary Gardens

 Fig. 7. Proposed modular plan and predicted dimensions of Humayun’s mausoleum. The vitasti
(V) equals 12 angulams and each angulam measures 1.763 cm. Notice the different manner in
which the length of each side of the structures has been divided.
Humayun’s Mausoleum
Analysis of the dimensions of the tomb revealed its association with the traditional
units of Arthasastra. Each side of Humayun’s tomb is 240V. One can then divide each
side in thirds, to give nine grids with each grid of size 80V x 80V. The grid based on this
division is marked with dark lines in fig. 7. However, in this case, this may not be an
appropriate grid size to appreciate its design because the finer dimensions of the structure
cannot be reconciled by intricate subdivisions of the 80V unit. The dimensions of
significant features like the width of doorways, etc. are not based on smaller subunits of
80V.
Careful examination of relative dimensions of various features of Humayun’s tomb
reveals that a grid pattern of 15V to the side better reflects the symmetry of the structure.
Most of the finer modular sub-divisions can be understood by considering 15V divisions.
For example, considering one side of the tomb, the length of the corner chamfer is 15V.
The distance from the chamfer to the small doorway is 15V, with the doorway occupying
a width of 30V. The width of the wall adjoining this small side doorway is 15V.
Immediately after is the central arched doorway whose entire width is 90V. The main
doorway entrance, located inside the doorway, is 60V. Similarly, the width of the entire
side doorway section containing the side doorway is 60V.
Several other division schemes can be used to highlight the relative positioning of
major structural features. Each side can first be simply divided into two halves (of 120V)
with each half further being divided as 75V from the centre line of the central doorway
to the centre line of the side door. Further, the distance from this location to the extreme

Nexus Network Journal – Vol.12, No. 2, 2010 281

corner is 45V, thereby giving the total half length as 120V. This scheme is indicated in
the top of fig. 7.
Another division scheme is at work when one considers the octagonal chambers in
the four corners of the mausoleum. Considering the centre of these chambers, another
grid can be generated dividing each side as 45V+150V+45V. This grid will have its
intersection points at the centre of the octagonal chambers on the four corners.
Another way to divide the side will be in the proportion of
15V+60V+90V+60V+15V. The end 15V units are for the chamfered end. The 60V
length is covered by the side faceway that contains the smaller doorway. The central 90V
occupies the central portion of the tomb.
In another scheme, one can visualize a grid pattern such that the central grid
coincides with the vertical and horizontal sides of the central octagonal chamber of the
tomb. In that case, the width of the central square will be 60V and therefore the side of
the grids adjoining this central grid will be 90V. The scheme of 90V+60V+90V division
of the side is indicated in the right hand side of fig. 7. The non-uniform grid drawn with
these dimensions will mark the central octagonal chamber and moreover, each corner
chamber will be now aligned with a perfect square.
Taj Mahal mausoleum

Fig. 8. Proposed modular plan and predicted dimensions of Taj Mahal mausoleum. The vitasti (V)
equals 12 angulams and each angulam measures 1.763 cm. Notice the different manner in which
the length of each side of the structures has been divided.

282 R. Balasubramaniam – On the Modular Design of Mughal Riverfront Funerary Gardens

 The proposed modular plan and dimensions of different sections of the mausoleum
are seen in fig. 8. In this figure, the vitasti (V) has been used. First, it is confirmed that
the Taj Mahal was designed using a master square of 270V to the side (see fig. 8). The
match of the predicted values of the dimension of the each side of the plinth with actual
measured values [Barraud 2006a; 2006b] is excellent (table 4). The usefulness of the
logical number 270 (= 33x10) for accomplishing further subdivisions is highlighted in
[Balasubramaniam 2009b]. The multiplication factor 10 involved in this number 270
facilitated decimal division of dimensions, which was important considering the intricate
inlay and exquisite mosaic work on the walls and floor of the Taj Mahal. It is important
to emphasize that the possibilities for the important triadic division of space are evident
only when the dimensions of the Taj Mahal are considered in units of vitasti.
It is interesting to explore how the plan of the Taj was divided. In the first instance,
the plan can be divided into nine smaller squares of side 90V (see fig. 8). The tripartite
composition and triadic symmetry of the Taj structure can be readily appreciated by
considering this grid division. The logical number of 270V for the length of the side was
divided into squares of 90V x 90V of the hasth bisith design. Further subdivision of the
90V length into thirds is evident in the width of the large arched doors (60V) and the
small arched doors (30V) on each (outer) face of the mausoleum (see fig. 8). The
predicted value for the width of the large door is 1269.36 cm. (=60V) and this is only
1.70% away from the average measured value of 1291 cm. [Barraud 2006a; 2006b].
The concept of triadic division can be extended further by considering division of
90V into three 30V segments. Further, these 30V units can be subdivided into three
divisions of 10V each. It is reasonable to propose that this triadic division of space was
very critical to and aided in the tripartite composition of the Taj Mahal.
The lengths of several major sections and architectural elements of the Taj Mahal
reveal that the side of length 270V was also divided into other different schemes, like
60V+150V+60V or 45V+180V+45V. These possibilities have been marked in fig. 8. The
division of the 270V side into unequal lengths of 60V+150V+60V results in the grid
pattern whose intersection points precisely match the centers of the octagonal chambers
on the four corners. This grid pattern has been shown by a dotted line in fig. 8. Using
the scheme of dividing the 270V length into lengths of 45V+180V+45V, it is possible to
recognize the design of the chamfered corners of the Taj Mahal. Considering each corner
square of dimensions 45V x 45V, it is first noted that the length of the corner section
does not correspond to the length of the diagonal of this square (namely (45/—2)V =
63.63V). The corner has been designed such that its length is 30V and the length of the
corner door archway is 20V (see fig. 8).
The important finding of the present study is that the modular design and inherent
symmetry of Humayun’s tomb and the Taj Mahal be appreciated only when the
dimensions are expressed in Arthasastra units. This implies the utilization of traditional
building principles of the subcontinent in the construction of these two Mughal
masterpieces. The continuity of traditional engineering technology of the subcontinent is
therefore highlighted by the analysis presented in this study. It is confirmed that there is
no drastic change in the units of measurement used in architectural constructions during
the Mughal period. The native artisans and designers presumably followed the standard
that was well known to them over a long period of time [Balasubramaniam 2008a,
2008b, 2009a, 2009b, 2009c].

Nexus Network Journal – Vol.12, No. 2, 2010 283

Conclusions
The modular design and inherent symmetry of two significant Mughal funerary
garden constructions, Humayun’s tomb and the Taj Mahal, have been revealed by
analyzing their dimensions based on traditional Indian units of measure mentioned in
the Arthasastra, in particular the dhanus (D) measuring 108 angulams and vitasti (V)
measuring 12 angulams, with each angulam taken as 1.763 cm. The low percentage of
errors between the predicted and actual dimensions of different sections of these
monuments confirmed the novel approach to metrological analysis of Mughal structures.
Acknowledgment
The author thanks the Archaeological Survey of India for support of his studies on
Humayun’s Tomb in New Delhi and Taj Mahal complex in Agra.
References
ASHER, C.B. 1992. Architecture of Mughal India. New Cambridge History of India. Cambridge
University Press.
BALASUBRAMANIAM, R. 2008a. New Insights on Metrology during the Mughal Period. Indian
Journal of History of Science 43: 569-588.
———. 2008b. On the Mathematical Significance of the Dimensions of the Delhi Iron Pillar.
Current Science 95: 766-770.
———. 2009a. On the Confirmation of the Traditional Unit of Length Measure in the Estimates
of Circumference of the Earth. Current Science 96: 547-552.
———. 2009b. New Insights on the Modular Planning of the Taj. Current Science 97: 42-29.
———. 2009c. New Insights on Metrology during the Mauryan Period. Current Science 97: 680-
682.
BALASUBRAMANIAM, R. and J.P. JOSHI. 2008b. Analysis of Terracotta Scale of Harappan
Civilization from Kalibangan. Current Science 95: 588-589.
BANDARANAYAKE, S. 2000. Sigiriya: City Palace and Royal Gardens. Colombo: Central Cultural
Fund, Ministry of Cultural Affairs.
BARRAUD, R. A. 2006a. Modular Planning of the Taj. Pp. 108-113 in E. Koch, The Complete Taj
Mahal: And the Riverfront Gardens of Agra. London: Thames & Hudson.
BARRAUD, R. A. 2006b. Factfile. Pp. 258-259 in E. Koch, The Complete Taj Mahal: And the
Riverfront Gardens of Agra. London: Thames & Hudson.
DANINO, M. 2005. Dholavira’s Geometry: A Preliminary Study. Puratattv, 35: 76-84.
DANINO, M. 2009. New insights into Harappan town-planning, proportions, and units, with
special reference to Dholavia. Man and Environment 33: 66-79.
HASAN, Z. 1997. Delhi Zail. Zail. Pp. 117-124 in Vol. III of Monuments of Delhi: Lasting
Splendour of the Great Mughals and Others. New Delhi: Aryan Books International.
KANGLE, R. P. 1986. The Kautilya Arthasastra. New Delhi: Motilal Banarsidass.
KOCH, E. 1997. The Mughal Waterfront Garden. Pp. 140-160 in Gardens in Times of Great
Muslim Empires: Theory and Design, Muquarnas, Supplement, Vol. 7, A. Petruccioli, ed.
Leiden: E. J. Brill.
———. 2002. Mughal Architecture: An Outline of Its History and Development (1526-1858).
New Delhi: Oxford University Press, New Delhi.
———. 2006. Chapter II: The Construction of the Taj Mahal. Pp. 83-101 in The Complete Taj
Mahal: And the Riverfront Gardens of Agra. London: Thames & Hudson.
MISRA, N. and T. MISRA. 2003. The Garden Tomb of Humayun: An Abode of Paradise. New
Delhi: Aryan Books International.
NATH, R. 1982-2005. History of Mughal Architecture, 4 vols. New Delhi: Abhinav Publications.
RAO, S. R. 1955-1962. Lothal A Harappan Port Town. New Delhi: Manager of Publications,
Government of India Press.
SHAMASASTRY, R. 1939. Kautilya’s Arthasastra (trans.), 3rd ed. Mysore: Mysore Printing and
Publishing House.

284 R. Balasubramaniam – On the Modular Design of Mughal Riverfront Funerary Gardens

About the author
R. Balasubramaniam is a full professor in the Department of Materials and Metallurgical
Engineering at the Indian Institute of Technology, Kanpur where he teaches courses on materials
science and engineering, corrosion, surface coatings technology and history of metallurgy. He
earned a Ph.D. in Materials Engineering at the Rensselaer Polytechnic Institute of Troy, USA in
1990 and his research activity is based on corrosion, materials-hydrogen interactions and Indian
archaeometallurgy. In the recent past, he has actively researched the metrological tradition of the
Indian subcontinent with emphasis on the unit of measurement used through the ages. The
widely-published scholar is the author of nine books and serves on the editorial board of several
international scientific journals.

Nexus Network Journal – Vol.12, No. 2, 2010 285

 Maryam Ashkan Research
Discontinuous Double-shell Domes
Department of Architecture,
Faculty of Built Environment,
through Islamic eras in the Middle East
University of Malaya
and Central Asia: History,
Lembah Pantai
50603 Kuala Lumpur, MALAYSIA
Morphology, Typologies, Geometry,
[email protected]

Yahaya Ahmad and Construction

Abstract. This paper presents a developed geometric approach
Department of Architecture, for deriving the typologies and geometries of discontinuous
Faculty of Built Environment, double-shell domes in Islamic architecture. Common
University of Malaya geometric attributes are created using a corpus of twenty one
Lembah Pantai domes that were built in the Middle East and Central Asia,
50603 Kuala Lumpur, MALAYSIA beginning from the early through to the late Islamic periods.
[email protected] An outline of the origin and development of the
Keywords: Geometric designs, discontinuous double-shell domes and their morphological
Islamic mathematics, Eastern features are addressed. Using the al-Kashi geometrical
dome history, Middle East and essences, a four-centered profile as an initial shape is
Central Asian domes, constructed based on new geometric parameters to deduce
discontinuous double-shell the geometric commonalities of the two aspects of formal
domes, dome typology language (typologies and geometries) of such domes.
Common geometric prototypes for typical profiles shared by
the study cases are generated and formulated according to a
proposed system. The theoretical frame work for the formal
language of discontinuous double-shell dome architecture is
structured to indicate a moderate development of this sort of
Islamic domes and highlight the specific geometric
relationship between the Islamic domical configurations and
practical mathematic rules for many decades. It can also be
established a basic approach for considering the geometric
compositional designs and the typological derivations of the
other eastern domes.

1 Introduction
It has generally been recognized that domes whether as single domical buildings or in
large complexes of buildings, have played significant role in Islamic architecture. They
are different considerably in sizes and types. The double-shell domes included the
majority of Islamic domes and had gradually developed from the early Islamic epochs
through to the late Islamic era. An Islamic or eastern double-shell dome, which is defined
as the dome whose two shells have noticeable distances, is the matter of this paper.
The developed four-centered profile with newly geometric constitution, including the
variances in angles, divisions, which have been relied on the al-Kashi geometrical
essences, presents a novel method for deriving the diverse typologies and geometries of
discontinuous double-shell domes. Using this method, geometric properties and
compositions of the eastern domes can widely be considered.
The discontinuous double-shell domes increasingly demonstrate the high level of
development of the Islamic dome architecture in the Middle East and central Asia. They
display a formal language encompassing geometric concepts, typologies variations, and

Nexus Network Journal 12 (2010) 287–319 Nexus Network Journal – Vol.12, No. 2, 2010 287
DOI 10.1007/s00004-010-0013-9; published online 5 February 2010
© 2010 Kim Williams Books, Turin
 morphological constitutions. These possess the specific geometric entities of the domical
formations that represent the physical link between practical mathematic and their
architectural expressions in order to emphasis the meaning of ‘centrality’ in the Islamic
architecture. The typological structure underlines the diversities of the discontinuous
double-shell domes that can be geometrically analyzed and syntactically systematized for
formulating a comprehensive compositional language. It also helps understand the styles
and aesthetic principles of these domes in Islamic architecture.
Despite several existing studies about the Islamic domes and their relative meanings,
the discontinuous double-shell domes still do not have completely known architectural
morphology, typology, geometrical context, and even associated terminologies.
The study of characteristics of the discontinuous double-shell domes and geometric
thoughts can be manipulated to give contemporary meanings to the traditional designs
and principles of the Islamic dome styles. Also, the developed geometric method has the
potential to be used analogically in analyzing and understanding the essences of different
sorts of the eastern domes.
This present paper contains the following parts: 1) a brief outline of the origin and
development of the discontinuous double-shell dome in historical architecture; 2) a brief
elaboration of the Islamic mathematician’s contributions and the al-Kashi geometry
essences; 3) an exposition of derived morphologies of the discontinuous double-shell
domes; 4) derivations of their typologies, geometric designs, and associated drawings of
such domes according to the developed geometrical method; and 5) a brief discussion on
their common construction methods and techniques.
2 Background
2.1 Historical outline of the origin and development of the discontinuous double-shell
domes in historical architecture
Generally in Islamic architecture, the regional terminologies used to name the distinct
building functions include qubbat (in Arabic), gunbad (in Iran and Afghanistan),
gumbaz (in Uzbekistan), and kumbett (in Turkey). They are normally referred to the
most distinguished forms of these domes in the Middle East and Central Asia.
After the introduction of the wooden domes in the near East, such as Sion church
(456-460 A.D.) in Jerusalem and the dome of Rock (621 A.D.) before and after the
coming of Islam [Smith 1971], thousands of masonry domes were built in this realm.
The absence of sufficient literature and meanings to assess the existing evidences had
regularly frustrated investigations about the Islamic domical styles and their architectural
configurations divorcing from their ancient and pre-Islamic origins. Nevertheless, the
Islamic domes certainly influenced on the Western domical architecture in the
nineteenth century [Grabar 1963; 2006].
Morphologically, in the construction of the eastern domes, the shell(s) can be put
together in three different ways (fig.1). These include one shell (the earliest type of the
eastern domes: (OS-Type 1 and OS-Type 2), two shells and three shells [Hejazi 1997].
However, few samples of these triple shells that emerged in comparison to large numbers
of the other sorts can thus verify its origin from the double-shell domes [Grangler 2004].
Regarding the double-shell types, two subdivision groups have been defined based on
how these two shells are composed together. They are the continuous and the
discontinuous groups.

288 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 In the continuous double-shell domes, sometimes, there exists no considerable
distance between the shells (CD-Type 1), or they are connected by brick connectors
(CD-Type 2), but very often the distance between these shells are small (CD-Type 3)
[Hejazi 1997]. It could thus be said that the continuous two shells domes are called
‘evolving’ from the one shell domes to the two shells domes in the Islamic dome
architecture development. The constructions of the one shell dome were continued up
to the late Islamic era [Stierlin 2002].
In the discontinuous double-shell domes, there are considerable distances between the
two shells. The discontinuity may start either from the base (DD-Type 3) or from the
top of the drum (DD-Type 1 and 2) [Hejazi 1997]. This is considered higher than the
other types of the Islamic domical typologies (DD-Type 2; TS-Type 1).

Fig. 1. Illustration of the Islamic dome typologies according to the composition of their shell(s),
after [Hejazi 1997]
One of the main advantages of discontinuous double-shell domes’ structure is the
separation the weathering surface from the internal shell and thereby, substantially giving
improved weather protection [Mainstone 2001]. Structurally, the weight of a given
overall breath of construction is reduced when using the light shells. In fact, its
construction method was also extremely successful, despite the seismic conditions in the
Middle East and central Asia [Hejazi 2003; Farshad 1977].Architecturally, it permitted
an increase in the external size and height of the dome thereby making it more imposing
without the necessary increase in its internal height which improved its aesthetical
meanings and splendors [Hillenbrand 1994; Michell 1978]. Mechanically, the internal
shell boost sound-reflected compared to the other types of the eastern domes. Any
whisper on one side of the domical chamber is easily heard because of its specific shape.
This principle is also applied to all forms of energy under the internal shell [Irfan 2002].
Overall speaking, the discontinuous double-shell domes as a vast group of Islamic
domes, in retrospect, resulted in the fairly continuous developments of generous

Nexus Network Journal – Vol.12, No. 2, 2010 289

proportions in dome practice in order to reach the meaning of ‘centrality’ in the Islamic
architecture [Michell 1978].
Historically, the most constant practices were not appeared up to the eleventh
century by the Seljuks architecture. Their architectural contributions and movements
involved innovative characteristics of various types of mausoleums [Creswell 1958]. They
showed enormous variety of appearances including octagonal, cylindrical (also called
tower, see fig.2a) and square shapes (fig. 2b, 2c) topped with two types of discontinuous
two shells: ‘conical’ roof formations (especially in Anatolia and Persia) or pointed’ shapes
(in Persia)[Saud 2003]. Nevertheless, the main origin of these compositions is the earlier
tower tombs of that area, especially, the prominent Gunbad-e Qabus at Gurgan (Iran,
1007 A.D.) [Saud 2003].

Fig. 2. Various types of Seljuks mausoleums: a) Döner Kümbet, Kayseri, Tuerkey [Hoag 2004];
b) Mausoleum of Fakhreddin Razi, Kunya Urgench, Turkmanistan [Source:
www.advantour.com/turkmenistan/dashoguz/il-arslan.htm]; c) Gunbad-i Surkh (Red tomb),
Maragha, Iran [Source: www. archnet. org, photographers: Sheila Blair and Jonathan Bloom 1984]

Fig. 3. The discontinuous double-shell domes placed over tomb towers, Kharraqan, Iran [Photo:
Seherr-Thoss 1968; Crosse-section after the Iranian Ministry of the Cultural Heritage (ICHO)]
In the conical roof (shell) cases, stone slabs or brick layers rested on the lower roof
(shell) with some internal voids for reducing the weight and protecting the lower parts
[Mainstone 2001]. In some cases, the tomb towers are topped with pointed shells. The
earliest known of such discontinuous double-shell domes were a couple of eleventh
century Iranian tomb towers (fig. 3). Both the internal and external masonry shells have

290 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
similar thickness and profiles as well as were composed without internal connections and
interconnecting wooden struts [Mainstone 2001]. Nevertheless, these tomb towers
demonstrated primary efforts in designing a conflict between the external appearance of
the domes and its aesthetic interior space in this period.
Another typical building typology, which showed considerable development under
the Seljuk patronage and were exceedingly used by Mongols later, is the domed
mausoleums were regionally built based on the developed architectonic configurations in
Greater Khorasan (now divided between Iran, Afghanistan, Tajikistan, Uzbekistan, and
Turkmenistan countries; see fig 4). Their structures consist of two cube-shaped stories
topped with huge dissimilar two shells domes [Pope 1971]. Study of the damaged
external shells of these pointed samples demonstrated the lacks of structural knowledge
and information of the proportional design.

Fig. 4. Two stories Seljuks mausoleums; a) Anonymous mausoleum, Khorasan, Iran [Memarian
1988]; b) Mausoleum of Sultan Sanjar (before conservation), Merv, Turkmenistan [Hoag 2004]
After the severe architecture degeneration, which was caused by the Mongol’s
invasions and their successor’s, Timur (that produced a gap in the dome construction
evolution), the material culture of the Middle East and central Asia flourished again
through Ilkhanids (in Iran) and Timurids (in Uzbekistan) [Stierlin 2002]. Architects and
artists from Asia Minor, Azerbaijan, the Caucasus, India, Iran, and elsewhere were
compelled to help the constructions of the often colossal state buildings for both sacred
and secular purposes [Michell 1978]. Subsequently, these were characterized by the
resemblance a huge of monumental discontinuous double-shell domes throughout this
realm. While the Ilkhanids domes were extensively built for the funerary usages, but the
Timurids domes were regularly attached to madrasa (religious school) and were often in
pairs instead of erecting on freestanding mausoleums and mosques [Hillenbrand 1994].
In this regard, the dome over the Sultan Bakht Aqa mausoleum in Isfahan (1351-52
A.D.) was the earliest known example of the discontinuous double-shell dome where the
external and internal shells substantially exhibited in different profiles with radial
stiffeners, as shown in fig. 5 [O’Kane 1998]. The dome of Sultaniya complex in Cairo
which were likely erected by Sultan Hasan about thirteen century, believed that is the
origin of formation of the Sultan Bakht Aqa tomb. Nevertheless, the use of internal
stiffeners between its two shells can be demonstrated its Persian origin [O’Kane 1998].
Consequently, the dome architectures were rapidly incorporated and altered into
local styles after the Timurids epoch by appearing three specific local dynasties
[Hillenbrand 1999] including: the Safavids in Iran (1501-1732 A.D.), the Shaybanids in
Central Asia (1503- ca. 1800 A.D.), and the Mongols in India (1525-1858 A.D., out of
scope of this paper).

Nexus Network Journal – Vol.12, No. 2, 2010 291

 Fig. 5. Sultan Bakht Aqa mausoleum with its internal stiffeners, Isfahan, Iran [Photo: Authors;
Cross-section after Memarian 1988]

Fig. 6. Uniform style of Timurid domes a) Bibi Khanum mosque, 1398/1405 A.D., Samarkand,
Uzbekistan, [Source: www.archnet. org; photographer: Hatice Yazar, 1990]; b) Khawaja Abu Nasr
Parsa mausoleum, 1460/1598 A.D., Balkh, Afghanistan [Source: www.archnet. org; photographer:
Stephen Shucart 2005]

292 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 They were dominated by the skilful use of diverse building materials (vernacular
architecture) and well-developed construction techniques which existed in each region
(fig.7) [Michell 1978]. At last, the application of innovative approaches in the Islamic
domical constructions became less important in the Middle East and Central Asia in the
middle of the late Islamic era. The constructions of Mongolian domes, however, survived
longer up to the end of late Islamic era in India [Stierlin 2002].
The most common prototype of discontinuous double-shell domes consists of the
external shell (the most importance component and the most visible part of dome), high
drum, internal shell, and radial stiffeners within the wooden struts. The latter was used to
fill the space between shells as well as integrating the whole components (fig. 8).

Fig. 7. The last generations of Islamic domes a) Tilla Kari madrasa, 1746/1660 A.D.(Shaybanids),
Samarkand, Uzbekistan [Source: www.archnet.org, photographer: Arul Rewal 1990]; b) Khawaja
Rabi mausoleum, 1617-22 A.D. (Safavids), Mashhad, Iran [Source: Authors]

Fig. 8. The distinct configurations of the discontinuous double-shell domes; a) Turabek Khanum
tomb, Seljuks, Uzbekistan. After [Hillenbrand 1994]; b) Gur-e Amir Timur, Timurids,
Uzbekistan. After [Hillenbrand 1994]; c) Shah mosque ( Masjid-e Imam), Safavids, Iran. After
[Stierlin 2002]

Nexus Network Journal – Vol.12, No. 2, 2010 293

2.2 Historical background of the mathematicians’ role in the design of Islamic domes
In an aforementioned development treatise of the Islamic discontinuous double-shell
domes, the role of the mathematicians can not be overlooked. Overall speaking, the
Islamic mathematic in contrast with the Greek mathematics can also be called “the
mathematics of practitioners due to close relationship of theory and practice [Özdural
2000]. Its proper demonstration can be derived from the works of al-Buzjani’s student
who recorded contexts of his meetings with master builders and architects to discuss
solutions to construction problems [Özdural 1995].
The locations of Islamic mathematic scientific centres were in the present-day Iran
and Iraq. The main used languages by mathematicians for writing mathematical treatises
were “Persian” and “Arabic” (principle language like Latin in Medieval Europe), and
“Turkish” (more translation versions). Because of this, it is often called “Arabic
mathematics”. It is interesting to observe how the mathematicians had to take into
consideration the master builders’ objectives for using geometry in design as well as how
the artisans had to realize the differences between precise and approximate approaches
[Katz 2007]. In fact, the main efforts aimed to define and formulate ‘an exact geometric
method’ rather than to determine certain proportions, not only for the dome designs, but
also for the arches and vaults compositions.
The primary textual signs of their assistance can be seen in the Ismail Samanid
mausoleum in Bukhara which was built in the early Islamic era (fig. 9) and its formation
was entirely designed by the geometrical principles of three celebrated mathematics: al-
Khorezmi, al-Fargani, and Ibn-Sino [Askarov 2007; Pope 1971].

Fig. 9. Geometric design of configuration of Ismail Samanid mausoleum, Bukhara, Uzbekistan

[Pope 1976]
A descriptive geometry method was used by Abu Sahl al-Quhi (circa. 1000) for
projecting circles on the sphere into the equatorial plane. Then, he rendered them back
onto the sphere in an outstanding visual manner. However, it would seem that he was
not interested in the practical mathematics [O’Connor and Roberstson 1999]. Another
known text is Kitab fi ma yahtaj ilayh al-kuttab wa'l-ummal min 'ilm al-hisab (a book on
the geometric constructions necessary for craftsmen), which was written by Abu’l- Wafa
Buzjani (circa. 1000) for practical use. Although neither the specific domical geometry

294 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
nor the related information was mentioned in this book, his particular geometric
methods have helped solve problems by simply using a ruler and a fixed compass [Jazbi
1997]. The long absence of developments during ca. 1000 until ca. 1400 was the result
of the genocide of scientists by Mongols and Timur troops [Stierlin 2002].
The Buzjani method was developed by the celebrated al-Kashi mathematician (1390-
1450 A.D.) to create the practical geometry for a variety of dome constitutions. Ghiyath
al-Din Jamshid Mas’ud Kashani (al-Kashi) ranks among the greatest mathematicians and
astronomers in the Islamic world. By far his extensive book is Key of Arithmetic (Book
IV), on “Measurements” where the last chapter, ‘measuring structures and building’, was
written for practical purposes by using geometry as tools for his calculations [Dold-
Samplonius, 1992, see fig. 10b].

Fig. 10. Traditional approaches for designing dome’s external shell: a) some geometrical shapes
from Suhayl al-Quhi’s book Fî istihraci mesaha al-muhassama al-maqafî or Risala-i abu Sahl
[Suleymaniya library, Ayasofya- 4832]; b) a page of al-Kashi book, IV Manuscripts [Memarian
1988]; c) Dold-Samplonius re-drawing of the al-Kashi method [Dold-Samplonius 2000]
Dold-Samplonius has discussed the several aspects of the al-Kashi calculation
principles. They include his good specific methods in approximating the surface areas
and the volume of the shell forming of the qubba (the dome). She elaborated his five
methods for drawing the “profile” of an arch from the al-Kashi’ s Key of Arithmetic (fig.
11) [Hogendijk and Sabra 2003; Dold-Samplonius 1992].
 The first and second approaches addressed the design of a three-centered profile
(centre points o, p, and q) according to the divisions of the circle into six parts
(approach 1) and eight parts (fig.11, approach 2);

Nexus Network Journal – Vol.12, No. 2, 2010 295

 Fig. 11. Five approaches for designing and drawing different types of pointed arches from Kâshânî’s Key of Arithmetic. After [Hogendijk and
Sabra 2003; Dold-Samplonius 1992; 2000]

296 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
Fig. 12. Illustrations of the profile generation and its geometrical properties [Source: Authors]
 Fig. 13. Illustrations of the four-centered initial profile and essences of its geometrical parameters [Source: Authors]

Nexus Network Journal – Vol.12, No. 2, 2010 297

  The third approach illustrated a four-centered arch which was drawn based on the
construction of a rectangle abpq under the span line and division of the span line
into eight parts (fig.11, approach 3);
 The fourth and fifth approaches depicted the geometries of two-centered arches
based on either the division of the span line or the construction of rectangle abpq
under the span line (fig.11, approaches 4 and 5).
On the geometrical point of view, the essences of al-Kashi drawings, which were
frequently employed for designing different types of arches, vaults, and domes [Dold-
Samplonius 2000], can be listed and concluded as follows:
 Using the variety of procedures of the division systems (divisions of circles in the
approaches 1 and 2, the specific constructions of rectangles under the span line in
the approaches 3 and 5, division of the span line into particular parts in the
approach 4) in order to get the centre points and for drawing the appropriate
shapes of arcs. The upper part of the profile (third and forth arcs) clearly showed a
longer radius compared to the lower part of the profile (first and second arcs).
Centre points of these arcs are located far from the profile than the lower part;
 Employing the different types of profiles (two-centered, three-centered and four-
centered) for designating the distinct types of the arch profiles for various building
usages.
3 The developed geometric method for analyzing geometries and typologies of
the discontinuous double-shell domes
A comprehensive geometric method with new parameters is required to be utilized as
a tool, not only for analyzing the configuration patterns of the various types of
discontinuous double-shell domes, but also for proposing a frame work in defining a
formal geometrical language for the different typologies of such domes according to their
initial profiles and geometric parameters. Subsequently, the external shell as the
dominant feature of such domes, which embraced the employed practice geometry and
presented typological features of such domes, is object of this stage of investigation.
Nevertheless, it is also necessary to generate the internal shell profile for deriving its
geometric variations and indications.
The key in understanding the geometric composition of the external shell is mainly
dealt with studying the dome cross-section when its thickness diminished [Huerta 2006].
This so-called ‘profile’ forms the basis of the dome geometric design. It consists of four
small arcs, namely, the lower part (first and second arcs) and the upper part (third and
fourth arcs) (fig. 12a, b). Its fundamental properties consist of the ‘span’ and the ‘rise’.
The horizontal distance between the two supporting members is called the span whilst
the rise is the vertical distance from the middle centre of the span line to the tip of the
profile (fig.12b). The span is the origin and fundamental to all the rules for obtaining the
proportion values.
The al-Kashi geometric approaches were used for designing common shapes of
profiles of arches, vaults, and domes at that time, and hardly covered the majority of
geometrical designs of the Islamic domes compositions with their different attributes at
various periods. Nevertheless, the al-Kashi four-centered profile has the potential to be
enhanced extensively through the expansion of the three geometrical parameters,
including the location of the first and second arcs, the location of the third and forth
arcs, and the positions of the breaking points.

298 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 On the other view, the frame work of the four-centered profile can be developed as ‘a
general initial shape’ of the dome context. It is defined to provide the flexibility in its
shape (or small arcs’ curvatures) through the application of the following essential
definitions (fig. 13):
1. The lower arcs: are the loci of the points whose centres are located on the two
vertices of rectangle iigg constructed above the span line. Values of its lengths
mi
and widths are gained based on the exact proportions of the span: s, (fig. 13a);
ni

2. The upper arcs: are the loci of the points whose centres are always set on the two
vertices of the rectangle ppqq constructed under the span line. Values of its
mic
lengths and widths are obtained based on the fractions of the span s (fig.13b);
nic
and
3. The Breaking points: are the couple of points a and b used for changing the
profile curvatures through two considered options; firstly, it can occur by crossing
the perpendicular lines aa and bb from the points a and b which are marked
mcc
from the end points of the span line based on the fraction of span s. Secondly,
ncc
the points a and b are gained from the certain values of the springing angles:
=25, 30 and 45.
In fact, the general initial profile (fig. 13c) contains the whole essential geometric
properties which are necessary in derivations of the typologies geometric prototypes and
their geometric designs. To facilitate presenting and setting of its geometrical indications,
a system is proposed {[R1],(B),[ R2]} = {Rectangular1, (Breaking point), Rectangular2}
whilst all values are calculated from the middle point of the span, O (0, 0). Where
ab=SSpan, and i=1, 2…5, then:
 R1: includes, respectively, values of a length and a width of the rectangle iigg
constructed above the span line. Two vertices of this rectangle are centre points of
the lower part arcs with the variable values as:
ª m1 º
«ig i' g ' ab
« n1 »
».
« m2 »
«ii ' gg ' ab »
«¬ n2 ¼

When ii=gg=0, then the centre points are located on the span line.
 B: shows two possible options of the breaking points either as the exact angular
values ‘O=25°, 30° and 45°, or the coordinates values of points which are
symmetrically positioned form the end points of span line as
(aa=bb=m3/n3ab,0);
 R2: describes, respectively, the values of a length and a width of rectangle pqpq
constructed under the span line. Two vertices of this rectangle are the centre
points of the upper arcs as follows:

Nexus Network Journal – Vol.12, No. 2, 2010 299

 ª m4 º
« p' q ' pq n ab »
« 4 ».
« m5 »
« qq ' pp ' ab »
¬ n5 ¼
Note that the string “var”, which means ‘variance’ in the drawings, is a specific
dimension or distance for having varied parameter on the specific direction. This
function helps in the flexibility of the analysis and in defining the rules for generating
common prototypes for the given typologies’ profiles.
The given lengths in this system should divide by two for obtaining the vertices i, g,
p, and q of the proposed rectangles. In fact, the obtained vertices are symmetrically
calculated and positioned on the both sides of the vertical axis.
4 The morphological features of the discontinuous double-shell domes in the
Middle East and Central Asia
4.1 Case studies
In order to define typologies of the discontinuous double-shell domes and to derive
the associated geometric profiles, a sample of twenty one cases were investigated in the
Middle East and Central Asia, including domes in Afghanistan, Iran, Azerbaijan, Turkey,
Kazakhstan, and Uzbekistan. The architecture chronology of the eastern dome with their
highlighted study periods is shown in fig. 14. The studied periods and the selected
dynasties are particular causes for developing the dome configuration design that it never
surpasses into the other periods later on. Fig. 15 exhibits twenty one assorted examples of
the discontinuous double-shell domes according to their studied dynasties.

Fig. 14. Eastern dome chronology and the specific periods of the discontinuous double-shell domes
4.2 Common morphological features of the discontinuous double-shell domes
In general, the discontinuous double-shell domes expose logical designs, with
dynamic articulations of distinct proportional components. They were extensively erected
in the mausoleum buildings rather than whether the mosques or madrasas which are
increasingly typical buildings of the late Islamic era.
As a result of the analytical analysis of the case studies, the common identical
components of this type of the Islamic domes are morphologically addressed and listed as
follows:
 External shell: this is what appeared from the outside of the dome buildings. It is
the only architectural item which was conceptually found in synonymous during
the several Islamic epochs. Its thickness is proportionally reduced from its base to
the tip at either 25° or 30° angles;

300 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 Fig. 15. Classification of selected samples based on the dynasties studied

 Internal shell: this covered the internal domical chamber and has a simple
geometric formation compared to the external shell. In fact, it is necessary that its
geometric shape fully conformed with the external shell for transferring forces
from the upper components to the transition tier in order to stabilize the dome
structure.
 Drum: Considerable thoughts and efforts were often given by the designers to
make the building as high as possible using the tall drum. In fact, the
discontinuous double-shell domes are considered as one of the highest samples of
the eastern domes, with an average height of 30-35 meters from the ground. Its
thickness must be sufficiently massive in order to transfer and neutralize the
vertical thrust of the external shell to the lower items, especially, the internal shell;
 Internal stiffeners with the wooden struts: these are architecturally specific
characteristics of such domes (except for the conical samples) called, ‘the
composition of the radial brick walls with the wooden struts’.

Nexus Network Journal – Vol.12, No. 2, 2010 301

302 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
Fig. 16. Illustrations of the various configurations of internal stiffeners and wooden struts placed between two shells; [Sources: b-1,
Authors; b-3, Memarian 1988; c-3, The Ministry of Culture & Youth Affair in Afghanistan]
 Generally speaking, they were built in the space between the two shells for the main
purpose of filling the empty spaces and to support the external shell. The radial walls are
divided into two types of primary and secondary (smaller than primary ones, see fig. 16,
a1-2). A different number of the radial walls observed were 4 (Sultan Bakht Aqa, first
sample), 6, 8, 14, to 18. Their heights also varied from 8 to 15 meters. Their settings and
sizes strongly affiliated with the size of the span. The walls thickness is also changed, but
not less than 40 cm has been seen (fig. 16, c-3).
It is necessary to say that the arrangement of the internal stiffeners and their numbers
have often been altered (fig. 16, c1-2)as a result of conservational interventions, such as
replacement of the wooden struts with steel meshes or even removal of the wooden
beams (fig. 16, b1, 2, and 3). Nevertheless, these removable wooden struts have basically
set into the devised holes located on the thickness of the brick radial walls (stiffeners).
They are arranged compositionally based on the vernacular architecture agreements.
For example, the vertical wooden posts were often seen in the Iranian samples (fig. 16,
d1, 2). However, the pairs of the small wooden pieces, which are connected the radial
walls to the drum’s body, are the predominant features in the Uzbekistan, Turkmenistan,
and Afghanistan (fig. 16, e1, 2).
5 The generative system for derivations of typologies and common geometries of
the discontinuous double-shell domes
5.1 Samples analysis approach
Islamic domes, however, present a wide variety of sizes and types, but some geometric
properties were repeatedly used in their composition designs. Nevertheless, no two
samples are exactly the same. Visually, the analysis of the external shells of samples
revealed three classifications of typologies including, conical, pointed, and bulbous which
embrace the different geometrical properties and architectonic characteristics.
In primary stage, the computation process of the samples started with generation of
both (external and internal) shells profiles as shown in stage 1 of fig. 17. In order to
categorize the derived profiles of the external shells into the identified typologies, in what
follows, the three shape-patterns of those typologies are mainly elaborated and
schematically designed:
 Conical pattern: it is a triangle which is circumscribed by a rectangle;
 Pointed pattern: it is the feature whose lower arcs (the first and second arcs ) are
tangent to the two vertical lines passing from the end points of the span line; and
 Bulbous pattern: it is the prototype where the vertical lines intersected the lower
arcs (the first and second arcs).
These patterns helped not only to distinguish between the pointed and bulbous
profiles but to understand also some geometric properties of the deduced profiles
constitutions which included location of the centre points of the lower arcs and
proportions of rise to span of the conical samples.
In the second stage, after categorizing profiles in the exact typologies, their geometric
parameters, proportions, angles, and breaking points locations were respectively deduced
to identify the typical attributes of each typology.

Nexus Network Journal – Vol.12, No. 2, 2010 303

 Fig. 17. A flow chart illustrating the different stages of analyzing and developing common
geometric prototypes for the identified typologies of the discontinuous double-shell domes

304 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
Fig. 18. Systematic geometric analysis of the conical typology samples. Cross-sections after:
[Hillenbrand 1994]; [Stierlin 2002]; [Pope 1971]; [Memarian 1988]; [The Iranian cultural
heritage organization documentation centre, (ICHO)]

Nexus Network Journal – Vol.12, No. 2, 2010 305

 In the third stage, the common geometric prototypes (see fig. 13c) of the identified
typologies are systematically generated from the initial profile, based upon the derived
geometric parameters from the samples of every typology. These common geometric
prototypes have also the potential to alter into the variations of subsets according to the
diversity of values that might be exited in parameters of each group of the identified
typologies examples.
Subsequently, the geometrical drawing steps of sample profiles were drawn, not only
to emphasize advantage of the proposed system, but also to verify the common geometric
prototypes. Using this method, any complex compositions of the dome conceptualisms
can be generated by employing additional values. They can be drawn based on the
elaborations of an analogous drawn profile of the discontinuous double-shell domes
(explained in following sections) in the modern language.
Variations in sizes and formations of Islamic domes expose a problem in generalizing
the typologies of the designs. Diversity in the shell formations due to conservational
interventions or other reasons are disregarded, even if, the developed initial profile
parameters are able to analyze their configurations. Slight inconsistencies in proportions
and angles are also ignored so as to assist more beneficial discussions on the classifications
of the common typological designs as well as to derive their common geometric designs.
Note that to prevent repetition of example’s names during the analyzing process, the
association numbers, which marked boldly under their pictures (see fig. 15), were used.
5.2 Conical Typology and Geometry
In spite of the complicated curvature characteristics of the pointed and bulbous
profiles, the external shell of the conical type exhibits a simple profile constitution. In the
primary stage, according to its shape- pattern definition, it is an isosceles triangle
circumscribed in a rectangle in such a way that one vertex is positioned on the middle of
its side. In the second stage, the edge ab in cases of the hexagonal prism forms (see fig.
18, cases 4 and 1) is divided into three equal parts (1/3 s).
In the third stage, common prototype of the conical typology (fig. 19) of the
discontinuous double-shell domes is logically generated by the ratio of the rise/span,
called n, where n d {1, 5/6, 7/8}:
 n=1: depicts the equal amounts of the span and rise;
 n< 1: means the smaller values of the rise in compared to the span.

Fig. 19. Common geometric prototype of the conical typology.

oH
ab
^
n d 1, 5 , 7 ,1
6 8
`

306 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 Depending on the design requirements and preferences, the forms of the internal
shell showed variations, even more than the pointed and bulbous examples such as, semi-
circular (case 6), saucer (cases 2, 4, and 7), and several shapes of the pointed form with
distinct geometrical properties including, the two-centered (cases 3, 5) and the four-
centered (case 1) forms. In the pointed cases, the geometric parameters of profiles are
systematically formulated based on the proposed system, which can facilitate their
drawing steps, as follows:
­°ª1 / 3ab º ªab º ½°
 Case 1: ®« », ‘45, « » ¾ ;
°̄¬0 ¼ ¬ab ¼ °¿

­° ª1 / 2ab º ½°
 Case 3: ®0,0, « »¾ ;
°̄ ¬1 / 8ab ¼ °¿

­° ª1 / 2ab º ½°
 Case 5: ®0,0, « »¾ ;
°̄ ¬1 / 2ab ¼ °¿

­°ª2 / 6ab º ½°
 Case 9: ®« »,0,0¾ .
°̄¬ 0 ¼ °¿

The drawings of the conical external shell profiles are limited to the designation of
triangles and rectangle, while the drawing steps of their internal shells with the pointed
profiles fully conformed to the pointed typology that would be comprehensively
elaborated in the next section.
No specific developments were discovered following the considerations of geometries
and the architectonic compositions of the Seljuks mausoleums, except for the domical
buildings which were partially reconstructed in the Timurids era.
5.3 Pointed Typology and Geometry
Historically, the pointed typology contained most number of the discontinuous
double-shell domes in the Middle East and central Asia. In the primary stage, using the
shape-pattern definition of this typology, if samples’ lower arcs (primary and secondary)
are tangent to the two vertical lines passing from the end of span, then the generated
profile is categorized in this group.
To embrace such a property, the centre points of the lower arcs (first and second arcs)
m1
have to be set on the span line, meaning that gg=ii= s=0. In addition, since the
n1
m3
proportions of the breaking points ( s) on the span line do not show clear values, the
n3
exact angles (such as, 25°, 30°, and 45°) have thus been proposed for determining the
breaking points.
In the second stage, as the results of the geometrical analysis of the examples, three
subsets of the pointed profiles have been recognized and configured according to the
number of their centre points (fig. 20):

Nexus Network Journal – Vol.12, No. 2, 2010 307

 Fig. 20. Systematic geometric analysis of the pointed typology samples. Cross-sections after:
[Hillenbrand 1994]; [Stierlin 2002]; [Pope 1971]; [Memarian 1988]; [The Iranian cultural
heritage organization documentation centre, (ICHO)]; [The Ministry of culture & Youth Affair in
Afghanistan]; [Grangler 2004]

308 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
  Two-centered profile: The case 11 was analyzed based on the al-Kashi method
and organized using the parameters of the new system:
­°ª2 / 8ab º ½°
®« »,0,0¾ .
°̄¬0 ¼ °¿
 Three-centered profile: It is only identified in the case 19 according to the
following geometrical parameters:
­°ª0º ª2 / 4abº ½°
®« » , ‘30, « »¾ .
°̄¬0¼ ¬1 / 3ab ¼ °¿
 Four-centered profiles: The majority of the cases 10, 13, 14, 15, and 20 are
categorized in this group based on the distinct geometrical indications:
­°ª4 / 8abº ª2 / 8ab º ½°
 Case 10: ®« » , ‘25, « »¾ ;
°̄¬0 ¼ ¬5 / 16ab ¼ °¿
­°ª2 / 8abº ª6 / 8ab º ½°
 Cases 13, 14, 20: ®« », ‘30, « »¾ ;
°̄¬0 ¼ ¬7 / 16ab ¼ °¿
­°ª2 / 6ab º ªab º ½°
 and Case 15: ®« », ‘45, « »¾ .
°̄¬0 ¼ ¬2 / 6ab ¼ °¿

In the third stage, the common geometric prototype of the pointed typology is
systematically generated based on the geometrical indications through this system
{[R1],‘B,[R2]}, where its common geometrical parameters are presented as follows (see
fig. 21):
­ ª m2 º½
°ª m1 u abº « n u ab » °
°« » , ‘B, « 2 » °¾ , ab=span, ‘B =25°, 30°, 45°, m is an integer digit and is
®« n1 » i
° «0 « m »
¬ »
¼ «
3
u ab » °
° ¬ n3 ¼ °¿
¯
greater than zero, ( mi d ni ); ni= 3, 4, 6, 8, 16, and i=1, 2, 3.

Fig. 21. Common geometric prototype of the Fig. 22. Classification of the pointed typology
pointed typology based on the variety of rises

Nexus Network Journal – Vol.12, No. 2, 2010 309

 Therefore, based on the foregoing epistemological syntax of geometric variables, any
kind of the pointed typologies of the Islamic discontinuous double-shell domes can be
organized using these geometric indications.
Furthermore, according to the variance in heights of the rise, the external shells of
this typology can also be categorized into shallow (see fig. 22, case 10), medium (cases
11, 13, 14, 19 and 20), and sharp (case 15). This property fully conformed to the scale of
the rectangle ppqqand the values of angle .
The internal shell forms are recognized as semi-circular (cases 10, 14, and 20), semi-
ellipse (cases 19y and 15), pointed (cases 11 and 13), and saucer forms (case 19z). On the
other perspective, the presented analysis approach with its systematical presentation tool
not only can utilize as the syntax elaboration of the geometric arrangements of the
external shell of the discontinuous double-shell domes, but it also has the potential to
compose any sort of the pointed features, such as the internal shell in given typologies as
follows (see fig. 20):
­°ª2 / 10ab º ª4 / 10abº ½°
 case 13: ®« », ‘30, « »¾ ;
°̄¬0 ¼ ¬4 / 10ab¼ °¿
­°ª2 / 10ab º ª8 / 10ab º ½°
 case 11: ®« », ‘30, « »¾ .
°̄¬0 ¼ ¬3 / 10ab ¼ °¿
To verify the common geometric prototype, the four-centered profile with its
geometric parameters for instance:
­°ª2 / 8ab º ª3 / 4ab º ½°
®« », ‘30, « »¾ ,
°̄¬0 ¼ ¬7 / 16ab ¼ °¿
which is considered as most common geometric attributes of the pointed samples, has
been selected. Note that the same method can also be adapted to draw any sort of
geometric prototype of the pointed profile. The construction of the sample is thus (fig.
23):
 Step 1: The baseline ab as span is constructed and divided into 8 equal parts. The
point O is set on the 1/2ab. The points i and g (named the centre points of the
first and second arcs) are marked symmetrically on the 1/8 ab from the point O.
 Step 2: By setting the compass on the points i and g, two circles can drawn.
 Step 3: The breaking points basically occurred at the angle of 30° from the point
O for gaining points a and b. The centre points of the upper arcs, p and q were
obtained by constructing the rectangle pqpq under the span line, where
pq=pq=6/8ab (marked systematically on the 3/8ab from the point o) and
pp=qq=7/16ab, as is shown in fig.23.
 Step 4: The compass is positioned on the point q and with the radius qa, the
final circle is drawn. The procedure is repeated with the compass placed at the
point p with the radius pb.
The exhaustive geometrical considerations and formulations of the samples of
pointed typology demonstrated the superior architectural development of this type of the
discontinuous double-shell domes. This might mainly occurred due to the relationship of
the mathematicians’ collaborations and the architect’s artisans in the medieval Islamic
era.

310 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
Fig. 23. Illustrations of geometrical drawing steps of the pointed profile typology

Nexus Network Journal – Vol.12, No. 2, 2010 311

5.4 Bulbous Typology and Geometry
The bulbous dome (also, named ‘onion’ in some literature; cf. [Grabar 1963; Michell
1978; Stierlin 2002]) is the most prominent shape of the discontinuous double-shell
domes and the last generation of the Islamic domes which were appeared, especially,
since the Safavids dynasty (after the sixteenth century) in the Middle East and central
Asia. Their external shell has a longer diameter than the drum that it rest on and its
height value (rise) usually surpass its width (span).
In the primary stage, according to its shape-pattern definition, the vertical lines
passing from the end points of the span should intersect the first and second arcs. To
embrace such a property, the centre points of the lower arcs have to be placed on the
rectangle iigg, constructed above the span line.
In the second stage, because no clear angle values have been recognized, the breaking
points were calculated depending on certain coordinates B=(m3/n3ab,0). These
coordinates’ values marked the locations of the perpendicular lines, aa and bb, which
are calculated from the end points of span.
In the analysis of the samples, the following subsets for the bulbous profiles were
deduced and formulated based on the geometric variable indications (fig. 24):
­°ª2 / 10ab º ª2 / 10ab º ½°
 Case 8: ®« », 1 / 10ab,0, « »¾ ;
°̄¬1 / 6ab ¼ ¬2 / 10ab ¼ °¿

­°ª2 / 16ab º ª8 / 16ab º ½°

 Case 12: ®« », 1 / 16ab,0, « »¾ ;
°̄¬1 / 16ab ¼ ¬2 / 16ab ¼ °¿

­°ª2 / 8ab º ª10 / 8abº ½°

 Cases 17 and 21: ®« », 1 / 8ab,0 , « »¾ ;
°̄¬1 / 8ab ¼ ¬5 / 8ab ¼ °¿

­°ª2 / 12abº ª14 / 12abº ½°

 Case 18: ®« », 1 / 12ab,0 , « »¾ ;
°̄¬1 / 10ab ¼ ¬4 / 12ab ¼ °¿

­°ª4 / 8ab º ª1 / 2abº ½°

 Case 16: ®« », 1 / 32ab,0, « »¾ .
°̄¬1 / 8ab ¼ ¬1 / 8ab ¼ °¿

In the third stage, the common geometric prototype of the bulbous profile is
consequently devised based on this system {[R1], (B), [R2]}, and its common geometric
parameters are shown as follows ( fig. 25):
­ª m1 º ª m4 º½
°« u ab » « n u ab» °
°« 1n
», m3 / n3 u ab,0 , « 4 » °¾ ,
®
° « m » « m »
2
u ab» «
5
u ab » °
°«¬ n2 ¼ n
¬ 5 ¼ °¿
¯

where mi is an integer digit (it is possible that m4 > n4) and is greater than zero,
ni=6,8,10,12,16,32 (only in n3), ab= span, and i=1, 2…5.

312 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 Fig. 24. Systematic geometric analysis of the bulbous typology samples. Cross-sections after:
[Stierlin 2002]; [Memarian 1988]; [The Iranian cultural heritage organization documentation
centre, (ICHO)]; [The Ministry of culture & Youth Affair in Afghanistan]

Nexus Network Journal – Vol.12, No. 2, 2010 313

 Fig. 25. Common geometric prototype of the bulbous typology
Using the premise parameters of the above formulation, any subset of the bulbous
conceptualism can be generated by applying different values.
Additionally, according to the variance in heights of the rise, the external profiles can
also be categorized into shallows (cases 8, 12, 16) and sharps (cases 17, 18, 21) (fig. 26).
They are fully affiliated to the scale of the two rectangles, pqpq and igig.

Fig. 26. Classification of the bulbous typology based on the variety of rises
Architecturally, the internal shell formations have been found that resemble the
pointed typology in most cases, such as the saucer (case 8), semi-elliptical (cases 12 and
18), and semi-circular (cases 16 and 21) forms. However, bulbous geometric attributes
(case 17) did not show the developing configuration compared to either the conical or
pointed typologies. The geometric parameters of this sample can be set out as follows:
­° ªab º ½°
®0,0, « »¾ .
°̄ ¬5 / 8 ab ¼ °¿

The bulbous profile of the case 12 with the geometric indications

314 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 ­°ª2 / 16ab º ª8 / 16ab º ½°
®« » , 1 / 16ab,0 , « »¾
°̄¬1 / 16ab ¼ ¬2 / 16ab ¼ °¿

was selected for the verification of common geometric prototype of bulbous. It is

interesting to say that the same approach can be used for drawing any kind of the
bulbous profiles. The construction of the bulbous profile is thus (fig. 27):

Fig. 27. Illustrations of geometrical drawing steps of the bulbous profile typology

Nexus Network Journal – Vol.12, No. 2, 2010 315

  Step 1: The base line of the span, namely, ab is constructed and then divided into
16 equal parts. The point O is set on the 1/2ab.
 Step 2: The rectangle igig is then marked out symmetrically on the 2/16 ab
(length) and 1/16 ab (width) from the base point O, constructed above the span
line. By setting the compass on the points i and g, two circles are respectively
drawn.
 Step 3: To get the breaking points a and b, two perpendicular lines are drawn
from points a and b which are located on the 1/16ab from the end points of
the span.
 Step 4: The rectangle ppqq is then constructed under the span line in order to
get the centre points of the upper arcs, p and q. By setting compass on the point q
and with the radius qa’, the third arc is drawn. This procedure is repeated with
the compass placed at the point p and the radius pb for drawing the fourth arc as
shown in step 4 of fig. 27.
6 Construction Method
Most domes were constructed from less flexible materials, such as stone, mud brick
and baked brick. Some times, brittle ceramics were utilized, not only to embellish the
exterior faces of the external shell and the drum, but to prevent the penetrations of rain
and snow.

Fig. 28. a, b )Illustrations of the brick rows arrangements in the constructions of both shells of the
discontinuous double-shell domes; c) Sayyidun tomb, Ilkhanids, Abarquh, Iran ; d) Domes of
Darb-i Imam shrine, Safavids, Isfahan, Iran. [Photos: www.archnet.org, photographers: Sheila Blair
and Jonathan Bloom 1984]

316 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
 The common method for erecting the discontinuous double-shell dome involves the
building of half of the internal shell within the main roof. The drum and the stiffeners
were then built together on the lower components. The last task is to close the external
shell and to set the wooden struts in the devised holes between the radial walls (fig. 28d).
Nevertheless, it was impossible to continue using the same construction method near the
apse of both the shells where the empty oculi were remained. Therefore, the rows of
bricks were arranged vertically (fig. 28c) at these parts. The oculus of the internal shell is
the last task of construction. The construction techniques for arranging the brick layers
varied depending on the eastern domical types.
In bulbous and pointed typologies, the directions of the bricks’ rows are almost
always perpendicular to the generated curve of the dome surface (fig. 28b) in both the
external and internal shells. The thicknesses of both shells were also gradually reduced at
whether 22.5 or 25 angles for decreasing weights of the shells (except for the external
shell of conical samples). In the conical external shells, bricks rows are arranged with their
horizontal directions in such a way that the bricks in the upper row are precisely laid on
the half part of those in the lower row (fig. 28a). Similarly, the construction techniques
of internal shell of the conical typology fully conformed to that of curvature’s typologies.
Conclusion
The discontinuous double-shell dome is defined as the dome whose shells have
considerable distances. The discontinuous double-shell domes presented the high level of
design and configuration in the Islamic dome architecture in the Middle East and
Central Asia. They also included the majority of typologies of the Islamic domes from
the early Islamic era through to the late Islamic epoch.
The present research involved a new frame work that, not only identified
systematically morphological features and typologies variations of the discontinuous
double-shell domes, but also offered the analytical understanding of their geometrical
prototypes and related parameters which can be appeared in the designs of the Islamic
dome architecture during various eras.
Based on the essences of the al-Kashi geometric approach, a theoretical frame work
was geometrically developed in order to exhibit the geometric principles of compositions
of the discontinuous double-shell domes. By using the developed initial profile with its
defined parameters, the formal language for the discontinuous double-shell domes has
systematically been specified. It included: redefinitions of identified typologies,
classifications of the typologies subsets according to their heights variations, derivations
of internal shell forms and their geometric indications. Twenty-one samples of the
discontinuous double-shell domes, which were built in the Middle East and central Asian
countries, were subjects of this analysis.
Four main components morphologically recognized such as, the external shell, the
internal shell, the drum, and the internal stiffeners. Three main typologies of such domes
have geometrically derived and redefined including, conical, pointed, and bulbous. Three
subsets for the pointed and bulbous typologies have commonly derived based on their
external shells heights such as, shallow, medium, and sharp.
The biggest challenge was the creation of flexibility in configuring the initial profile,
especially, in locations of its breaking points that allowed the possibilities of covering
various dome designs. The results geometrically included two optional characters for the
breaking points.

Nexus Network Journal – Vol.12, No. 2, 2010 317

 The generated initial profile with its developed geometrical parameters, and the
proposed system, has the potential of offering a unique computational method for the
design analysis and geometric derivations of the Islamic domes, such as the Mongolians,
one shell domes, etc. Simultaneously, the results of this systematic geometrical analysis
could be recognized by using whether shape grammar or genetic algorithm, not only to
generate any kind of the Islamic domical configuration, but also to develop a more
advanced archive and retrieval system for similar data.
References
ASKAROV, S. 2003. Splendour of Bukhara. Archive of San’at magazine 4.
(http://www.sanat.orexca.com/eng/4-03/history_art2.shtml, accessed 19 Sep 2008)
CRESWELL, K. A. C. 1958. A short account of the Early Muslim architecture II. Oxford:
Clarendon Press.
DOLD-SAMPLONIUS, Y. 1992. The 15th century Timurid mathematician Ghiyath al-Din Jamshid
al-Kashi and his computation of the Qubba. Amphora : Festschrift for Hans Wussing on the
occasion of his 65th birthday. S. S. Demidov et al. (eds.), Basel, Boston, Berlin, p. 171-181.
———. 2000. Calculation of Arches and Domes in 15th Century Samarkand. Nexus Network
Journal 2, 3: 45-55.
FARSHAD, M. 1977. On the shape of Momentless Tensionless Masonry Domes. Journal of
Building and Environment (12): 81-85.
GRABAR, O. 1963. The Islamic dome: Some considerations. The journal of the society of
Architectural Historians 22, 4: 191-198.
———. 2006. Islamic Art and Beyond. United Kingdom:Variorum Press.
GRANGLER, A. 2004. Bukhara: The Eastern dome of Islam. Stuttgart: Axel Menges Press.
HEJAZI, M. M. 1997. Historical Buildings of Iran: their Architecture and Structure. Southampton:
Computational Mechanics Publications.
———. 2003. Seismic Vulnerability of Iranian historical Domes. Journal of Earthquake Resistant
Engineering Structures 4: 157-165.
HUERTA, S. 2006. Galileo was Wrong: The geometrical Design of Masonry Arches. Nexus
Network Journal 8 (2). 25-52
HILLENBRAND, R. 1994. Islamic architecture: form, function, and meaning. New York: Columbia
University Press.
———. 1999. The Ilkhanids and Timurids. In Islamic Art and Architecture. London: Thames
and Hudson. (p. 196-202)
HOAG, J. D. 2004. History of World Architecture: Islamic Architecture. Milan: Phaidon Press.
HOGENDIJK, J. P. and SABRA, A. I. (eds). 2003. The enterprise of the Science in Islam: New
perspectives. Cambridge: MIT Press.
IRFAN, H. 2002. Sacred Geometry of Islamic Mosque. Islamonline- News Section. (
http://www.islamonline.net/English/Science/2002/07/article02.shtml, accessed 5 Sept 2007)
JAZBI, S. A. (ed). 1997. Applied Geometry: By Abolvafa mohammad ibn mohammad al-Buzjani
(in Persian and Arabic).Tehran: Soroush Press, ISBN 964 435 201 7.
KATZ, V. J. (ed). 2007. The Mathematics of Egypt, Mesopotamia, China, India, and Islam: A
source Book. New Jersey: Princeton University Press.
KENNEDY, E.S. 1960. A Letter of Jamshid al-Kashi to His Father: Scientific Research at a Fifteenth
Century Court. Orientalia, Nova Series 29, 191-213.
MAINSTONE, R. J. 2001. Vaults, domes, and curved membrances. In Developments on Structural
Form. London: Architectural Press. (p. 124)
MEINECKE, M. 1985. Mamluk Architecture, Regional Architectural Traditions: Evolution and
Interrelations. Damas-zener Mitteilungen 2: 163-175.
MEMARIAN, G. Hossein. 1988. Statics of Arched Structures ( Nîyâresh-e Sâzehâye Tâghî), Vol. 1.
Tehran: Iran University of Science and Technology Press. (In Persian)
MICHELL, G. (ed.). 1978. Architecture of the Islamic world: Its History and Social Meaning. New
York: Thames and Hudson.

318 Maryam Ashkan – Discontinuous Double-shell Domes through Islamic eras in the Middle East and Central Asia:
O’CONNER J. J. and ROBERTSON E. F. 1999. Abu Sahl Waijan ibn Rustam al-Quhi. In School of
Mathematics and Statistics, University of St Andrews, Scotland. (http://www-history.mcs.st-
andrews.ac.uk/Biographies/Al-Quhi.html, accessed 12 April 2009).
O’KANE, B. 1998. Dome in Iranian Architecture. In Iranian Art and Architecture.
(http://www.cais-soas.com/CAIS/Architecture/dome.htm, accessed 22 June 2008).
ÖZDURAL, A. 1995. Omar Khayyam, Mathematicians, and “Conversazioni” with Artisans. Journal
of the Society of Architectural Historians 54, 1: 54-71.
———. 2000. Mathematics and Arts: Connections between Theory and Practice in the medieval
Islamic world. The Journal of Historia Mathematica 27, 2:171-201.
POPE, A.U. 1971. Introducing Persian Architecture. London: Oxford University Press.
———. 1976. Introducing Persian Architecture. In Survey of Persian Art, A. U. Pope and P.
Ackerman (eds). Tehran: Soroush Press.
SAUD, R. 2003. Muslim Architecture under Seljuk Patronage (1038-1327). Foundation for
Science, Technology, and Civilization (FSTC).
(http://www.muslimheritage.com/features/default.cfm?ArticleID=347, accessed 3 Feb 2009).
SEHERR-THOSS, S. P. 1968. Design and Color in Islamic Architecture: Afghanistan, Iran, Turkey.
Washington, D.C.: Smithsonian Institution Press.
SMITH, E. B. 1971. The dome: A study in the History of Ideas. New Jersey: Princeton Press.
STIERLIN, H. 2002. Islamic art and Architecture: From Isfahan to the Taj Mahal. London: Thames
and Hudson.

About the authors

Maryam Ashkan was born in Iran in 1978, and graduated with the M. Arch degree in 2003 from
Islamic Azad University/Qazvin, Dept. of Architecture and Urban Design. She started her career as
conservator architect in Jameh Kavir Construction Company (2003-2006). Since July 2006, she
has been a Ph.D candidate in the University of Malaya, Kuala Lumpur, in the Faculty of Built
Environment specializing in historic Eastern domes in the Middle East and Central Asia.
Associate Professor Dr. Yahaya Ahmad received his Bachelor of Arts in Architecture (1986), Master
of Construction Management (1987) and Master of Architecture (1988) from Washington
University, USA; and his Ph.D in Conservation Management from the University of Liverpool,
UK (2004). He has published many academic papers on conservation and is directly involved in
many conservation projects. He was involved in the drafting of the National Heritage Act (2005)
Malaysia, headed the nomination team to prepare the nomination dossier for Melaka and George
Town for listing on the World Heritage List, and headed expert teams in the re-construction of
Melaka fort. He was seconded to the Department of National Heritage as Deputy Commissioner
of Heritage Malaysia 2007-2009, and was elected as ICCROM Council Member 2007-2011.

Nexus Network Journal – Vol.12, No. 2, 2010 319

 Giulio Magli Research
Dipartimento di Matematica At the Other End of the Sun’s Path:
Politecnico di Milano
P.le Leonardo da Vinci 32 A New Interpretation of Machu Picchu
20133 Milano ITALY
Abstract. The Inca citadel of Machu Picchu is usually
[email protected]
interpreted as a “royal estate” of the Inca ruler Pachacuti.
Keywords: Machu Picchu, Inca This idea is challenged here by a critical reappraisal of existing
architecture, Inca astronomy sources and a re-analysis of existing evidence. It is shown that
such evidence actually point at a quite different interpretation
suggested, on one hand, by several clues coming from the
urban layout (the interior arrangement of the town, the
ancient access ways, the position with respect to the landscape
and the cycles of the celestial bodies in Inca times), and, on
the other hand, by a comparison with known information
about the Inca pilgrimage center on the Island of the Sun of
Lake Titicaca. Altogether, these clues lead us to propose that
Machu Picchu was intentionally planned and built as a
pilgrimage center connected with the Inca “cosmovision”.

1 Introduction
This paper analyzes one of the most beautiful and enigmatic achievements ever
accomplished anywhere in the world by architecture. It is an ancient Andean town whose
original name is unknown; it is famous with the name Machu Picchu. Although it may
seem strange at a first glance for such a renowned archaeological site, the reason why the
town was built, the date at which it was built, the ruler who ordered its construction, the
reason why it was abandoned, in a word, the interpretation of this place, are unknown.
For reasons we do not know, Machu Picchu was abandoned and forgotten; it was
brought again to the attention of the world only with the famous expedition of Hiram
Bingham in 1911 (see [Bingham 1952] or [Salazar-Burger 2004] for an up-to-date
account). Immediately after its “re-discovery” the site was enveloped by a halo of
mystery. Bingham himself thought it to be the “lost capital” of the last Inca reign,
Vilcabamba, an interpretation that we know to be untenable today; various errors and
misunderstandings further contributed to the confusion, such as, for instance, an
exaggeratedly high estimate of the percentage of female bones found in the burials, which
led to the hypothesis that Machu Picchu may have been a sanctuary inhabited by Inca’s
“Virgins of the Sun”. Today, this as well as other, even more unsound theories have been
canceled by modern archaeological research (for instance, the true excess of female with
respect to male bones is around 1.46 to 1). Modern research also helped, for instance, to
clarify the day-to-day life of the inhabitants (see [Burger 2004] and references therein).
However, contemporary with such important developments, nothing has really come out
that helps to explain why and when the Incas built the town (or at least, this is the
opinion of the present author, to be substantiated in what follows).
Living without interpretative schemes is extremely difficult in any science (e.g.,
physics) and archaeology is no exception. Therefore, archaeologists have adopted a
scheme, a sort of dogma, about the true meaning of the town: the idea that Machu
Picchu is to be identified as one of the "royal estates" of the Inca Pachacuti [Rowe 1990].

Nexus Network Journal 12 (2010) 321–341 Nexus Network Journal – Vol.12, No. 2, 2010 321
DOI 10.1007/s00004-010-0028-2; published online 4 May 2010
© 2010 Kim Williams Books, Turin
 Fig. 1a. Map of Machu Picchu highlighting the sites discussed in the text (north on top)

322 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
 Fig. 1b. Machu Picchu. Photograph by the author
What is today customarily called an Inca “royal estate” was a land property,
nominally owned by the king and managed by his family clan. A royal estate was
typically composed of agricultural lands and “palaces” intended to serve as residences for
the ruler and the elite. Such places were used for amusement (such as hunting) and
perhaps also for dealing with state affairs. A good example of royal estates is Chincero,
property of Topa Inca, described in detail by the chronicler Betanzos as a property
“where to go for recreation” and thoroughly analyzed by Niles [1999]. Other important
Inca sites have been interpreted as royal estates as well, in particular Pisac and
Ollantaytambo. To the best of the present author’s knowledge, however, there is no
textual evidence whatsoever showing that Pisac was a royal estate (Pisac is never
mentioned in the Spanish chronicles). More convincing is the case of Ollantaytambo
[Protzen 1993]. This place is in fact associated with the ruler Pachacuti in some Spanish
accounts and, in particular, Sarmiento de Gamboa says that the king “took as his own the
valley…where he erected some magnificent buildings”. As far as Machu Picchu is
concerned, it is easy to get the impression from most of the scholarly work that there
exist firm textual evidence that this town too was one of Pachacuti’s royal estates.
However, as discussed in full detail here in Appendix 1, this is not the case. It is the aim
of the present paper to propose a completely different theory regarding the conception
and construction of the town. To “make room” for such an interpretation I am going to
use a scientific instrument which is typical of a physicist’s way of thinking: Occam’s
razor. According to this principle, what is unproved and unnecessary not only can, but
should be cut away from any scientific approach to a problem. I thus attempt to show
(Appendix 1) that actually there is no proof whatsoever, either textual or archaeological,
that Machu Picchu was built as a private estate for Pachacuti (or, for that matter, for any
other Inca ruler). The interpretation as a royal estate therefore is both unproved and

Nexus Network Journal – Vol.12, No. 2, 2010 323

unnecessary: Occam’s razor allows us to cut it away. Further, although I do not exclude
that the site may have had secondary functions, the “multi-functional” interpretation
adopted by some scholars, who view Machu Picchu as a royal estate but also as a sacred
and perhaps administrative center as well (see e.g. [Reinhard 2007], [Malville and Ziegler
2007]) is also refuted here. The reason again is that this point of view does not help to
explain those characteristics of the site (such the urban layout, for instance) which render
Machu Picchu unique among the Inca towns, and those other characteristics (such as the
clear directionality in access and fruition) which render Machu Picchu almost unique,
the sole possible comparison being a place whose main function had nothing to do with
royal estates or multi-functional centers: the sanctuary on the Island of the Sun (discussed
in § 4).
All in all, as a working hypothesis in what follows I will neglect any pre-existing
interpretation. This means trying to explore the problem of the meaning of Machu
Picchu starting from, and only from, the very beginning: the town itself.
2 A “non-standard” description of Machu Picchu
Machu Picchu lies at 2400 m. above sea level, built like a condor’s nest between the
two twin peaks (Huayna Picchu to the north and Machu Picchu to the south) which
form a sort of peninsula, surrounded on three sides by the gorges of the Urubamba River
some 80 km. northwest of Cusco, the capital of the Inca empire. It would be of course
out of the scope of the present paper to give a full description of the site; however, we
need a clear idea of the urban layout and the relationship of the town with the landscape.
Here we immediately encounter a curious problem. Indeed, since the city is stretched in
an approximate southeast-northwest direction (in conformance to the general direction
of the Machu Picchu-Huayna Picchu ridge), a space-saving device on the printed page
usually results in – starting from the very first Bingham’s plan –north appearing on the ut
lower right. However, this way of mapping the town completely changes the correct
perspective in which the place is – and, most important, was – actually visited by a
newcomer. Further, this unusual orientation makes it difficult to understand at a glance
the orientation of the buildings with respect to the path of the sun and of the stars, as
well as the relationships of the layout of the town with the cardinal points (and the
mountains associated to them). Therefore, although it may seem a somewhat trivial
point, I consider it fundamental for a correct interpretation of the site to refer to a map
that shows north on top (to this aim I have rotated one of the original maps made by
Bingham; see fig. 1).
Further, to understand the layout of Machu Picchu, it is important to keep in mind
that the complete urban plan was conceived, planned, founded and built from scratch,
following the precise will of the planner on the basis of a global, unitary project laid out
after an accurate and complete survey of the area and, in particular, of the natural rock
outcrops which were scattered around. Many such rocks were leveled; others, however,
were “interpreted” in an artistic way, in accordance with that special feeling of the Inca
artists for the stone and, in particular, for the shapes of stones that “replicate” natural
elements. Huge boulders and existing caves were therefore wonderfully integrated into
the project. Each time the planner felt it necessary, the terrain was leveled with the use of
a sophisticated technique of superimposed foundation layers. It is thought that as much
as 60% of the work involved in the construction of the town is buried in its basements
[Wright and Valencia Zagarra 2001]. Finally, it is important to notice that the town is
lacking in water sources, and therefore a careful project was needed to construct an

324 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
aqueduct which brings water from a spring located higher on the north flank of Machu
Picchu mountain.
The town was abandoned when this huge building program was near to completion,
so that some elements – such as for instance the area of the Temple of the Three
Windows – were left unfinished, as it was, probably, another magnificent architectural
project of the Incas, namely the Sacsahuaman “fortress” in Cusco [Protzen 2004]. It is of
course difficult to estimate the length of time employed to bring the site to the present
almost-finished state (to the best of my knowledge, nobody has ever tried to figure out
how long this took). In any case, it is hard to believe that from the beginning of the
planning to the state visible today – with the inclusion of hundreds of agricultural
terraces – it could have taken less than, say, a number of decades (decidedly a long time
to wait for a private estate to be ready).

Fig. 2. Comparison between maps of Machu Picchu and Huanuco Pampa (north
on top in both, but maps are not at the same scale)
The rigorous, unitary project inspiring the construction of the town remained fully
unaltered after the conquest, contrary to what happened to the other Inca settlements.
There is, however, a notable exception: the provincial site called Huanuco Pampa
[Morris & Thompson 1985]. This town was created by the Incas as an administrative
center, and as such it has nothing to do with Machu Picchu from a functional point of
view. However, it is very useful as a term of comparison in order to gain a better
understanding of the inspiring principles of Inca town planning [Hyslop 1990].
Huanuco was founded in the central highlands of Chinchaysuyu, at 3700 m. above sea
level. The Spaniards rapidly gave up the idea of living at such an unfriendly altitude, and
therefore the place was soon abandoned (fig. 2, right). It exhibits four main “quarters”

Nexus Network Journal – Vol.12, No. 2, 2010 325

corresponding roughly to the cardinal directions and surrounding an enormous plaza,
which is empty except for a rectangular building probably used for ceremonial purposes.
The eastern sector of the town – connected to the center by a spectacular east-west
alignment of double-jamb doorways – is the sole part built in fine Inca stonework and
was certainly devoted to the rulers and to ritual activities.
If we compare the two plans, we immediately see that also Machu Picchu was
apparently conceived in a pretty similar, quadripartite way (fig. 2, left).1 The division of
the town into sectors which will be used here (slightly different from the one usually
adopted) is given in fig. 3; in it we can recognize a key difference with Huanuco Pampa:
blocks of buildings at Machu Picchu are present only in two sectors (I and II). Most of
the buildings in these two sectors can be understood through stylistic analyses of Inca
architecture [Gasparini and Margolies 1980; Kendall 1985; Niles 2004]. Indeed most of
the Kanchas (blocks) were clearly conceived as residences of the elite, as shown by the
fine stonework and the presence of double-jamb doorways. In particular, the block of
Sector I, which is furnished with a private garden and direct access to the first of the
gravity fountains (i.e., the purest water), was very probably the private apartment of the
ruler when he visited the site. Scattered around in both these residential quarters are,
however, many buildings which are clearly conceived for ritual purposes, as shown, for
instance, by their astronomical alignments (see § 3).
Proceeding anti-clockwise we encounter Sector III (the northernmost). This Sector
appears to have been intentionally left without buildings. In a sense we can say that it is
actually occupied by the Huayna Picchu mountain. The central plaza ends at the sides of
Huayna Picchu; to the right, near the start of the path ascending to the summit, a visitor
encounters the so-called Sacred Rock complex, a small, leveled plaza closed on the west
by a huge natural rock sculpted so as to resemble the profile of Mount Yanatin, visible at
the distance. Finally, Sector IV contains, instead of blocks of buildings, a sequence of
very peculiar structures. The sequence of such structures as they are viewed by a
newcomer arriving to the town is the following:
1) The gate. Contrary to almost all Inca cities, such as Huanuco Pampa, Machu Picchu
was indeed fenced by a wall, with a doorway near the western end. The function of
the wall was to create a physical separation between the settlement and the outside
without, however, serving a defensive purpose; in other words Machu Picchu was not
“fortified” (accordingly, as mentioned, the town had no springs or water reservoirs).
2) A “disordered” zone where stone-quarrying activities were carried out. It is perhaps
worth noticing that those stones which exhibit regularly spaced drill-bits holes were
not worked by the Incas but are the results of modern “archaeological” experiments;
actually this method – typical of the Romans – was not used by the Incas. They
instead used hard hammer-stones and took advantage of natural fissures of the rocks.
Inca methods are clearly visible in many other points of the same area.
3) The so-called Sacred Plaza, a small space open to the western horizon and closed on
the other three sides. To the east, in particular, one finds the so-called Temple of the
Three Windows, actually a three-sided building. The windows, a spectacular feat of
engineering composed of huge, perfectly dressed polygonal blocks, are located on the
east wall facing the central plaza.
4) The so-called Intihuatana, a steep terraced pyramid on the summit of which lies a
carved stone of white granite.

326 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
 Fig. 3. The proposed “quadripartition” of Machu Picchu; see text for details
3 Machu Picchu and the landscape
Machu Picchu is wonderfully integrated into the landscape; this is so in a quite
“global” sense, including not only the earthly landscape, but also the sky at the time of
construction. The word “landscape” is therefore meant here in the broad sense which is
commonly used today in archaeoastronomy (see e.g. [Magli 2009]).
To describe the relationship between the town and the landscape, it is worth starting
from the analysis given by Reinhard [2007] of the position of the town with respect to
the surrounding mountains. Indeed, as is well known, the mountains were a fundamental
part of the Inca’s huacas (shrines).
To visualize the position of Machu Picchu we can draw a cardinally oriented cross
with the town at the center, and then follow the four cardinal directions (fig. 4). In this
way we meet four important peaks of the region. First of all, north is identified by the
“proprietary” mountain Huayna Picchu, which is, in a sense, a part of the town itself.
The summit (2720 m) was – and is – accessed via a spectacular path, essentially a steep
flight of steps partly carved into the rock which circles the hill on the west. On the top
there are features clearly connected with ritual activities. In particular, a building has its
windows looking out toward the Aobamba-Santa Teresa ridge and the recently restudied
site of Llactapata, where a corresponding construction focuses the view on Huayna
Picchu [Ziegler, Thomson and Malville 2003]. The highest point is shaped as a sort of
arrow which points due south. In this direction of the cardinal cross, the sight at the
horizon is blocked by the Salcantay peak, one of the highest mountain of the region of
Cusco (6271 m). Salcantay was certainly revered already in Inca times and is today
perceived as the brother of the (slightly higher) Ausangate, the highest peak of the Inca
heartland, located east of Cusco.

Nexus Network Journal – Vol.12, No. 2, 2010 327

 Fig. 4. The geographical position of Machu Picchu with respect to “cardinal”
mountains and to the Urubamba river
Important mountains are also located in correspondence to the east and the west of
Machu Picchu. Indeed if we take the (slightly elevated) point of view of an observer
located on the Intihuatana platform, looking east we can immediately discern one of the
two “pyramidal” peaks of Mount Veronica (5750 m). To the west instead, the peaks of
the Pumasillo (6075 m) range span the southwestern horizon, with the northernmost end
of the range located due west.
Interestingly, in accordance with the fact that everything in Inca religion had dual,
complimentary aspects, the relationship of Machu Picchu with the mountains is no
exception. In fact, the mountains, with respect to which the town occupies such a
“special place”, are “replicated” in many sculpted “image rocks” scattered in the town
itself.
Finally, the position of Machu Picchu must also certainly be considered “special”
with respect to the Urubamba (Vilcanota) River, since – as already mentioned – the river
makes a three-sided turn around the Huayna Picchu-Machu Picchu ridge, which appears
therefore as a sort of peninsula marking a neat topographical end to the valley.
To give an overview of the relationships of Machu Picchu with the sky – i.e., of the
archaeoastronomy of Machu Picchu – it is necessary to distinguish neatly between the
“built”, residential part of the town, and therefore the eastern flank (sectors I-II) and the
“ceremonial” part located on the western flank (sector IV). Indeed, in sectors I-II a clear
interest in accurate celestial observations has been documented in many buildings:2
1) The so-called Torreon: a P-shaped building of fine Inca masonry which
encircles a shaped stone. The motion of the rising sun near the winter solstice
could be followed here by means of the shadows cast by a cord affixed at the
window [Dearborn and White 1989]. Also the Pleiades and Scorpio were
probably observed here.
2) The underlying cave (customarily called the Royal Mausoleum) beneath the
Torreon, also aligned to the June solstice sunrise [Malville and Ziegler 2007].
3) The cave usually called Intimachay: a natural cavity with a carefully cut tunnel-

328 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
 window which allows a quite precise measurement of the sun rising at the
summer solstice [Dearborn, Schreiber and White 1987]
4) A general “solstitial” planning of the whole “royal” sector.
People living in the residential sectors had therefore access to devices aimed at
following the timing of the solstices with good precision. Vice-versa, only a generic,
rather symbolic interest in the sky is recognizable in the ceremonial sector. Actually, and
in spite of several existing claims (regarding for example the astronomical function of the
Intihuatana stone) to the best of this author’s knowledge no clear interest in precise
astronomical measurements has ever been convincingly documented here. It is therefore
interesting that astronomical sightings were aimed at the rising of the sun and/or other
celestial bodies, while the likely zone where rites were carried out seems to be suited for
“popular” observations of phenomena at setting. In particular, from the Intihuatana the
highest summit of Pumasillo roughly corresponds to the azimuth of the setting sun at the
December solstice while the sun at the equinoxes is seen as setting at the northern end of
the same range.
Finally, to the possible astronomical observations carried on from Machu Picchu we
should add (as noted by Reinhard [2007]) the fact that, since the culmination of the sun
between the two zenith passages and that of all the stars of the southern portion of the
sky obviously occurs due south, the imposing mountain of Salcantay could be used as a
useful, distant foresight.
The place were Machu Picchu was built therefore seems to have been chosen because
it satisfied an impressive number of geographical/symbolical requirements. It is, of
course, possible – although unlikely – to think that such relationships are due to chance,
or that they were ancillary and are not a fundamental key to explain the choice of the site.
However, the position of Machu Picchu also exhibits another interesting feature. To
discuss it, we observe that, together with the cardinal directions, the directions
characterized by a relatively thin “void” of azimuths between 135° and 155° degrees (as
measured from Cusco) are of tantamount importance in understanding the complex
connections between religion, astronomy, cosmology, and sacred geography among the
Incas [Urton 1978; Zuidema 1982a; Magli 2005].
First of all, it must of course be observed that this “void” characterizes the “preferred”
direction of the orography of the geographical region of interest here, which spans some
500 km. starting from Tiahuanaco, following Lake Titicaca and the course of the rivers
up to the Viracocha temple in Ratqui, then Cusco (the Inca heartland) and, finally,
bends slightly west to follow the Urubamba valley. Machu Picchu stands, in a sense, at
the ideal end of this corridor since the river makes a complete turn around its ridge.
As we shall explain in detail later on, this “void” – clearly not by chance – also
characterized the “ideal direction” of the Inca cosmological myth. These two aspects
merge with yet another very important one, namely, the fact that the “void” happens to
be connected with orientation. Indeed, in the southern hemisphere a “pole star” is never
available (since precession never brings the south celestial pole sufficiently close to a
bright star). As a consequence, the natural way to establish (roughly) the position of the
south pole is – and has always been – to follow a bright star or group of stars up to
culmination. The most natural choice at the Cusco latitudes is to follow those bright
stars of the Milky Way which culminate relatively near the pole, and indeed the principal
constellations of the Incas were located along the Milky Way. Among them, particularly

Nexus Network Journal – Vol.12, No. 2, 2010 329

relevant for orientation were the stars of the constellations called Crux-Centaurus by
today’s astronomers, and the dark constellation of the Llama (a dark zone of the Milky
Way near Centaurus, which looks like the form of a crouching animal) [Urton 1982].
These celestial objects were raising in Inca times (today precession has shifted these values
a bit) just in correspondence to the “void”; for instance Gamma-crux and Alpha-crux
were rising in 1430 AD at azimuths (viewed from Cusco with a flat horizon) around
146° and 152° respectively, while the head of the Llama was rising roughly between 141°
and 151°. It has even been suggested that the rising azimuth of Alpha-crux may have also
influenced the choice of the borderline between the two southern suyus (parts) of the
empire [Urton 1978]. In any case, as we shall see more in depth later on, there is little
doubt that Mayu,, the Milky Way, seen as a huge “double branched” celestial river,
played a fundamental role in the Inca “cosmovision”.
4 Machu Picchu as a pilgrimage center
It has been already proposed by some scholars that Machu Picchu was a sacred center,
visually and symbolically connected with several other huacas of the region and was
therefore a destination of pilgrimages. In particular, this idea inspired Reinhard [2007],
and similar conceits also appear as part of the “multi-functional” interpretation discussed
by Ziegler and Malville [2007]. In both these works, however, the “royal estate” theory is
in any case maintained. As mentioned, it is instead the aim of the present work to
propose that Machu Picchu not only was a sacred center, but that it was conceived,
designed and built specifically to be a place of pilgrimage, the last part of it actually
taking place inside the town. In this respect, at least to the best of the author’s
knowledge, the interpretation proposed here can be considered as new.3
To develop this interpretation, it is fundamental to start from a comparison between
Machu Picchu and the unique Inca pilgrimage site which is historically well documented
and has been the subject of a exhaustive study: the Island of the Sun [Bauer and Stanish
2001].
The Island of the Sun is a rocky islet located near the southern end of Lake Titicaca.
For some reason that nobody has ever dared to investigate, an apparently insignificant,
natural rock formation present in the northern part of this island was identified as
nothing less than the place of origin of the sun, and therefore of the Incas: this place
actually appears, although with different details, in most versions of the Inca
cosmological myth. As a consequence, a very important Inca sanctuary was located on the
island. The whole site was administrated by the state, and the Incas removed the existing
population replacing them with colonists from various parts of the empire; also, a
specialized group of women was established with the purpose of serving the sanctuary.
The sanctuary area of course included the “sacred rock” where the sun was born, which
was the final destination of the pilgrims. The pilgrimage took place in successive stages
(fig. 5):

1) Pilgrims gathered at what is today Copacabana, and then sailed to the island
from the south.
2) Once landed they followed a path oriented – as the island – in a southeast-
northwest direction. The path ultimately brought them to the most sacred part,
which was enclosed by a (low) wall.
3) Apparently not all of the pilgrims were allowed to pass the entrance to the

330 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
 northern part of the island; those not admitted could still have a look anyway at
the rites from terraces appositely leveled out of the wall.
4) After the wall, the sacred path passed by some other gates and “stations”. In
particular, near the building today called Mama Ojlia, the pilgrims could look
at the “footprints of the sun”, an area where the exposed bedrock contains
natural marks resembling huge footprints.
5) Finally, people gathered in the plaza in front of the Sacred Rock, where they
witnessed rites and, at the time of the winter solstice, the hierophany of the sun
setting between two pillars on a ridge to the northwest.

Fig. 5. Map of the Inca sanctuary on the Island of the Sun. The position of the
pillars used to observe the winter solstice sunset from the sanctuary area is
indicated. After [Bauer and Stanish 2001]
Bauer and Stanish describe the layout of the site as follows:
It is not by chance that the final destination of the pilgrims was on the
point of land farthest from the mainland, on the northwest side of the
island of the Sun. The sanctuary was, like many pilgrimage centers of the
world, situated in a remote location that served to emphasize its
otherworldliness [Bauer and Stanish 2001: 247].
Clearly, the very same words may be applied to Machu Picchu, which is located in a
similar way: it suffices to substitute the lake surrounding the Island of the Sun with the
three-sided gorge of the Urubamba River. Actually, as we shall see, the list of similarities
is much longer and more impressive than this. Let us, indeed, turn our attention to
Machu Picchu again.

Nexus Network Journal – Vol.12, No. 2, 2010 331

 There are two main ways to approach Machu Picchu coming from the Inca
heartland. The first one is the world-famous, spectacular route – usually called the Inca
trail – which departs from the Urubamba River at Pikillacta (Llactapata), some 30 km.
from the Machu Picchu ridge. It leads to the Inca site of Winay-wayna and then to the
so-called Intipunku, a building clearly meant to provide a monumental, controlled access
and located south of Machu Picchu, in plain view of the town.
The “Inca trail” is indeed certainly Inca, since it is in many points carved in the rocks
and passes through or near several Inca settlements. However, it is a very uneasy route,
rising for example as much as 4200 m to cross the “Dead Woman Pass”. The journey
from Cusco to Machu Picchu following this route certainly took at least four full days for
the Inca on his human-transported chair and his retinue. However, especially in the dry
season (May to September, which is also the season during which Machu Picchu is
supposed to have been visited by the Inca) there is another, much easier and natural route
from Cusco to Machu Picchu: just follow the Urubamba valley up to the Machu Picchu
ridge and then ascend up to Winay-wayna and the town. This path may become
dangerous or even impassable in case of heavy rains, but in the dry season it allowed a
three-day, quiet journey from Cusco. This one was obviously also an “Inca trail”, and
indeed several Inca ruins are scattered near the river (e.g., Torontoy). Interestingly, a
further Inca path ascending directly from the river to the town has recently been found in
the deep forest on the east flank of the Machu Picchu ridge [Valencia Zagarra 2004].
This path as well is endowed with artistic fountains and resting spaces, confirming a
“cerimonial” in addition to a practical function.
All in all, it appears that the “classical” Inca trail was conceived mainly as a
ceremonial route, not as a functional one, at least in the dry season. Its very existence
contributes to the casting of serious doubts on the royal estate theory: indeed the ruler
traveling to Machu Picchu to spend the winter (dry) season there did not need to take
such a long, uneasy route to arrive to his (in any case very distant) estate.
Let us now re-join the path of the pilgrims approaching Machu Picchu from
Intipunku. All the roads (including that coming from the northwest) meet at the so-
called upper Agricultural Area outside the main gate. Here a huge kallanca, Machu
Picchu’s largest building, is located. Clear traces of ritual activities are present here, in
particular a “replica stone” which, as does the sacred rock of the town, replicates Cerro
Yanatin. In the area there are also a number of piles of small stones of different natures,
and at least some were evidently carried to the site from distant places; for instance, a
number are rounded river rocks. Probably, thus, at least some of these stones are offerings
left upon arrival. The whole area presents clear similarities with the area located outside
of the innermost sanctuary on the Island of the Sun as it offers an obstructed view on the
western sector of the town and the plaza. We can thus speculate that, exactly as occurred
on the Island on the Sun, people who were not admitted further could still look anyway
at the events taking place in the town from the platform and the terraces. Curiously, the
sector of the Sacred Rock is not visible from here, and perhaps for this reason a “replica”
of the same monument was created in this place.
Continuing their way down, the pilgrims eventually reached the gate of the town.
The town itself was thus characterized as a closed space, and this is unusual for the Incas
(actually at Machu Picchu we encounter most of the few examples of Inca stone
doorways which have devices – rings and recesses – meant for closing them with wooden
lintels).

332 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
 It is in any case clear that the fencing wall was not meant for defensive purposes, since
it would have been exceedingly easy for any enemy to block the water channel (which
comes in the town from outside) with just a few stones and in Machu Picchu there are no
water reservoirs. The aim of the wall is therefore, as we said, to furnish a controlled access
and, consequently, to stress the separation of what is inside from what remains outside.
The visitors admitted perhaps left their ritual offerings just near the entrance wall,
since many peculiar stone pebbles (mainly of obsidian) have been recovered there. Then,
people were confronted by a corridor (service buildings such as stables and magazines are
located on the left of the entrance but are separated by a wall) with the imposing view of
the Huayna Picchu mountain, the likely final destination of the pilgrimage, just in front
of them.
Today, at this point most tourists turn right, visiting the “residential sectors” first.
However, in ancient times these sectors – which begin with the Royal residence – were
very likely closed to public; therefore, a person entering would have had by necessity to
proceed straight in the western sector encountering the sequence of structures we became
familiar with in the preceding section: first, the so-called quarry; second, the Temple of
the Three Windows and, finally, the Intihuatana platform. But why?
It is my aim here to maintain that sector IV in Machu Picchu, oriented and
“scheduled” towards north and Huayna Picchu, was conceived as an image, a replica of
the path followed by the first Incas in the cosmological myth. This myth has been
recounted, although sometimes with different details, in many chronicles. Is starts at the
Island of the Sun, from which the first Incas traveled underneath the earth following the
“void” (southeast-northwest) direction, emerging at a place called Tampu-tocco.
According to Sarmiento de Gamboa [1999] and Cobo [1983], this name means “the
house of windows” because “it is certain that in this hill there are three windows” and the
first Incas came out from one of these windows. Tampu-tocco is located south of Cusco:
the Incas now traveled therefore to the north, up to the summit of the Huanacauri hill,
where one of them was turned into stone, becoming a fundamental huaca of the future
empire before arriving in the Cusco valley.
There is, at least in my view, little doubt that the key elements of the myth find a
close correspondence in the structures of Sector IV. In fact, a newcomer encounters in
succession the three main elements of the myth, symbolized respectively by:
1) The quarry. Although it is a true quarrying area, it is also a zone intentionally
left in “disorder” located very near the “royal” and the “ceremonial” sectors. It
shows signs of ritual activities as well, such as carvings of serpents on the rocks.
It is thus tempting to associate it with Pachamama and to think that it was
somehow connected with an image of the underground travel of the first Incas.4
2) The sacred plaza. Here the Temple of the Three Windows is to be associated
with Tampu-Tocco.
3) The Intihuatana pyramid. This might have been conceived as a replica of the
Huanacauri hill, the Intihuatana itself resembling in shape the sacred mountain
Huayna Picchu located at the end of the path [Reinhard 2007] as well as the
sacred stone-huaca which was located on Huanacauri.
It should be noted that, inspired precisely by the Temple of the Three Windows,
Hiram Bingham proposed that the town had to be identified with Tampu-tocco. It is not

Nexus Network Journal – Vol.12, No. 2, 2010 333

my intention here to revive his theory. In fact, there is no doubt about the identification
of the Huanacauri hill as one of the most important huacas of the Cusco ceque system
[Bauer 1998] and actually most chronicles associate Tampu-tocco with the Pacariqtambo
hills south of Cusco (the imposing Inca site of Maukallakta was probably built there to
recall the mythical events, see [Bauer and Stanish 2001]). The idea here is rather that
Machu Picchu was conceived, as many other sanctuaries around the world, as a powerful
and tangible replica of the holy places of the myth. It is in fact well known that
sanctuaries are often built to offer “replicas” related to the same sacred event and/or place
in different sites and at different stages of monumentality (think, for instance, of the
various “copies of the Holy Land” constructed in Europe during the Middle Ages and the
Renaissance). It is an easy process to render such places sacred as well, for instance with
the presence of holy images, relics, or oracles.5 In this connection, note the following
passage appearing in the chronicle of Bernabè’ Cobo:
The Incas had founded a town on the site of Pacariqtambo, and they built
on it, in order to make it famous, a magnificent royal palace with a
splendid temple. The ruins of this palace and temple remain even today,
and in them some stone statues and idols are seen. At the entrance of that
famous cave at Pacariqtambo there is a carefully cut stone window in
memory of the time when Manco Capac left through it [Cobo 1983].
This passage is somewhat strange. First of all, it hardly fits with what is visible – at
least today – in Pacariqtambo, while it could be easily referred to Machu Picchu. Further,
it gives the impression that the author is speaking about a place where he has never been.
It is actually difficult to believe that “statues and idols” could have been left on a site so
near to Cusco and known to the Spaniards at the time Cobo was writing.
The connection between “power and replica” is especially true for those sanctuaries
which were aimed at the foundation of the temporal power in Durkheim’s [1912] sense,
such as, without a doubt, those administered by the Inca state. Again, the best known
example is that on the Island of the Sun, where cyclical ritual activities took place at
scheduled times during the year, with the culmination at the winter solstice hierophany.
It is thus tempting to attribute a similar role and function at Machu Picchu as well, and
to conclude that the place was administered by a dedicated group of “priest-astronomers”
(amautas ) in charge of “controlling” the whole course of the pacha – the sacred “space
and time” of the Incas – as shown by the astronomical alignments of the buildings of the
residential sectors (which include both solstices).
Finally, although only at the level of what may be simply an interesting coincidence,
it can be noted that there are impressive similarities between the NW sector of Machu
Picchu and some seventeenth-century drawings appearing in the so-called Miccinelli
Documents [Laurencich-Minelli 2006, 2007] and in the Poma de Ayala Chronicle.
Actually, as already noted several times in the literature (see e.g. again [Reinhard 2007])
the representation of the Huanacauri idol in the rites of the Inca month of March
depicted in this document closely resembles the Intihuatana (fig. 6, left). Of course,
many huacas of this type existed before the conquest; however, the resemblance becomes
more impressive if we also consider the Poma representations of Tampu-tocco (fig. 6,
right). This place is indeed shown with its three windows at the base of the Huanacauri
hill. Perhaps the choice of the month is not casual either: March is the only case in which
the Inca name of the month mentioned by Poma exhibits the suffix -pacha [Laurencich,
Minelli and Magli 2010]. The equinox was a moment at which the sun (hanan) and the

334 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
moon (hurin – thought of as representing the dark hours) were equilibrated having the
same “strength”; in this month sacrifices to Viracocha were made and, according to
ethnographical data, the full moon before the equinox was the time of a feast devoted to
the hanan dead, who were in turn associated with the summit of the mountains.

Fig. 6. Poma de Ayala folii 242 (right) and 266 (left)

5 Discussion
As explained in detail in Appendix 1, any interpretation of Machu Picchu is destined
to remain speculative unless we eventually find out for sure how the town was called by
the Incas and discover written texts mentioning both the town and its function in an
explicit way. Clearly, however, the interpretation proposed in the present paper would be
strengthened if it helps to explain why the town was built in such an accurately chosen,
special position. We are thus led to investigate if it is possible to identify a link between
topography on the earth and celestial cycles in the sky at the times of the Incas.
As a general observation, it must be noticed that these kinds of issues are extremely
delicate in archaeoastronomy in general. Actually, very few of the proposed examples of
monuments or human-built landscapes anywhere in the world associated with “terrestrial
images” of the sky are securely proved (see [Magli 2009] for a complete discussion).
However, in the case of the Incas we have a quite solid starting point which relies on the
already mentioned fieldwork of Urton [1982] among the contemporary Misminay
people. This research has made particularly evident the fundamental role played by the
Milky Way. In the centuries before the Incas, precession brought the solstitial points near
the intersections between our galaxy and the ecliptic (fig. 7).

Nexus Network Journal – Vol.12, No. 2, 2010 335

 Fig. 7. The intersection of the Milky Way with the Ecliptic. The positions of the
solstices in the Milky Way, respectively in Sagittarius and near Gemini, are shown.
Four star-to-star “cross” constellations of today’s Misminay people are also shown:
a) a combination of stars in the area of Gemini/Orion; b) our Southern Cross; c)
the Belt of Orion plus probably Rigel; d) five stars in the head of Scorpio. From
[Urton 1982]

Fig. 8. The pilgrimage to the Vilcanota Hill according to Zuidema [1982]

336 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
 As a consequence, the position of the sun with respect to the Milky Way could be
used to estimate the times of the solstices, which were, in turn, fundamental dates among
the yearly Inca rituals.6 The bright arc of our galaxy in the sky was perceived as a
“celestial river” having a “terrestrial counterpart” in the Vilcanota River, which, in
Urton’s words, is “equated with the Milky Way”. This is, therefore, a hint to the
possibility that Machu Picchu, located in such a special position with respect to the
Vilcanota river, may have had a special significance connected with the sky as well.
Actually, written documentations of two Inca pilgrimages along the course of the
Vilcanota exists in the work of the chronicler Cristobal De Molina. The first (studied in
detail in [Zuidema 1982b]) moved south of east from Cusco and was connected with the
cosmological myth (this pilgrimage was in all likelihood a Capac-Cocha, namely, it
ended with a human sacrifice). It occurred at the winter solstice and was directed “toward
the place where the sun was born” (fig. 8). The final station was a huaca located on the
summit of the Vilcanota hill, some 150 km. as the crow flies from the capital at the
border between the Vilcanota and the Titicaca watershed. Coming back, the priests
followed the course of the river, bringing offerings to other huacas. The second
pilgrimage was instead directed north of west, i.e., in the opposite direction, and
occurred a few weeks after the summer solstice. During the night, illuminated by torches,
runners started from Cusco and reached a bridge in Ollantaytambo, where Coca leaves
were offered to the river.
In both cases, a fundamental role was played by three key elements: the opposite
solstice timing, the opposite “void” direction, and the Vilcanota river alongside it. The
two pilgrimages thus took place on the earth, along the terrestrial image of the Milky
Way, towards two “ends” which – due to the timing – were clearly connected with the
two “ends” of the sun’s path at the horizon, the journey of our star throughout the fixed
stars during the year. Of course, the path of the sun is the Ecliptic, not the Milky Way,
but the starting and ending points (the winter and the summer solstice) were both
located at the crossroads between the two. Interestingly, the southern pilgrimage ended at
the sources of the river, but still far from the “true” place of birth of the sun, the Island of
the Sun. Perhaps a similar role was played by Machu Picchu with respect to
Ollantaytambo in the case of the northern pilgrimage; actually, the “inter-cardinal”
position of the town with respect to the landscape also recalls the presence in the sky of
cruciform star-to-star Inca constellations, which were concentrated near the two solstitial
points (see again [Urton 1982] and fig. 7).
All in all, the hints of different nature and origin presented in this paper all seem to
point in the same direction. They suggest that Machu Picchu could have been conceived
and built as the ideal – hanan – counterpart of the Island of the Sun, in accordance with
the duality of the sacred which is typical of the Andean world.7
Machu Picchu was indeed located at the ideal, opposite crossroad between the
terrestrial and the celestial rivers: the other end of the sun’s path.
Acknowledgments
The ideas set forth in this paper benefited very much from several discussions with
my friends and colleagues Laura Laurencich Minelli, who taught me the concept of “Inca
space-time”, and Jean-Pierre Protzen, whose comments greatly improved a first draft of
the present work.

Nexus Network Journal – Vol.12, No. 2, 2010 337

Appendix 1. Review of the evidence for the “royal estate” theory.
As mentioned in the introduction, an Inca “royal estate” was a land property,
nominally owned by the king and managed by his family clan, endowed with agricultural
lands and residences for the ruler and the elite. The very same interpretation has been
repeatedatly proposed for Machu Picchu, and it is the aim of this appendix to give what I
hope is a complete overview of the evidence – or better, of the almost complete lack of
evidence – for this idea.
First of all, to gain any textual information about an ancient place we should know
how it was called. As a matter of fact however, we cannot be sure about how the town
now known as Machu Picchu was called by the Incas. Indeed, we only know that it lies
between two paired peaks which were called by inhabitants of the area – at the moment
of Bingham’s “discovery” – by the names of Machu Picchu, the old peak, and Huayna
Picchu, the young one. These names were reported to Bingham and he decided to name
the town after one of the peaks. However, a town with this name (or, for that matter,
with the similar name Picho; see below) is never mentioned in the Spanish historical
chronicles (at least to the best of the author’s knowledge). Of course, although often
confused and/or captious, the chronicles constitute the fundamental written sources of
information about Inca life and history. Therefore, no attempt to identify the role and
the meaning of the town can be based on historical sources, and any attempt is,
consequentially and necessarily, admittedly speculative (this holds of course for the
proposals set forth in the present paper as well).
That said, we can proceed to analyze the “royal estate” hypothesis. This hypothesis
was formulated by Rowe [1990] starting from the fact that a place called Picho (Picchu)
is mentioned in some sixteenth-century documents. In particular, it appears in the
account of a journey to Vilcabamba. In this manuscript the author, Diego Rodriguez de
Figueroa, cites Picho – apparently without having visited it – as a place located on a
certain path between Condormarca and Tambo. The same place is also mentioned in a
legal document of 1562, first discovered by Glave and Remy [1983] where the existence
of a Picho cacique (local chief) farming coca is also cited. On the basis of this evidence,
Rowe proposed that Bingham or his informants had confused the true name of the town
– which he thinks was originally called Picho – with that of the nearby mountain.
This interpretation appears reasonable, although it is, of course, difficult to be certain
of its validity for a series of reasons. First of all, the name Picchu (peak) is exceedingly
common : for instance, one of the sacred hills northwest of Cusco is called precisely
Picchu. Further, it is at least possible that the place was located in the same area of
Machu Picchu but down in the valley, along the course of the Urubamba river, to which
the administration of agricultural activities referred. Finally, it cannot be forgotten that
there is no archaeological evidence for the presence of Spaniards in the town at any time,
in spite of existing claims; further, there are no visible signs of intentional destruction of
Inca “idols” – a fact exceedingly common in all the Cusco huacas.
In any case, even if we admit that Machu Picchu was called Picchu and therefore
mentioned in these writings, we are in any case very far from having any interpretation of
the town, and the idea of the “royal estate” has to be justified by other means. To this
end, attention is called to the fact that a list of agricultural properties attributes the
farmed lands of the Urubamba valley between Ollantaytambo and Chaullay to the
private possession of their conqueror, the Inca Pachacuti. Picho is not explicitly
mentioned in this context, but “since the terrains of the valley bottom belonged to
Pachacuti, it is quite probable that the places at higher quotas in the same zone were part

338 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
of the royal estate of the king as well” ([Rowe 1990], my translation). Again, however,
even if admit the validity of this second implication – and of course, we are not obliged
to do this – the fact that the citadel was part of the royal properties does not show that it
was used as a personal estate of the Inca. This too is, therefore, an inference, and indeed
Rowe cautiously concludes “we can suppose [podemos suponer] that the Inca ruler
choose Machu Picchu as a personal estate and as a memorial of war campaigns in the
zone of Vitcos” (my translation). In spite of such correct prudence, this paper is
commonly considered (perhaps without its having been read) as giving conclusive proofs
for the interpretation of Machu Picchu as a royal estate of Pachacuti. Instead, it is clear
that this is only a sort of vague possibility based on as many as three interdependent
implications (Machu Picchu was called Picho -> Picho was a property of Pachacuti ->
Pachacuti built it as his personal royal estate), none of which is securely proved.
Of course, together with the written sources, it has to be ascertained if archaeological
evidence exists to support the theory. As far as the monumental architecture of the town
and its relationship with the landscape and the sky are concerned, this ascertainment is
fully developed in the main body of the present paper. Here I will instead mention some
recent studies which may be of relevance for the interpretation of the town (see [Burger
2004] and references therein). One is the re-analysis of the bones found by the Bingham
expedition. As already mentioned, the percentage of female bones was originally over-
estimated; however, a 3:2 excess of female bones remains. It is known that the Incas
established a group of women in the sanctuary of the Island of the Sun whose specific
role was to serve the sanctuary; perhaps the same might have happened in Machu Picchu.
The re-analysis of human remains has also shown that the inhabitants (which should
have numbered around 750) exhibited an high degree of population diversity. They
therefore came from the different parts of the empire; likewise, a significant part of the
pottery recovered in tombs comes from distant provinces. This is relevant for our
discussion because it is known that royal estates were staffed by mitima, retainers of
different ethnic groups removed from their original villages. However, although valid,
this observation cannot be used as a proof for the “royal estate” theory, since the very
same thing holds for many other Inca state projects and in particular, again, for the
sanctuary on the Island of the Sun. Finally, another relevant recent analysis is the
reconstruction of the ancient climate at Machu Picchu, which turned out to have been
quite similar to that of today: warmer than in Cusco, but also more rainy. It has to be
definitively excluded, therefore, that Machu Picchu could have been considered a more
pleasant place to stay than the capital during the rainy season (October to April), while a
relative improvement of the climate temperature might actually have been obtained by
moving to Machu Picchu in winter.
Notes
1. It is worth noting that trials have been made in the past to divide the plan of the town into the
two traditional moieties – the upper, or hanan, and the lower, or hurin – which are typical of
the organization of the Andean space. For instance, the capital Cusco was ideally divided
according to this principle. Contrary to Cusco, however, where this division is well
documented in the chronicles and essentially corresponds to the northern and the southern
parts respectively, at Machu Picchu the situation is unclear. In any case, the standard proposal
is to identify the hanan part as that including the “royal residence” and the quarry (sectors I
and IV in fig. 3).
2. The level of precision of such measurements has been the subject of much debate starting from
a paper by Dearborn and Schreiber [1986]. This topic is not of specific relevance here however;
the interested reader can consult the anthology recently edited by A. Aveni [2008].
3. The possible existence of other functions for the town (for instance, administrative) are not

Nexus Network Journal – Vol.12, No. 2, 2010 339

 excluded a priori, but have to be considered as subsidiary to the main one. Similarly, although
the “royal estate” theory is refuted here, this of course does not mean that the Inca did not visit
the town. On the contrary, it is very likely that he did, and the evidence for a “royal residence”
is actually quite compelling.
4. A similar “symbolic + functional” interpretation has been proposed by who writes for the
famous quarry of Ranu Raraku on Easter Island, where several huge statues were left at
different stages of extraction and completion [Magli 2009].
5. It can be noted that important oracular shrines have been documented in the Andean world
since extremely ancient times, e.g., in Chavin and Pachacamac, and that some of the caves at
Machu Picchu (especially the Temple of the Condor) may suggest the same interpretation.
6. It is worth recalling that the coincidence between the solstices and the intersections of the
Ecliptic with the Milky Way was fundamental in Maya cosmology, see e.g., [Schele, Freidel
and Parker 1995]
7. People living in the northern hemisphere associate “north” naturally with “up”, because their
celestial pole – and therefore the center of the apparent rotation of the stars, the place where
the axis mundi meets the sky – is the northern one. Interestingly, it is apparent that north was
associated with “upper” (hanan) by the Incas as well, although they were living in the southern
hemisphere. The anthropological reasons for this fully merit, at least in the author’s view,
further investigations.

References
AVENI, A. 2008. Foundations of New World Cultural Astronomy: A Reader With Commentary.
Boulder: University Press of Colorado.
BAUER, B. 1998. The Sacred Landscape of the Inca: The Cusco Ceque System. Austin: University
of Texas Press,.
———. 2004. Ancient Cuzco: Heartland of the Inca. Austin: University of Texas Press.
BAUER, B., and DEARBORN, D. 1995. Astronomy and Empire in the Ancient Andes. Austin:
University of Texas Press.
BAUER, B., and STANISH, C. 2001. Ritual and Pilgrimage in the Ancient Andes: The Islands of the
Sun and the Moon. Austin: University of Texas Press.
BINGHAM, H. 1952. The lost City of the Incas: The Story of Machu Picchu and its Builders.
London: Phoenix House.
BURGER R. and SALAZAR-BURGER L.C. (Eds) 2004) Machu Picchu: Unveiling the Mystery of the
Incas. New Haven and London: Yale University Press.
COBO, B. 1983. History of the Inca Empire (1653). Roland Hamilton, trans. Austin: University of
Texas Press.
DEARBORN, D. S. and SCHREIBER, K. 1986. Here Comes the Sun: The Cuzco–Machu Picchu
Connection. Archaeoastronomy 9:15–37.
DEARBORN, D.S., SCHREIBER S. and WHITE, R. 1987. Intimachay: A December Solstice
Observatory at Machu Picchu, Peru. American Antiquity 52, 2: 346-352.
DEARBORN, D.S. and WHITE, R. 1989. Inca observatories: their relation to the calendar and ritual.
Pp. 462-469 in World Archaeoastronomy, A.F. Aveni, ed. Cambridge: Cambridge University
Press.
DE GAMBOA, S. 1999. History of the Incas (1572). Clements Markham, trans. New York: Dover
Publications.
DURKHEIM, E. 1912. The Elementary Forms of Religious Life . Rpt. 1965, London: The Free
Press.
GASPARINI, G. and MARGOLIES, L. 1980. Inca Architecture. Bloomington: Indiana University
Press.
GLAVE, L. and REMY M. 1983. Estructura agraria y vida rural en una región andina.
Ollantaytambo entre los siglos XVI y XIX. Archivos de Historia Andina. Cusco: Centro de
Estudios Rurales Andinos Bartolomé de las Casas.
HYSLOP, J. 1990. Inka settlement planning. Austin: University of Texas Press.
KENDALL, A. 1985. Aspects of Inca architecture: description, function and chronology. Oxford:

340 Giulio Magli – At the Other End of the Sun’s Path: A New Interpretation of Machu Picchu
 British Archaeological Reports.
LAURENCICH MINELLI L. 2006. UEl mito utopico de Paititi desde un documento jesuitico
parcalmente inedito del siglo XVII. Archivio per l’ Antropologia e la Etnologia CXXXV: 183-
202.
———. 2007. Exsul Immeritus Blas Valera Populo Suo e Historia et Rudimenta Linguae
Piruanorum. Indios, gesuiti e spagnoli in due documenti segreti sul Per del XVII secolo.
CLUEB Bologna.
LAURENCICH MINELLI, L. and MAGLI, G. 2010. A calendar Quipu of the early 17th century and its
relationship with the Inca astronomy. Archaeoastronomy 22 (in press). Pre-print
arxiv.org/abs/0801.1577.
MAGLI, G. 2005. Mathematics, Astronomy and Sacred Landscape in the Inka Heartland. Nexus
Network Journal 7, 2: 22-32.
———. 2009. Mysteries and discoveries of Archaeoastronomy. New York: Springer-Verlag.
MALVILLE, J. M., and ZIEGLER, G. 2007. Machu Picchu, Inca Pachacuti’s Sacred City: A multiple
ritual, ceremonial and administrative center. http://www.adventurespecialists.org/mapi1.html.
Accessed 23 March 2010.
MALVILLE, J. M., THOMSON, H. and ZIEGLER, G. 2004. El observatorio de Machu Picchu:
Redescubrimiento de Llactapata y su templo solar. Revista Andina 39: 9- 40.
NILES, S. 2004. The nature of Inca royal estates. Pp. 49-68 in Machu Picchu: Unveiling the
Mystery of the Incas, R.Burger and L.C. Salazar-Burger, eds. New Haven and London: Yale
University Press.
NILES, S. 1999. The shape of Inca History. Iowa City: University of Iowa Press.
PROTZEN, J.P. 1993. Inca Architecture and Construction at Ollantaytambo. Oxford: Oxford
University Press
———. 2004. The fortress of Saqsa Waman: Was it ever finished? Ñawpa Pacha 25–27: 155-175.
REINHARD, J. 2007. Machu Picchu: Exploring an Ancient Sacred Center. New York: Cotsen
Institute of Archaeology.
ROWE, J. 1990. Machu Picchu a la luz de documentos de siglo XVI. Historia 16, 1: 139-154.
SCHELE, L., FREIDEL, D., and Parker, J. 1995. Maya Cosmos. New York: Quill.
URTON, G. 1978. Orientation in Quechua and Incaic Astronomy. Ethnology 17, 2: 157-167.
———. 1982. At the Crossroads of the Earth and the Sky: An Andean Cosmology. Austin:
University of Texas Press.
WRIGHT, R. M. and VALENCIA ZEGARRA, A. 2001. The Machu Picchu Guidebook. Boulder:
Johnson Books.
ZUIDEMA R .T 1982a. Catachillay: The Role of the Pleiades & Southern Cross and Alpha and Beta
Centauri in the Calendar of the Incas. Pp. 203-230 in Ethnoastronomy and Archaeoastronomy
in the American Tropics, Anthony F. Aveni & G. Urton, eds. New York: New York Academy
of Sciences, vol. 385.
———. 1982b. Bureaucracy and systematic knowledge in Andean civilisation. Pp. 419-458 in The
Inka and Aztec States 1400-1800, G.A. Collier, R.I. Rosaldo and J.D. Wirth, eds. Studies in
Anthropology. New York: Academic Press.

About the author

Giulio Magli is a full professor in the Faculty of Civil Architecture of the Politecnico of Milan,
where he teaches the only official course of Archaeoastronomy ever established in Italy. He earned a
Ph.D. in mathematics at the University of Milan in 1992 and his research activity developed in the
field of General Relativity Theory, with special attention to problems of relevance in astrophysics.
In recent years, however, his research interests have focussed mainly on archaeoastronomy, with
special emphasis on the relationship between architecture, landscape and the astronomical lore of
ancient cultures. On this subject he has written several papers and the book Mysteries and
Discoveries of Archaeoastronomy, published in 2005 (in Italian) by Newton & Compton editors,
and in English edition by Springer-Verlag (2009).

Nexus Network Journal – Vol.12, No. 2, 2010 341

 Tessa Morrison Research
The School of Architecture The Body, the Temple and the
and Built Environment,
The University of Newcastle Newtonian Man Conundrum
Callaghan, NSW, 2308
Abstract. From his early days at the University of Cambridge until
AUSTRALIA
his death, Isaac Newton had a long running interest in the Temple
Tessa.Morrison@newcastle.
of Solomon, a topic which appeared in his works on prophecy,
edu.au
chronology and metrology. At the same time that Newton was
Keywords: Isaac Newton, working on the Principia, he reconstructed the Temple and
Temple of Solomon, commented on the reconstructions of others. An important part of
Vitruvius, Vitruvian man, his investigations concerned the measurements of the Temple,
Newtonian man which were harmonic and were built “exactly as the proportion of
proportions architecture demands.” Newton considered these proportions to
be in accordance with Book III and IV of De Architectura.
However, while insisting on exact architectural proportions,
Newton moved away from the traditional proportions of the
Vitruvian man; he derived a Newtonian man. This poses an
interesting conundrum: Newton accepted the Temple’s
architectural proportions as outlined in Vitruvius’s Book III, yet
he rejected the human model Vitruvius used as the foundation of
these proportions. At the same time Newton accepted the human
frame as the basis of all ancient measurements and attempted to
estimate the length of the sacred cubit based on the lengths of the
parts for the body and the measurements set out by the ancient
writers such a Vitruvius.

The Temple of Solomon, Babson Ms 0434 and ‘A Dissertation upon the Sacred
Cubit of the Jews’
When Newton died in 1727 he left hundreds of unpublished manuscripts, some
dating back to his early days in Cambridge in the 1660s. Newton’s heirs invited Thomas
Pellett to examine the manuscripts and report on their suitability for publication. After
just three days of examining these hundreds of manuscripts, Pellett dismissed the
majority of manuscripts as “not fit to be printed” [Gjertsen 1986: 426], “of no scientific
value” and as “loose and foul papers” [Manuel 1974: 14]. He only found two sets of
manuscripts suitable for publication. The first were two manuscripts on prophecies, and
although Pellet claimed that the text on prophecy was imperfect, they were nevertheless
worthy of publication. This was eventually prepared for press by Newton’s nephew
Benjamin Smith [Gjertsen 1986: 399] and published in 1733 as The Observations upon
the Prophecies of Daniel and the Apocalypse of St John. Observations proved to be one
of Newton’s best sellers in the eighteenth century and it was also translated into Latin
and German shortly after its first edition [Hall 1992: 372] .
The second was a set of manuscripts on chronology, which were compiled and
arranged by John Conduitt, husband of Newton’s niece Catherine, and published in
1728 as Chronology of Ancient Kingdoms Amended. The book cannot be considered a
success and is exceptionally dull. It is arranged in six chapters, five of which were
chronologies of the ancient empires of Greece; Egypt; Assyria; Babylon, and Persia. The
other chapter is a description of Solomon’s Temple, which is not only an intriguing

Nexus Network Journal 12 (2010) 343–352 Nexus Network Journal – Vol.12, No. 2, 2010 343
DOI 10.1007/s00004-010-0029-1; published online 6 May 2010
© 2010 Kim Williams Books, Turin
 addition to a book on chronology of ancient kingdoms, but is curiously placed after the
chapter on the Babylonian Empire, which destroyed the Temple.1
The beginning of the chapter on the Temple is quite dismissive: “The Temple of
Solomon being destroyed by the Babylonians, it may not be amiss here to give a
description of that edifice” [Newton 1988: 332]. The chapter consists of a brief (barely
3,000-word long) description of its floor plan, with three illustrated floor plans. There is
no mention of the style of architecture, its splendour or its significance. The description
lacks any enthusiasm and is highly clinical. Its architectural description is problematic
and there are parts that do not make structural sense. For example in the Chronology
Newton claimed that:
The porch of the Temple was 120 cubits high, and its length from south
to north equaled the breadth of the House: the House was three stories
high, which made the height of the Holy Place three times thirty cubits,
and that of the Most Holy three times twenty [Newton 1988: 342-343].
Since the porch, the Holy Place and the Most Holy of Holies adjoined each other,
this description created a strange and confused stepped structure which appears to have
no precedents, Biblical or otherwise. The three illustrated floor plans are very detailed,
but that detail is not backed up by the text. Furthermore, both the text and the
illustrations include an external wall that surrounds the precinct wall, with four gates on
the western side, the Gate of Shallecheth, the Gate of Parbar, and the two Gates of
Assupim. But these were part of the Second Temple and not Solomon’s Temple [2
Samuel 6:11-12].
From this chapter it would be easy to conclude that Newton had no knowledge of
architecture and that his interest in Solomon’s Temple was only as a biblical symbol and,
with its destruction, an important historic event. However, the converse is true. Not only
did Newton have a good working knowledge of Vitruvian theory, he had a long running
interest in the Temple of Solomon that spanned over fifty years.
Over this fifty years Newton wrote many manuscripts that related to the Temple (for
example, [Newton undated(a), undated(c), ca. mid-1680s and ca. 1690s]). Two
manuscripts are of primary interest for this paper. The first was written in the mid-1680s
and was entitled by Newton “Introduction to the Lexicon of the Prophets, Part two:
About the appearance of the Jewish Temple”, more commonly known by its call number
Babson Ms 0434. The other is an earlier unpublished manuscript of the 1680s, an
appendix entitled “De magnitudine cubiti sacri” [Newton c1680s(b)], which is a part of a
draft on Solomon’s Temple that developed into Babson MS 0434. In 1737 this appendix
was translated from Latin into English and published as “A Dissertation upon the Sacred
Cubit of the Jews”, in Miscellaneous Works of John Greaves Professor of Geometry at
Oxford [Newton 1737].
“Dissertation” is a work of metrology. Newton examined the measurements taken by
John Greaves (1602-1652), Savilian Professor of Astronomy at the University of Oxford,
who conducted a survey of the Pyramids of Giza which resulted in the publication of
Pyramidographia in 1646. Greaves’s measurements in English feet, taken at the
Pyramids, were used to calculate the Royal cubit, Memphis cubit and the Egyptian cubit.
From Greaves’s calculations of the ancient cubits, Newton proceeded to calculate the
measurement of the Jewish sacred cubit, which was essential to understanding the
Temple structure.

344 Tessa Morrison – The Body, the Temple and the Newtonian Man Conundrum
 Newton’s “Dissertation” begins: “To the description of the Temple belongs the
knowledge of the sacred cubit; to the understanding of which, the knowledge of the
cubits of the different nations will be conducive” [Newton 1737: 405]. Newton used
Greaves’s measurements of the Great Pyramid and systematically compared them with
measurements given by ancient sources such as Herodotus, Vitruvius, Strabo, Josephus,
Hesychius of Alexandria, Lucius Iunius Moderatus Columella, Philandrier, Gnaeus Julius
Agricola, Publius Clodius Thrasea Paetus, the Talmud and more contemporary writers
such as Willebrord Snellius, Samuel Purchas and Juan Bautista Villalpando. Newton also
cited Arabic sources, such as Ibn Abd Alhokm (321-405) [Newton 1737: 408].
In Babson Ms 0434 Newton systematically reconstructed the Temple of Solomon.
His primary source for his reconstruction was the Book of Ezekiel. However, Newton
examined the changes in the Temple over time. The building of the second Temple by
Zerubbabel followed the same foundations but with a great deal less grandeur. It had the
same dimensions and was a pragmatic house of worship, but its architecture was
mundane and it was nothing to look at [Newton c1680s(a) 5r]. Cyrus the Great ordered
the building of the Temple and the internal atrium but nothing else was added. This was
the sanctuary that was maintained up to the time of Alexander the Great, as reported by
the pagan writer Hecataeus. The Temple was further fortified under Simeon the Just,
until Herod built a more sumptuous building for the sanctuary. According to Newton,
“God, predicting all these things, thus he corrected them through the prophet Ezekiel”
[Newton 1680s(a): 7r]. But Ezekiel left out details of the building and the Angel who
revealed the Temple and its measurements to Ezekiel did not show him the entire
Temple. Thus by examining the architectural features through time and with the writing
of ancient writers, such as Philo, Hecataeus, Josephus, Maimonides, the Talmud and the
Septuaginta, Newton was able to reconstruct the Temple of Solomon, removing the
features that had been added by the later builders. He stated “we complete the
description of the Temple [of Solomon by] comparing all the Temples to each other and
supplying what Ezekiel omitted relative to the Temples of Solomon and of Herod”
[Newton c1680s(a): 59r]. From this description of the Temple, Newton claimed that it is
possible to distinguish the plan of the Temple of Solomon. Since Zerubbabel had built
on the foundations of the Temple of Solomon, everything that Zerubbabel and Herod
added, or anything that is irregular, must be rejected. Symmetry and harmony in the
design of the Temple were important factors in the layout of the Temple plan. He stated
that, “The structure is made valuable by such great simplicity and harmony of all its
proportions” [Newton c1680a: 65r].
Babson Ms 0434 is a working document; it is incomplete and it contains two
reconstructions, the second of which, the more detailed of the two descriptions, is a
refinement of the first. The illustration of the Temple precinct in Babson Ms 0434 is in
fact the first reconstruction and the second reconstruction is only verbally expressed but
in sufficient detail to reconstruct it. In both reconstructions, symmetry was of
paramount importance to the floor plan.
Newton’s knowledge of Vitruvius, the measurements of the body and the Jewish
cubit
In his reconstruction Newton not only outlined the structure of the Temple, he
examined the colonnades: the numbers of columns, their height, their thickness, their
intervals and their style. These he claimed were determined according to the proportions
of architecture. Newton revealed that he was familiar with the architectural theory of

Nexus Network Journal – Vol.12, No. 2, 2010 345

Vitruvius’s De Architectura, particularly Book III and IV. When Newton derived the
width of the intercolumniations from the measurement of the column given by Josephus,
he paraphrased De Architectura Book III, Chapter III, “The Proportions of
Intercolumniations and of Columns”. He stated, “The intervals of these pedestals,
according to the proportions of architecture, should not be less than the pedestals”
[Newton c1608s(a): 29r]. From Vitruvius, Book IV, Chapter III and the measurements
of Josephus, Newton estimated the height of the columns as being “six times the
thickness according to the Doric style” [Newton c1680s(a): 36r]. In Ezekiel 40:14 the
measurement of the height of the doorway is given as twenty cubits; thus Newton
concluded that “the width of the doorway was of ten cubits and the height according to
the rules of the architects, should be double the width” [Newton c1680s(a): 45r;
Vitruvius 1960: IV, vi, 6]. For Newton most of the measurements of the Temple are
“exactly as the proportion of architecture demands” [Newton c1680s(a): 10r]. However,
according to Newton the architecture of the Temple sometimes surpassed the beauty that
classical architecture demands. He confirmed that there was a row of twenty-one
columns and twenty inter-columns in the Royal colonnade from the measurement
described by Josephus. Newton stated:
the Royal colonnade will occupy seventeen, twenty or twenty-four [spaces]
between the columns of the same magnitude. But seventeen, according to
the architectural proportions, will be too few, and twenty-four will be
excessive if the columns were estimated to be equal to those of the other
atriums, and, in one and another case, are set too far apart by the numbers
of Josephus, therefore it should be twenty [intercolumniations]. According
to this proportion, the columns will be less numerous than in the
proportion of the eustyli of Vitruvius, but more beautiful; and here, where
instead of the architrave there are large blocks of marble that cannot be
broken, it does not fit the objections of Vitruvius [Newton c1680s(a): 37r]
(this and all other quotes from Babson 0434 are translated from the Latin
by the author).
Harmony and symmetry in the design of the Temple were important elements in the
layout of the Temple plan. Any element that was described by the ancient writers that
was irregular had to be rejected. He stated that, “The structure is made valuable by such
great simplicity and harmony of all its proportions” [Newton c1680s(a): 62r]. The
perfection of the measurements was of paramount importance to the design.
In Babson Ms 0434 Newton evaluated the measurements, revealing their
proportional perfection. In the Book of Ezekiel, the prophet is guided by an angel
measuring each part of the Temple as they move around the Temple precinct. The
measurements of the Temple were in Jewish cubits. There were two types of Jewish
cubits, described by Newton as the sacred cubit and the vulgar cubit. The description of
the cubits in Ezekiel is very confusing as he claimed that, “The cubit is a cubit and a
palm breadth”, leaving the distinction between the two cubits ambiguous. In 2
Chronicles 3:2, Solomon instructed that the Temple be built in cubits “after the first
measure”. A cubit was equal to the length of the forearm from the elbow joint to the end
of the middle finger. This simple measurement is inscribed in Egyptian hieroglyphs. The
hieroglyph for a cubit is the image of the forearm [Glazebrook 1931: 413].
In “A Dissertation upon the Sacred Cubit of the Jews”, Newton not only attempted
to resolve this ambiguity, he also attempted to estimate the length of both cubits. He

346 Tessa Morrison – The Body, the Temple and the Newtonian Man Conundrum
quoted Vitruvius’s measurement of the Roman and Greek cubits as being one and a half
Roman feet [Newton 1737: 405; Vitruvius, 1960: III, i]. A Roman foot is 0.97 of an
English foot. Newton examined both the Roman and Greek cubits and feet; measuring
them in palms and digits, together with the Greek orgyiae, the span of the arms fully
outstretched, because these measurements were defined by the ancient authors. To
estimate their value Newton approached the problem of the variations in the
measurements of the ancient authors by assessing the limits of each and then comparing
them to each other.
Newton reasoned that the builders of the Great Pyramid would have used a uniform
unit of measurement in their design, which would have been the ancient measurement of
a cubit. In his calculations he claimed that one Greek orgyiae is equal to four Memphis
cubits. After converting the measurement of Greaves from English feet to Roman palms
and digits he then compared them with the measurements of the ancients. From this
Newton concluded, “And it is my opinion that the Pyramid was built throughout after
the measure of this (Memphis) cubit” [Newton 1737: 413]. Newton supported this
argument that the ancient buildings were built to a standard unit of measurement by
considering the measurement of Babylonians bricks. They were all uniform in size,
according to the measurements of sixteenth-century travel writer Samuel Purchas, their
length was one foot, the width was eight inches and the thickness was six inches. Thus
the combination of two brick lengths, three brick widths and four brick thicknesses
formed square cubits.
Newton claimed that all measurements which exceeded human proportions, such as
the Roman calamus, clima, scruplum, actus and many others, were deduced from the
multiples of human proportions. The ancient nations rounded off their large numbers
into even numbers of cubits – the cubit of man [Newton 1737: 416]. Newton derived
the length of many nations’ cubits: Memphis, Egyptian, Greek, Roman, Arabian and
Babylonian. Although the different lengths of these cubits conformed to the cubits of
man, with the exception of the Babylonian cubit which he estimated at two English feet
[Newton 1737: 414], this may be an error of inscription, since later in the paper he
referred to the Babylonian cubit as being two Roman feet, but both measurements are
larger that the human elbow; still, he continued to maintain that the Babylonians built in
cubits.
Greaves found that the modern Egyptian cubit was 1.824 English feet, exceeding that
of the ancient Egyptian cubit or Memphis cubit. “The measurements of feet and cubits
now exceed the proportion of the human members” [Newton 1737: 417]. According to
Greaves’s measurements of the Egyptian monuments, the human stature was the same as
it was in ancient times. The measurements had increased in length because of human and
instrument error.
Feet and cubits were first used (as a measurement) in every nation
according to the proportion of the members of a man, from which they
were taken. For the foot of a man is to the cubit or lower part of the arm
of the same man as about 5 to 9 [Newton 1737: 419].
Newton confirmed this ratio 5:9 between the foot and cubit with other ancient
measurements. He considered that the Jewish measurements were determined in the
same manner.

Nexus Network Journal – Vol.12, No. 2, 2010 347

 He claimed that the Jewish vulgar cubit cannot exceed the cubit of a tall man.
Newton claimed that:
The stature of the human body, according to the Talmud, contains about
three cubits from the feet to the head; and if the feet be raised, and the
arms lifted up, it will add one cubit more and contain four cubits. Now
the ordinary stature of men, when they are barefoot, is greater than five
Roman feet, and less that six Roman feet, and may be best fixed at five feet
and an half [Newton 1737: 421].2
According to Erubin 48a in the Talmud, the area of “his place” is “three cubits for his
body and one cubit to enable him to take up an object at his feet and put it down at his
head.” Newton moved away from the classical “Vitruvian” man. In Vitruvius, the height
of man is set at six Roman feet; Vitruvius claimed that the number six was perfect and
this perfection was further expressed in the Roman cubit, which equaled six palms or 24
digits, but to Newton and the Bible six palms was a sacred Jewish cubit. Newton’s
measurements of the stature of a man, five to six Roman feet, equaled three vulgar Jewish
cubits of five palms each cubit; thus a vulgar cubit was to be no less than 20 Roman
unciæ3 and no more than 24 unciæ,4 also from this measurement the sacred Jewish cubit,
of six cubits, he calculated to be no less than 24 Roman unciæ and no more than 28.8
unciæ.

Fig. 1. The Newtonian Man

348 Tessa Morrison – The Body, the Temple and the Newtonian Man Conundrum
 Newton gave two examples from ancient literature, where he further defined the
limits of the sacred cubit. In the first, Josephus wrote that the columns of the great court
of the Jewish Temple could be embraced by three men with their arms joined. Vitruvius
stated, “For if we measure the distance from the soles of the feet to the top of the head,
and then apply that measure to the outstretched arms, the breadth will be found to be the
same as the height” [Vitruvius 1960: III.i.3]. However, Newton claimed that although
the orgyia, or the length of the outstretched arms of a man, was supposed to be the same
as the height of a man, in fact it was a palm wider [Newton 1737: 422]. Newton
abandoned the traditional image of the Vitruvian man, which is circumscribed by the
circle and the square, by adding an extra palm to the length of a man’s outstretched arms,
giving a slightly more elliptical and rectangular image to the geometry of man (fig. 1).
The circumference of the columns, according to the Talmud and Josephus is eight cubits;
for Newton, this is equal to three times the height of a man plus three palms, i.e., greater
than 15.75 Roman feet and less than 18.75 Roman feet. This further defined the sacred
cubit as greater than two Roman feet and less than two and a third Roman feet.
In Newton’s second example of the use of the cubit from the ancient literature, the
Sabbath-day’s journey, in the opinion of what Newton called the ‘unanimous’ content of
the Talmud and all the Jews, was two thousand cubits. According to Josephus, this
measurement is not so consistent; in one place he claimed that the Sabbath-day’s journey
is five stades (three thousand Roman feet) and in another place, six stades (three
thousand-six hundred Rome feet) [Josephus 1963: V.2.3; XX.8.6]. Newton, who was
very familiar with the work of Josephus, used the reference from the Talmud instead and
claimed that instead of “cubits” the Jews sometimes substituted “paces”. Walking on the
Sabbath is not hurried but is of a moderated speed: “Now man of a middling stature, in
walking in this manner, go every step more than two Roman feet, and less that two and a
third. And within these limits was the sacred cubit circumscribed” [Newton 1737: 424].
Turning to Vitruvius for the correct architectural height of a step [Vitruvius 1960:
III.iv.4], Newton claimed that the middling proportion referred to by the Jews was about
13.5 unciæ and from this he calculated that a pace or sacred cubit was more that 24
unciæ and less than 27 unciæ. From the examples of the height of a man, the
circumference of the columns and the Sabbath-day’s walk, Newton defined the limits of
the sacred cubit and rejected “the erroneous opinions of other writers”. Newton
concluded that the vulgar cubit was five palms, the cubit of man was equal to 21.4 unciæ
or 1.717 English feet, while the sacred cubit was six palms [Newton 1737: 427], the
cubit of man plus a palm, was equal to 25.6 unciæ or 2.068 English feet.
The conundrum of the Newtonian man
Unlike many commentators of his time Newton does not directly include or refer to
any anthropomorphic element in his reconstruction of the Temple, where the figure of
man/God was reflected in the measurements and geometry of the Temple, which
prefigured the perfection of the mystical body of the Church. While Newton insisted on
exact architectural proportions, he moved away from the traditional proportions of the
Vitruvian man, which had been an important element in other contemporary
reconstructions. This poses an interesting conundrum: Newton accepted the Temple’s
architectural proportions as outlined in Vitruvius’ Book III yet he rejected the human
model Vitruvius used as the foundation of these proportions. At the same time, Newton
accepted that the human frame was the basis of all ancient measurements, and he

Nexus Network Journal – Vol.12, No. 2, 2010 349

attempted to estimate the length of the sacred cubit with the lengths of the parts of the
body and the measurements of ancient writers, such as Vitruvius.
When writing Babson Ms 0434 and “Dissertation upon the Sacred Cubit of the
Jews” he was at the height of his intellectual power and was completing the first edition
of the Principia. The rejection of the Vitruvian man as a model for the proportion of the
Temple cannot be dismissed as an oversight by Newton. Newton was aware of Book III
of De Architectura and the image of the Vitruvian man was also well represented in
architectural text. Furthermore, the images of the cosmological/Vitruvian man were
strongly linked in the Renaissance with the rise of Hermetic philosophy; and in Newton’s
unpublished papers he demonstrated an interest in the symbolism of Hermetic
philosophy [Newton undated(a), undated(b), undated(c); Anonymous (Trismegist)
2002]. The framing of the model from the Book of Erubin in the Talmud and adding a
palm to the span of a man’s outstretched arms must have been a conscious alternative.
Conclusion
Newton’s manuscripts on the Temple span over fifty years, and the majority of these
papers are theological rather than architectural in nature. However, the architecture of
the Temple plays an important role in Newton’s work on the language of the prophets.
The prophets could only be interpreted through “hieroglyphs” [Newton 1957] and one
of those hieroglyphs was the framework of the architecture and rituals of Solomon’s
Temple. This view of the Temple is not only confirmed by his unpublished papers but
also by Newton’s title for Babson Ms 0434, “Introduction to the Lexicon of the
Prophets, Part two: About the appearance of the Jewish Temple”. Newton believed that
the ancient religion, which he claimed was the original religion of God, understood the
mathematical principle of God’s orderly design which sustained the solar system. He
perceived that they had a pure knowledge of the workings of natural philosophy [Newton
c1690s]. The symbol of the Temple was important to Newton, and he returned to the
topic many times over the fifty year period. At first he refined it, but eventually, towards
the end of his life, he sanitized his work. Before his death, Newton was preparing the
Chronology of Ancient Kingdoms Amended for publication; by then he had a legacy to
maintain, and to maintain it he sanitized a lot of his work, disguising his religious beliefs
[Westfall 1980: 817] which at the time were heretical and would have seen him publicly
disgraced.5 The chapter on the description of Solomon’s Temple in the Chronology had
become so sanitized that it is virtually nonsensical and had lost the brilliance that
characterised his early work.
The conundrum of the Newtonian man is an interesting puzzle but it is one that has
no solution, for Newton left no clue as to why he moved away from the traditional image
of the Vitruvian man. However, what Babson Ms 0434 and “A Dissertation upon the
Sacred Cubit of the Jews” does reveal is that he had a good working knowledge of
Vitruvian theory and an interest in architectural aesthetics, are two aspects of his
character that are not normally associated with one of the greatest scientists in history.
Notes
1. There has been some excellent research carried out on the theological implications of the role
of the Temple of Solomon in Newton’s work (see [Mandelbrote 1993 & 2007; Faur 2004;
Goldish 1998]). However, none of these have examined Newton’s knowledge of architectural
and Vitruvian theory. In particular, “A Dissertation upon the Sacred Cubit of the Jews”
discussed in this paper, although frequently referenced as proof of Newton’s interest in the

350 Tessa Morrison – The Body, the Temple and the Newtonian Man Conundrum
 length of the Jewish cubit (for example [Leshem 2003: Popkin 1992; Westfall, 1980]), is a
neglected paper that has not been discussed in its own right.
2. Newton’s references to the Talmud are incorrect. His reference is Mishnaioth, Tract. De
Ghaburim, cap. 4; it should be Talmud, Erubin, 48a.
3. An unciæ is a Roman inch, 20 unciæ equals 1.612 English feet.
4. 24 unciæ equals 1.934 English feet.
5. William Whiston, former pupil and successor to Newton as Lucasian Professor at Cambridge,
also shared Newton’s beliefs. However, he made his beliefs public, which ended his career at
Cambridge. He was later charged with heresy and, although not convicted, he never held an
academic position again [Force 1985].
References
ANONYMOUS (Hermes Trismegist). 2002. Tabula Saragdina. Pg. 274 in The Janus Faces of
Genius, B. J. T. Dobbs, ed., Isaac Newton, trans. Cambridge: Cambridge University Press,
274.
BIRCH, Thomas. 1737. Miscellaneous Works of John Greaves. London: Printed by J.Hughs for J.
Brindley.
FORCE, James E. 1985. William Whiston. Cambridge: Cambridge University Press.
FAUR, Jose. 2004. Newton, Maimonidean. Review of Rabbinic Judaism 6, 2/3: 215-49.
GLAZEBROOK, Richard T. 1931. Standards of Measurement, Their History and Development. The
Proceedings of the Physical Society 43.
GJERTSEN, Derek. 1986. The Newton Handbook. London and New York: Routledge & Kegan
Paul.
GOLDISH, Matt. 1998. Judaism in the Theology of Sir Isaac Newton. London: Kluwer Academic
Publishers.
GREAVES, John. 1646. Pyramidographia: Or, a Description of the Pyramids in Aegypt. London.
HALL, A. Rupert. 1992. Isaac Newton: Adventurer in Thought. Cambridge: Cambridge University
Press.
JOSEPHUS. 1963. Jewish Antiquities. 9 vols. London: William Heinemann Ltd.
LESHEM, Ayval. 2003. Newton on Mathematics and Spiritual Purity. International Archives of the
History of Ideas, 183. Dordrecht and Boston: Kluwer Academic Publishers.
MANUEL, Frank E. 1974. The religion of Isaac Newton. Cambridge: Cambridge University Press.
MANDELBROTE, Scott. 1993. A Duty of the Greatest Moment: Isaac Newton and the Writing of
Biblical Criticism. The British Journal for the History of Science 26, 3: 281-302.
———. 2007. Isaac Newton and the exegesis of the Book of Daniel. Pp. 351-375 in Die
Geschichte der Daniel-Auslegung in Judentum, Christentum und Islam, K. Bracht and D.S.
du Toit, eds. Berlin: Walter de Gruyter.
NEWTON, Isaac. undated(a). Yahuda Ms 1.1 (unpublished manuscript). Jerusalem: Jewish National
and University Library.
———. undated(b). Irenicum (unpublished manuscript). Cambridge: King’s College.
———. undated(c). Notes from Ramon Lull (unpublished manuscript). Stanford: California,
Stanford University Library.
———. undated(d). Experimenta Raymundi (unpublished manuscript). Cambridge: King’s
College.
———. undated(e). Tabula Smaragdina and Hieroglyphica Planetarum (unpublished manuscript).
Cambridge: King’s College.
———. ca.1680s(a). Introduction to the Lexicon of the Prophets, Part two: About the appearance
of the Jewish Temple (unpublished manuscript). Babson Ms 0434. Wellesley, MA: Babson
College.
———. ca. 1680s(b). Drafts Concerning Solomon’s Temple and the Sacred Cubit (unpublished
manuscript). Yahuda Ms 2.4. Jerusalem: Jewish National and University Library.
———. ca. mid-1680s. The First Book Concerning the Language of the Prophets (unpublished
manuscript). Keynes Ms 5. Cambridge: King’s College.
———. 1690s. The Origins of Religions (unpublished manuscript). Yahuda Ms 41. Jerusalem:
Jewish National and University Library.

Nexus Network Journal – Vol.12, No. 2, 2010 351

———. 1737. A Dissertation upon the Sacred Cubit of the Jews. In Miscellaneous Works of John
Greaves Professor of Geometry at Oxford. London.
———. 1957. Of an universall language. Translation and commentary by Ralph W. V. Elliott.
The Modern Language Review III, 1 (1957): 1-18.
———. 1988. The Chronology of Ancient Kingdoms Amended. London: Histories & Mysteries
of Man.
POPKIN, Richard Henry. 1992. The Third Force in Seventeenth-century Thought. Leiden and
New York: E. J. Brill.
VILLALPANDO, Juan Bautista and Jerónimo del Prado. 1604. Ezechielem Explanationes Et
Apparatus Urbis Hierolymitani Commentariis Et Imaginibus Illustratus. Rome.
VITRUVIUS. 1960. The Ten Books on Architecture. Morris Hicky Morgan, trans. New York:
Dover Publications.
WESTFALL, Richard S. 1980. Never at Rest: A Biography of Isaac Newton. Cambridge: Cambridge
University Press.

About the Author

Dr Tessa Morrison is an Australian Research Council post-doctoral Fellow in the School of
Architecture and Build Environment at the University of Newcastle, Australia, Her academic
background is in history, mathematics and philosophy and she has published articles on geometric
and spatial symbolism, and architectural history. Her current research project is “Isaac Newton’s
Temple of Solomon and his analysis of sacred architecture: An interpretation and discussion of
Babson Manuscript 0434”. This project focuses on Newton’s architectural interests and places it
into perspective with his other works and into the context of his times.

352 Tessa Morrison – The Body, the Temple and the Newtonian Man Conundrum
 Book Review
Nigel Hiscock
The Symbol at Your Door: Number
and Geometry in Religious
Architecture of the Greek and Latin
Middle Ages
Aldershot: Ashgate, 2007
Reviewed by Areli Marina
Keywords: church, baptistery, School of Architecture
basilica, centrally-planned University of Illinois
building, chapter house, Middle 117 Temple Buell Hall, MC 621
Ages, meaning, symbolism, 611 Lorado Taft Drive
geometry, harmonic proportion, Champaign, IL 61820 USA
measurement, number, Platonic [email protected]
solids, sacred geometry, shapes

In his previous book, The Wise Master Builder: Platonic Geometry in Plans of
Medieval Abbeys and Cathedrals (Aldershot: Ashgate, 2000), Nigel Hiscock sought to
demonstrate that Platonic geometry underlay the ground plans of Western European
Romanesque cathedrals and abbey churches, but refrained from arguing that these
geometric and numerological foundations had symbolic meanings. In The Symbol at
Your Door, he sets out to prove that medieval church design is pervaded with numbers
and geometric forms that express metaphysical concepts central to Christian neo-
Platonism. In Hiscock’s view, this symbolism was deliberately inscribed into the
structures by the buildings’ designers at the behest of their patrons. He argues that
religious buildings that can be parsed into measures that correspond to Pythagorean
number theory and broken down into fragments that coincide with Platonic geometry
exist in such large numbers that the correspondences cannot be coincidental, but rather
must have been intended by either patrons or builders. Furthermore, because there is so
much evidence that well-educated medieval persons associated particular numbers, ratios,
and geometric forms with Christian significance (such as the association of the number
three with the Trinity, the square with the shape of the Holy Jerusalem, or the sphere
with heaven), Hiscock concludes that patrons and designers must have meant to elicit
those meanings in the mind of their viewers.
As the author acknowledges, this controversial hypothesis contradicts an influential
strain of medieval architectural historiography that asserts that medieval builders used
geometry and number purely for functional and not for representational purposes. This
school of thought is exemplified by the work of Lon R. Shelby, John Harvey, François
Bucher, or (more recently) Robert Bork, and emerges from the French nineteenth-
century rationalist tradition of Eugene Emmanuel Viollet-le-Duc. Hiscock positions
himself in a different genealogy, identifying his conceptual roots in the iconographic
investigations of Richard Krautheimer, Erwin Panofsky, and Paul Frankl, among others.
In fact, The Symbol at Your Door can be read as a prolonged meditation on questions
suggested by Frankl in “The Secrets of the Mediaeval Masons” in 1945 (Art Bulletin 27,

Nexus Network Journal 12 (2010) 353–355 Nexus Network Journal – Vol.12, No. 2, 2010 353
DOI 10.1007/s00004-010-0031-7; published online 6 May 2010
© 2010 Kim Williams Books, Turin
n. 1, pp. 46-60): What means did medieval builders use to develop their buildings’ plans,
elevations, and decoration? How were these ideas communicated in the workshop? Why
were the equilateral triangle and the square so important to medieval design practice?
How were the metaphysical concepts about shape and measure presented by Plato in the
Timaeus transmitted in the Middle Ages? What do these practices reveal about the
relationship between the underlying geometry, workshop practice, aesthetic pleasure, and
religious ideology? Although these are not new questions, most have never been
conclusively resolved, and Hiscock’s in-depth exploration of them is a welcome addition
to the discourse.
The author divides his investigation into four parts. In the substantial Prologue, he
examines the historiography of medieval architectural history and introduces the reader
to Pythagorean and Platonic thought, its transmission in the Middle Ages, and its impact
on architectural culture. This section digests material treated more fully in The Wise
Master Builder. It outlines the relationship between arithmetic and the numbers
perceived to be at the root of universal order (particularly 1, 2, 3, and 4) and the
meanings attributed to them by ancient philosophers and Christian theologians.
Chapters 1 and 2 constitute the second part of the book, in which the author explores the
significance of two solid geometric forms, the sphere and the cube, as well as their
deployment in cruciform churches. Chapters 3 through 6 are dedicated to the
architectural use and meaning of plane figures: the equilateral triangle and the numbers
derived from it (3, 6, 12); the square, square schematism, quadrature, and its associated
forms, including the 1:2 rectangle; the pentagon; and, in the last and longest chapter,
circles, octagons, and other polygons. The book closes with an Epilogue in which the
author summarizes his findings and considers their post-medieval implications.
To counter the arguments of the “rationalist school” and prove his hypothesis,
Hiscock surveys and presents two bodies of classical and medieval textual sources and
juxtaposes them with his analyses of medieval buildings. The first body of textual sources,
including citations from authors such as Augustine, Boethius, and Dionysus, the Pseudo-
Aeropagite, demonstrates that number and geometric form conveyed specific
metaphysical meaning to educated audiences. The second shows that when writing
ekphrases and other eulogistic texts on buildings, or using buildings as metaphors,
medieval authors such as Eusebius and William Durandus often adopted the same
numerological and geometric topoi. The range of sources cited by Hiscock is impressive;
the bibliography includes 157 citations of ancient and medieval works. Relatively few of
these writers, however, present direct evidence about a specific building at or around the
time of its construction, and Hiscock focuses on those examples only briefly (pp. 37-38).
He is most persuasive when discussing the abundance of evidence relating to a single
building structure, such as Hagia Sophia in Istanbul (Chapter 1), or when he introduces
new questions, such as why the expressive content of the stone tracery of stained glass
windows is seldom considered when discussing their iconographic programs (Chapter 6).
The book’s impressive breadth of scope is also its principal weakness. Hiscock
conjures so many examples, cites so many authors, and illustrates so many buildings to
prove his point that the vigor of his argument is dissipated in seemingly endless, often
redundant, and occasionally contradictory exposition. Even the author seems to get lost
in this thicket of evidence, suggesting, for example, that the burial and commemorative
function of English cathedral chapter houses did not account for the choice of a
polygonal plan on page 255, only to assert on page 257 that because chapter houses
functioned as memorial shrines, they often imitated the octagonal form of baptisteries,

354 Areli Marina – The Symbol at Your Door: Number and Geometry in Religious Architecture of the Greek and Latin...
mausolea, and martyria. Also, his persistent use of the passive voice obscures the agency
with which he wants to endow patrons, designers, and builders, and often obfuscates his
meaning. Judicious editorial pruning would have focused the narrative, and resulted in a
more legible and ultimately more convincing text.
In the end, Hiscock cannot quite dispel the lingering doubt that, although highly-
educated clerics and patrons may have been well-versed in the metaphysical aspects of
number and geometric form, there is little consistent proof that either designers, builders,
or audiences were equally familiar with them, much less across several centuries and from
the British Isles to Asia Minor. Indeed, in cases where texts directly addressing geometry
and number in relation to active workshops survive, such as the well-known debate
regarding the design of Milan cathedral, they are silent on the subject of meaning and
symbolic content (see especially pp. 366-370).
The philosopher Karl Popper has suggested that scientific theories cannot be proven
true; at best, scientists can only refute untrue hypotheses. The volume of evidence
amassed by Hiscock powerfully suggests, but does not prove, a strong causal relationship
between the existence of Pythagorean and Platonic numbers and figures in medieval
religious buildings and the symbolic interpretations of the buildings and building
components made by their contemporaries. As in Frankl’s day, the weak link remains the
connection between the mind of the patron and the builder’s hand (Frankl, 50). Insights
into this aspect of the problem may lie, not within the purview of architectural history,
but in the hands of our colleagues exploring the histories of medieval education and
science.
In sum, The Symbol at Your Door is an ambitious book that is not afraid to engage a
complex and controversial question. Hiscock’s introduction to the metaphysics of
number in the medieval period and the veritable anthology of medieval writings on
architecture that he has assembled will be useful launching pads for future studies of how
(or perhaps whether) the numbers and geometric forms embedded in specific medieval
religious buildings had symbolic meaning. The wide range of his analyses reminds us of
the richness and diversity of signification that architectural forms convey to their past and
present audiences.

About the reviewer

Areli Marina, who trained as an art historian at the Institute of Fine Arts, New York University,
teaches medieval and Renaissance architectural history at the University of Illinois. Her research
focuses on the intersection of public rhetoric, national identity, and civic art production,
particularly in relation to the semiotics of architecture and urban form; the problematic
historiography of the Romanesque, Gothic, and Renaissance styles; and the role of antiquity in
medieval and Renaissance art and architecture, with particular emphasis on northern Italy. Her
work has been supported by a Rome Prize Fellowship at the American Academy in Rome, a Getty
Foundation Fellowship, the Gladys Krieble Delmas Foundation, and the University of Illinois. Dr.
Marina has recently completed a book on the medieval piazza, The Italian Piazza Transformed:
Parma’s City Center in the Communal Age, which will be published by Pennsylvania State
University Press. During 2010-11, she hopes to make significant progress on a new book on
medieval and Renaissance baptisteries while a Villa I Tatti fellow at the Harvard Center for Italian
Renaissance Studies in Florence.

Nexus Network Journal – Vol.12, No. 2, 2010 355

 Conference Report
Architecture and Mathematics.
A seminar to celebrate Professor
emeritus Staale Sinding-Larsen’s 80th
birthday
Trondheim, Norway, 25 November 2009
Report by Eir Grytli
Keywords: architecture and
mathematics, Staale Sinding- Faculty of Architecture and Fine Art
Larsen Norwegian University of Science and Technology
Trondheim, NORWAY
[email protected]

Trained as an art historian, Professor Staale Sinding-Larsen held the chair of

architectural history at the Norwegian University of Science and Technology (NTNU,
former NTH), from 1970 to 2000. During this period he also served for seven years as
Director General of the Norwegian Institute in Rome. Despite his age, Professor
Sinding-Larsen is still an active researcher and writer, with a continuing curiosity and
open-mindedness towards the advantages of cross-disciplinary approaches to architectural
research. On the occasion of his 80th anniversary, his colleagues at the Faculty of
Architecture and Fine Art wished to honour him with a seminar focusing on one of the
topics that is closest to Professor Sinding-Larsen’s heart, namely the relationship between
architecture and mathematics. In order to make the interdisciplinary research on the
meeting-point between architecture and mathematics visible, the seminar was organized
in co-operation between the Faculty of Architecture and Fine Art and the Department of
Mathematical Sciences at NTNU.
Professor Knut Einar Larsen, the initiator and organizer of the seminar, opened the
event. He emphasized the open-mindedness with which Staale Sinding-Larsen has always
approached his own profession, art history, and how his curiosity had introduced new
ways of understanding and assessing architecture, ways which still today influence the
way architectural history is taught at NTNU.
The programme of the seminar was organized in two sections. The first section,
Historical Reflections, presented research about how use of mathematical thought and
methods (and particularly geometry) has influenced architectural design in the past. The
second section, The Contemporary Scene, focused on how mathematics is used in
contemporary architectural design, especially regarding the possibilities made available by
computer design.
Sylvie Duvernoy opened the first session, speaking on the topic Roman Architecture
and Greek Mathematics – a Case Study: Pompeii’s Amphitheatre.
Sylvie Duvernoy trained as an architect in Paris, and earned a doctorate from the
University of Florence in 1998. She is professor of architectural drawing at the University
of Ferrara.

Nexus Network Journal 12 (2010) 357–360 Nexus Network Journal – Vol.12, No. 2, 2010 357
DOI 10.1007/s00004-010-0036-2; published online 26 June 2010
© 2010 Kim Williams Books, Turin
 Participants at the seminar to celebrate Professor emeritus Staale Sinding-Larsen’s
80th birthday: 1-Staale Sinding-Larsen; 2-Knut Einar Larsen; 3-Tore Haugen; 4-
Finn Hakonsen; 5-Sylvie Duvernoy ; 6- Fabian Scheurer; 7-Georg Glaeser; 8-Dag
Nilsen; 9-Sverre Smalø. (photos by Georg Glaeser)
Duvernoy presented a case study on the amphitheatre in Pompeii. The study is a part
of her comprehensive research of Roman amphitheatres, aimed at revealing the use of
geometric diagrams in their designs through measured surveys and mathematical analysis
of their curves. Around the beginning of the second century B.C the amphitheatre

358 Eir Grytli – Architecture and Mathematics. A seminar to celebrate Professor emeritus Staale...
appeared as an architectural novelty in Roman culture, and was characterized by a closed
elliptic shape that had never been previously adopted in architectural design. The period
of time corresponds to the culmination of the Golden Age of Greek mathematics. It may
be hypothesised that the necessity of designing a new building type provided theoretical
mathematics with a successful field of direct and immediate experimentation. In her
presentation, Duvernoy also called attention to how beauty is a related concept in
architecture and in mathematics/geometry.
Dag Nilsen presented research entitled A Mathematical Game. Studying ratios of
measures as a possible method in building archaeology? Dag Nilsen is trained as an
architect, educated at NTNU (former NTH). Nilsen is associate professor at the Faculty
of Architecture and Fine Art, NTNU, where he teaches architectural history and
conservation.
When trying to reveal the history of a building, written sources are rarely sufficient to
determine either its form and outline in earlier stages, or the intention of the builders.
Nilsen maintains that through examining the building itself, it may be possible to
develop a method to interpret specific questions encountered in building archaeology. By
analyzing ratios of measures in two Norwegian medieval churches, it appears to be
possible to explain previously unsolved problems of their design. In extensive studies of
vernacular Norwegian architecture, Nilsen has also found indications that
mathematical/geometrical tools have been used for design of building types not normally
associated with professional designers.
Finn Hakonsen gave a talk on the subject Ornament and Geometry – The case of
Tuse village church in Denmark. Finn Hakonsen is an architect trained at the Royal
Academy of Fine Art in Copenhagen. His teaching and his writing are mainly based on
the theories of the Tectonic Culture in Architecture.
The Tuse church at northern Seeland was built around the year 1200, in the
Romanesque period of the country, and is one of more than a thousand medieval stone
village churches in Denmark. The study presented is based on theories of architect
Mogens Koch and its point of departure was the typical consecration crosses (hjulkors)
found in the church, which appear as circular ornaments, usually associated with religious
symbolism. The presentation launched the theory that these ornaments also embody a
proportional system of the building. The study reveals correlations between the ornament
and the proportions of the built structure. This leads to the question of whether the
ornament, in addition to carrying religious meaning, also is a carrier of information for
the aesthetical and structural design, a recipe for the entire building process.
Georg Glaeser opened the afternoon session with a talk entitled The Infinite Variety
of Curved Surfaces. Georg Glaeser is a professor in mathematics and geometry at the
University of Applied Arts in Vienna. He is also the author of several books about
computer-based geometry. In the lecture, he presented the geometric programming
system “Open Geometry”. With the help of a C++ compiler, the user is able to create
images and animations with arbitrary geometric context. The system provides a large
library that makes it possible to carry out virtually any geometric task. It provides a
pedagogical approach to understanding how curved surfaces are described as a function
of mathematical formulas. His lecture was a fascinating demonstration of the possibilities
of simulating curved surfaces using computer-aided programming tools, as well as
illustrating how mathematics can also have an aesthetical dimension. The demonstration
did not focus specifically on architectural structures, but showed the possibilities of

Nexus Network Journal – Vol.12, No. 2, 2010 359

modelling curved forms by mathematical formulas in a digital programming system
which can be utilized in computer-aided architectural design.
Fabian Scheurer presented The realization of the impossible: Geometrical Challenges
in Contemporary architectural Concepts. Fabian Scheurer is a computer scientist and
partner of the company Designtoproduction in Zürich.
Today, mathematical and geometrical knowledge is of fundamental importance for
the design and realization of complex, curved forms which are increasingly used in
contemporary architecture design. Computer-aided technology has expanded the
possibilities for architectural design and especially through the technology providing
direct transfer from the design phase to production of components and elements, offering
solutions for irregular shapes that otherwise could never have been realized. In his lecture,
Scheurer also called attention to the challenges comprised between the seemingly infinite
design possibilities and the complexity of the production and construction process, and
how important it is that the computer tool be understood by the architects using it.
Finally, the central person of the day, Staale Sinding-Larsen, shared his Meta-
perspective on architecture and mathematics, based on his view of an object; he dealt not
with an object in itself, but rather what can be done with it. This approach to
understanding objects can be referred back to the object-oriented paradigm in the digital
world: an object consists of a set of attributes and methods. Methods are groups of
instructions with reference to the attributes. A model for understanding mathematical
influence on architectural design based on this theory will be presented in his
forthcoming publication, Borromini’s Spire.
Due to the busy end-of term period, only a limited number of students of
architecture were able to dedicate the whole day to the seminar. This was a pity, because
knowledge about how mathematics has influenced architecture as a design tool – and
continues to do so increasingly – is vital for both a historical understanding of the built
environment and for future design processes. In his lecture, Fabian Scheurer emphasized
the need for architects to understand digital tools in order to obtain optimal benefits in
complex design tasks; such an understanding requires a deeper knowledge of
mathematics than is normally taught in architecture schools.
Personally, not being an expert myself on mathematics in architecture, I found the
seminar most interesting and relevant. The deeper understanding of how mathematics
and geometry has been a creative force for the design of buildings through all times
provides a language for “reading” historic built structures which adds to the historical
knowledge about the creation of buildings. Maybe it also provides a better understanding
of why contemporary changes and additions to historic buildings often look alien to the
original, if the architect who planned the addition did not know the mathematical
language used in the original building.
About the reviewer
Eir Grytli is an architect, with a Ph.D in vernacular development history (1993). She is Professor at
the Faculty of Architecture and Fine Art, Norwegian University of Science and Technology
(NTNU), Trondheim, Norway, where she teaches architectural history and building conservation.

360 Eir Grytli – Architecture and Mathematics. A seminar to celebrate Professor emeritus Staale...
 Tomás Erratum
García-Salgado Erratum to:
National Autonomous University of
Mexico The Sunlight Effect of the Kukulcán
Palacio de Versalles 200
Col. Lomas Reforma MÉXICO
Pyramid or The History of a Line
D. F. C.P. 11930 Erratum to: Nexus Network Journal 12 (2010) 113-130
[email protected]
DOI 10.1007/s00004-010-0019-3
The present erratum corrects an error in fig. 12, p. 124 of
“The Sunlight Effect of the Kukulcán Pyramid or The
History of a Line”.

Original fig. 12

Corrected fig. 12 (Summer and Winter Soltice sunrise positions inverted)

The online version of the original article can be found under doi:10.1007/s00004-010-0019-3.

Nexus Network Journal 12 (2010) 361 Nexus Network Journal – Vol.12, No. 2, 2010 361
DOI 10.1007/s00004-010-0037-1; published online 6 July 2010
© 2010 Kim Williams Books, Turin
NEXUS NETWORK JOURNAL Architecture and Mathematics

Copyright Subscription information

Submission of a manuscript implies: that the work described has not ISSN print edition 1590-5896
been published before (except in form of an abstract or as part of a pub- ISSN electronic edition 1522-4600
lished lecture, review or thesis); that it is not under consideration for Subscription rates
publication elsewhere; that its publication has been approved by all For information on subscription rates please contact:
co-authors, if any, as well as – tacitly or explicitly – by the responsible Springer Customer Service Center GmbH
authorities at the institution where the work was carried out. The Americas (North, South, Central America
The author warrants that his/her contribution is original and that and the Caribbean)
he/she has full power to make this grant. The author signs for and [email protected]
accepts responsibility for releasing this material on behalf of any and Outside the Americas: [email protected]
all co-authors. Transfer of copyright to Springer (respective to owner
if other than Springer) becomes effective if and when the article is Orders and inquiries
accepted for publication. After submission of the Copyright Transfer The Americas (North, South, Central America
Statement signed by the corresponding author, changes of author- and the Caribbean)
ship or in the order of the authors listed will not be accepted by Springer Journal Fulfillment
Springer. P.O. Box 2485, Secaucus, NJ 07096-2485, USA
The copyright covers the exclusive right and license (for U.S. Tel.: 800-SPRINGER (777-4643), Tel.: +1-201-348-4033
government employees: to the extent transferable) to reproduce, (outside US and Canada), Fax: +1-201-348-4505
publish, distribute and archive the article in all forms and media of e-mail: [email protected]
expression now known or developed in the future, including reprints, Outside the Americas
translations, photographic reproductions, microform, electronic form via a bookseller or
(offline, online) or any other reproductions of similar nature. Springer Customer Service Center GmbH
All articles published in this journal are protected by copyright, Haberstrasse 7, 69126 Heidelberg, Germany
which covers the exclusive rights to reproduce and distribute the article Tel.: +49-6221-345-4304, Fax: +49-6221-345-4229
(e.g., as offprints), as well as all translation rights. No material published e-mail: [email protected]
in this journal may be reproduced photographically or stored on micro- Business hours: Monday to Friday
film, in electronic data bases, video disks, etc., without first obtaining 8 a.m. to 8 p.m. local time and on German public holidays
written permission from the publisher (respective the copyright owner
if other than Springer). The use of general descriptive names, trade Cancellations must be received by September 30 to take effect at
names, trademarks, etc., in this publication, even if not specifically the end of the same year.
identified, does not imply that these names are not protected by the Changes of address: Allow six weeks for all changes to
relevant laws and regulations. become effective. All communications should include both old and
An author may self-archive an author-created version of his/her new addresses (with postal codes) and should be accompanied by a
article on his/her own website. He/she may also deposit this version mailing label from a recent issue.
on his/her institution’s and funder’s (funder’s designated) reposi- According to § 4 Sect. 3 of the German Postal Services Data Pro-
tory at the funder’s request or as a result of a legal obligation, tection Regulations, if a subscriber’s address changes, the German
including his/her final version, provided it is not made publicly Post Office can inform the publisher of the new address even if the
available until after 12 months of official publication. He/she may subscriber has not submitted a formal application for mail to be
not use the publisher’s PDF version which is posted on www.spring- forwarded. Subscribers not in agreement with this procedure may
erlink.com for the purpose of self-archiving or deposit. Furthermore, send a written complaint to Customer Service Journals,within 14
the author may only post his/her version provided acknowledge- days of publication of this issue.
ment is given to the original source of publication and a link is in- Back volumes: Prices are available on request.
serted to the published article on Springer’s website. The link must Microform editions are available from: ProQuest. Further informa-
be accompanied by the following text: “The original publication is tion available at: http://www.il.proquest.com/umi/
available at www.springerlink.com”. Electronic edition
The author is requested to use the appropriate DOI for the article (go An electronic edition of this journal is available at springerlink.com
to the Linking Options in the article, then to OpenURL and use the link Advertising
with the DOI). Articles disseminated via ww.springerlink.com are in- Ms Raina Chandler
dexed, abstracted and referenced by many abstracting and information Springer, Tiergartenstraße 17
services, bibliographic networks, subscription agencies, library net- 69121 Heidelberg, Germany
works, and consortia. Tel.: +49-62 21-4 87 8443
While the advice and information in this journal is believed to be true Fax: +49-62 21-4 87 68443
and accurate at the date of its publication, neither the authors, the springer.com/wikom
editors, nor the publisher can accept any legal responsibility for any e-mail: [email protected]
errors or omissions that may be made. The publisher makes no warranty,
express or implied, with respect to the material contained herein. Instructions for authors
Special regulations for photocopies in the USA: Photocopies may be Instructions for authors can now be found on the journal’s
made for personal or in-house use beyond the limitations stipulated website: birkhauser-science.com/NNJ
under Section107 or 108 of U.S. Copyright Law, provided a fee is paid. Production
All fees should be paid to the Copyright Clearance Center, Inc.,222 Springer, Petra Meyer-vom Hagen
Rosewood Drive, Danvers, MA 01923, USA, Tel.:+1-978-7508400, Journal Production, Postfach 105280,
Fax:+1-978-6468600, http://www.copyright.com, stating the ISSN of 69042 Heidelberg, Germany
the journal, the volume, and the first and last page numbers of each Fax: +49-6221-487 68239
article copied. The copyright owner’s consent does not include copy- e-mail: [email protected]
ing for general distribution, promotion, new works, or resale. In these Typesetter: Scientific Publishing Services (Pvt.) Limited,
cases, specific written permission must first be obtained from the Chennai, India
publisher. Printers: Krips, Meppel, The Netherlands
The Canada Institute for Scientific and Technical Information Printed on acid-free paper
(CISTI) provides a comprehensive, world-wide document delivery Springer is a part of
service for all Springer journals. For more information, or to place an Springer Science+Business Media
order for a copyrightcleared Springer document, please contact springer.com
Client Assistant, Document Delivery, CISTI, Ottawa K1A OS2, Ownership and Copyright
Canada (Tel.: +1-613- 9939251; Fax: +1-613-9528243; e-mail: © Kim Williams Books 2010
[email protected]). Printed in The Netherlands