On Dividing the Sky

Preliminary version!
Final version in preparation!
Rüdiger Plantiko

31st July 2004

Abstract

An abundant amount of astrological house systems has been developed through

the times. By studying the evolution of house systems, it will be shown that there
is an underlying guiding principle for the construction of houses as well as for the
so-called ”primary” directions, that has been approached with different accuracy.
The notion of ”mundane position” and of ”spherical house systems” will help
classify and understand the different ways of dividing the Primum Mobile that
have been proposed by astrologers.

1
CONTENTS 2

Contents
1 Introduction 3

2 Primitive first concepts 4

2.1 House position ’by counting’ . . . . . . . . . . . . . . . . . . . . . . 4
2.2 The equal houses . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3 Ptolemy 6
3.1 Houses and Domification . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Definition of Mundane Position . . . . . . . . . . . . . . . . . . . . 7
3.3 Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4 After Ptolemy 17
4.1 Division of ecliptical longitudes . . . . . . . . . . . . . . . . . . . . 17
4.2 The Equator as reference . . . . . . . . . . . . . . . . . . . . . . . 18

5 Spherical House Systems 19

5.1 The projection method . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.2 Campanus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.3 Haly Abenragel . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.4 Regiomontanus . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.5 Placidus de Titis . . . . . . . . . . . . . . . . . . . . . . . . . . 27

6 After Placidus 30
6.1 Kuehr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.2 Koch Houses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

7 Discussion 33
1 INTRODUCTION 3

1 Introduction
The introduction of the Ascendant, the rising point of the Ecliptic, into individual
astrology 1 marks the starting point of many attempts to integrate the motus
primi mobilis, the daily motion of the sky around the observer into the building
of Astrology.
This became particularly important for two parts of the subject:

• The Ascendant was taken as the starting point of a new division of the Eclip-
tic, which, after an intermediate period working with eight sections, 2 very
soon arrived at a system of twelve houses, which were thought in analogy to
the twelve signs of the zodiac. The traditional zodiacal position of a planet
was complemented by the house position, i.e. the number (I,II,. . . XII) of the
house that it occupied. Dividing the sky or at least the Ecliptic into twelve
parts, taking into account in some way or another the daily motion of the
sky, is called the domification.
• In the theory of directions, a part of the Horoscope is taken to the mundane
position (or house position) of another part by the daily motion of the sphere.
For example, if a planet is rising at birth time, then a second planet or aspect
point will form a contact to the former when it reaches the horizon. The angle
of rotation, the so-called directional arc, to make this coincidence happen, is
translated into lifetime using a proportionality factor which traditionally was
simply the equation of one degree of arc with one year of the native’s life.

The notion of mundane position may be regarded as a quantification of the position

of a celestial body in relation to its daily motion. When such a quantification is
given,

• the domification will be achieved by computing the Ecliptic points corre-

sponding to certain fixed values of mundane position (i.e. having integer
multiples of 30◦ as mundane position values);
• the direction problem is solved by computing the mundane position of the
first planet, and then determining the angle by which the sphere has to be
rotated until the second planet achieves the mundane position of the first
planet.

This paper is a continuation of [25] and [26]. The aim is to give a formal description
of the process of domification and to derive some consequences. By using nowadays
mathematical notations, the idea of house division and its historical manifestations
can be worked out more clearly than it was done in the mentioned predecessors.
As sources for the history of house systems, I am using mainly [12] and [13], see
also [7], [2] and [10].
I owe special thanks to Dieter Koch for many valuable discussions.
1
see for example [13], pp. 11, for an outline of this development.
2
the oktatopos, for a first written reference see [20], II.864-970
2 PRIMITIVE FIRST CONCEPTS 4

2 Primitive first concepts

Since there is some understanding of spherical trigonometry and astronomy re-
quired to understand the problem of mundane position, the first concepts of mun-
dane position were of quite an ad hoc nature.

2.1 House position ’by counting’

One of the first house systems – if it is allowed to call it like this – is the method of
reckoning signs: The first house simply is the sign of the Ascendant. The second
house is the subsequent sign, and so on. The ”cusps” of this very primitive method
are therefore the beginnings of the zodiacal signs. Neither Ascendant nor MC are
house cusps in this system. While the Ascendant at least always is located in the
first house, the MC may be in house IX, X or even outside of these. Instead of
being house cusps, Ascendant and MC play the role of additional ’sensitive points’
– if they are taken into respect at all in the Horoscope.
According to Knappich and Gil Brand, this method originates in the early
graeco-egyptian (’Hermetic’) astrology scene which was not interested in such so-
phisticated details as tables of ascensions; instead, they needed easy and handy
algorithms to do their predictions. It had been established by Bouch é-Leclerq
in his Astrologie greque with many examples[3] that the counted houses and the
counted ’lots’ (like the Lot of Fortune, but computed on the level of zodiac sign)
were the methods of choice for these astrologers.

Figure 1: Horoscope from P. Oxy. 235 (about 20 CE).

The Horoscope figure from the Oxyrhynchus Papyrus (redrawn by Neugebauer in

2 PRIMITIVE FIRST CONCEPTS 5

[22], p.18) dates to about 20 C.E. It shows a circular Horoscope diagram, arranged
like a modern Horoscope scheme with the ascending sign Taurus to the left, the
signs in counterclockwise order, so that the lower half corresponds to the invisible
part of the Ecliptic. This figures illustrates the precision of average Horoscopes of
that time – up to one sign – which necessarily implies that aspects and houses can
only be identified on the level of zodiac sign. With other words: the method of
counting houses has its roots in the lack of computational precision.
Gil Brand quotes the carmen astrologicum of Dorotheus Sidonius (first
century C.E.) with a random example, 3 a Horoscope with Ascendant in 18◦ l and
A in 6◦ 500 l, considered as Hyleg, although above the horizon, which can only be
explained by the usage of counted houses, equating the complete zodiac sign l
with the first house.
Later, the method is described in a classical textbook of Vedic astrology due to
Varaha Mihira and referred to as Rasi Chakra. 4 Interesting enough, this sim-
ple method survived through the Indian and Arabian times of astrology. Even
nowadays, Vedic astrologers are working with those ’counted houses’ ([12], p.20).
For the Arabian astrologers, note that an early representant of that epoch as
Messahalah uses the counted houses simultaneously with another system, like in
the following quotation of his Astrological History (which is used as an arbitrary
example for many similar places):

... and Jupiter and Mars in the eleventh [house], they will be by division
in the tenth...5

g
4° 12°39’ 19° 27°17’ 29°
MC K XI eq. F XI div.

Figure 2: Ecliptical house divisions (example).

The example refers to a Horoscope with Ascendant in 19 ◦ i, K in 12◦ g and F

in 27◦ g. Since g is the eleventh sign, counted from the sign of the Ascendant
i, K and F are therefore in XI ”by counting”. The method ”by division” on the
other hand cannot be an equal house system, since then K would be in X and F
in XI (the house cusp for XI would be at 19 ◦ g). Since for the latitude of Bagdad,
the above Ascendant results in a Medium Coeli of about 4 ◦ g and a cusp XI of
3
[4], p. 212. Mind that the original version of the carmen astrologicum is lost, however.
4
[21], quoted in [12], p. 20
5
[8], p.40
3 PTOLEMY 6

about 29◦ g, the ’division method’ could be the Eutokios method (see section
4.1 below).

2.2 The equal houses

The equal houses start with the Ascendant which is determined as usual as inter-
section point of the Horizon with the Ecliptic. The cusps of the subsequent houses
are obtained by adding integer multiples of 30 ◦ to the longitude of the Ascendant.
Compared to the counted houses, the equal houses mark a slight progress, since
they take into account the actual longitude of the rising point of the Ecliptic. The
Horoscopos is taken more seriously, and the insight is growing that it should be
integrated as astronomical point into the house construction.
Althouh the Ascendant in this system forms the cusp of the first house, the Medium
Coeli, is not a house cusp. It remains a sensitive point, usually positioned some-
where in the ninth or tenth house.

3 Ptolemy
3.1 Houses and Domification
Although the houses are present at many places of the ”astrologers’ Bible”, the
Tetrabiblos contains no details about the domification. Ptolemy assumes the
readers’ knowledge of how to compute the houses, as well as some basics about
their classification (e.g. the meaning of angular, intermediate and cadent houses).
The astrological aphorisms of the Tetrabiblos are working mainly with planets,
signs and fixed stars. Only marginally, the meaning of particular houses are taken
into consideration. E.g. the connection of the fifth house with children, or of the
sixth and twelfth house with servants are mentioned. But there is no systematical
explanation of the meanings of all the houses.
From the beginning of section III.10 Of Length of Life it has been induced that
Ptolemy used the equal system of houses: In order to define the aphetical places,
he divides the Ecliptic into 12 houses of 30 ◦ , the first house beginning 5◦ before
the Ascendant and ending 25◦ after it. From this passage6 , it seems that the tenth
house ( 
  ) is nothing else than the part of the Ecliptic in right square
aspect (= −90◦ ) to the first house.
On the other hand, in this same chapter he works at a later place with the Medium
Coeli as intersection of the Meridian with the Ecliptic. With no doubt, Ptolemy
knew that this point does not always form a 90 ◦ angle on the Ecliptic to the
Ascendant. Since he uses the same word 
  for the Medium Coeli, there
is a contradiction between these two statements.
6
[28], p.273
3 PTOLEMY 7

Later astrologers7 have tried to harmonize these statements by introducing the

so-called mundane aspects. But it seems more plausible to me that Ptolemy re-
produces two different techniques in parallel, taken from two different traditions.
Each tradition may have had its own area of validity: The ”hermetic” house divi-
sion may have been the right method for determining the aphetical places, while
the astronomically correct Medium Coeli had been used for the computation of
directions.
To let the first house start 5◦ before the Ascendant, is a speciality of the Ptole-
meic variant of the equal house system that had some influence on later domi-
fication methods; it was probably led from the idea that the Ascendant has an
influence by aspect (due to its orb) in both directions. This Ptolemeic shift of
5 degrees has been taken over into other house systems of late antiquity (Por-
phyrios and Rhetorios, see sections 4.1 and 4.2 in this paper).

3.2 Definition of Mundane Position

In Tetrabiblos, section III.10 Of Length of Life, Ptolemy gives a definition of what
we would call the locus of mundane position, and in the sequence he exemplifies
his method of computing directions with three examples.

    
  !#" $%$!
'&)(*
+ ,
, -/.  0 1 
2 -
 3

For a place is similar and the same if it has the same position in the
same direction with reference both to the horizon and the meridian. 8

The points on the visible part of the Meridian are therefore all ”similar to each
others” – they are all in culmination, and ”similar” to each others are the points
on the eastern semicircle of the Horizon – they are all rising. But what about the
points between rising and culmination? How can their ”intermediateness” between
these two positions be quantified? How can we compare two different points with
relation to the angles?
In this chapter, we will work out a formal framework for candidates of ”systems
of mundane position”, claiming to be able to answer this question.
First of all, we want to restrict the problem to those points of the sphere that,
additional to culmination, are subject to rising and setting at the given location. 9
We therefore disregard the circumpolar regions:

Definition I As region of definition D Φ for systems of mundane position, we

define the spherical layer resulting from the celestial sphere by excising the two
7
e.g. Placidus, see[11], Thesis 12, p. 256
8
[28], III.10, p.291
9
This restriction is not arbitrary, but originates in the nature of the problem: It doesn’t make sense
to express the ”intermediateness” between rising and culmination, for a point which doesn’t rise at all.
3 PTOLEMY 8

N
Meridian

Horizon

Equator

Figure 3: The domain DΦ

circumpolar caps. If S denotes the sphere, the points being specified by their coor-
dinates α for right ascension and δ for declination, and Φ the geographical latitude
of the location, we thus define

DΦ := {(α, δ) ∈ S : |δ| < 90◦ − |Φ|}

We can think of DΦ as the maximal region of the sphere for which it makes sense
to speak of a mundane position, because the points of this region are rising, cul-
minating and setting.
Figure 4 demonstrates our convention to measure mundane position which is led
by the analogy of houses to zodiac signs (I = a, II = b etc.). The Ascendant
is therefore in analogy with the Vernal Point, and the houses are counting in the
same direction like the signs. The mundane position for Ascendant and Descendant
has to be 0◦ and 180◦ , respectively. MC and IC get the values 270 ◦ and 90◦ ,
respectively. From the beginning of the *
  up to the contemporary
astrology, the analogy between signs and houses is one of the essential analogies
for interpreting houses: The quality of a house can be thought of as a ”material”
or ”mundane” specializations of the traits of the corresponding zodiac sign.
We are now able to give a minimal definition for the notion of ”concept of mundane
position”:

Definition II Let t be the sidereal time of birth and Φ the geographical latitude
of the birth place. A concept of mundane position is a family of maps µ Φ;t : DΦ →
[0◦ , 360◦ [, such that
3 PTOLEMY 9

270°
300° 240°

330° X IX 210°
XI VIII
XII VII
0° 180°
I VI
II V
30° III IV 150°

60° 120°
90°

Figure 4: Mundane Position

1. The functions µΦ;t , considered as maps to the unit circle, are continuous. 10
2. The angular semicircles are mapped to multiples of 90 ◦ :
µ−1 ◦
Φ;t (0 ) = Eastern semicircle of the Horizon
µ−1 ◦
Φ;t (90 ) = Invisible points of DΦ in Meridian transit
µ−1 ◦
Φ;t (180 ) = Western semicircle of the Horizon
µ−1 ◦
Φ;t (270 ) = Visible points of DΦ in Meridian transit

3. For each parallel circle contained completely in D Φ;t , i.e. for fixed declination
δ (with |δ| < 90◦ −|Φ|) the mundane position µ(α, δ) is a strictly monotonous
function of the meridional distance M D := α − t reduced to the interval
]0, 360].
4. The function is equivariant with respect to translations of t and α, meaning
more specifically that
µΦ;t+τ (α, δ) = µΦ;t (α + τ, δ)
for all τ . This implies that the mundane position in fact depends only on the
meridional distance α − t rather than on α and t independently.

In the course of this paper, we will surpress the dependence of t in the notations.
This is justified by the translational equivariance (condition II.4). We consider the
sidereal time t arbitrary but fixed.
10
This may be formulated more formally by postulating the continuity of the composition
(cos µΦ;t , sin µΦ;t ).
3 PTOLEMY 10

Condition II.3 ensures that the mundane position is decreasing strictly with in-
creasing sidereal time.
The definition is minimal, since there still is an infinite quantity of such concepts
of mundane position. It only covers its formal requirements. The requirement
that such a system should be derived in a conceivable manner from the spherical
geometry of the rotating celestial sphere, can itself not be formalized!
If a house system really reflects the daily motion of the celestial sphere, then it
should be derivable from an underlying concept of mundane position. This leads
us to the following

Definition III A house system will be called spherical if the corresponding domifi-
cation can be extended to the sphere, i.e. if it is derived from a concept of mundane
position.

As simple and minimal as they are – the formal requirements of the definition II
already lead to some strong consequences:

Consequence 1 From the definition of a concept of mundane position, it follows:

A parallel circle of fixed declination contained in D Φ is mapped bijectively to the
interval [0; 360[.

This is a standard property of strictly monotonous functions (II.3), combined with

the continuity as map to the circle (II.1).

Consequence 2 A concept of mundane position can neither be based on the eclip-

tical longitude nor on the right ascension - more precisely:

1. The mundane position µΦ cannot be a function of the ecliptical longitude

alone.
2. The mundane position µΦ cannot be a function of the right ascension alone.
3. The mundane position µΦ cannot be a function of the oblique ascension with
any fixed polar elevation ϕ alone.

Indeed: there are points with different ecliptical longitudes on the part of the
Meridian contained in DΦ . If mundane position depended on ecliptical longitude
alone, µ−1 (90◦ ) would be more than the meridional semicircle, as was required in
condition II.2: This semicircle should then have to be joined by the semicircles
of points having those longitudes. The same argument, applied to the Equator
with the Horizon instead of the Meridian, excludes a dependency of µ Φ of the
right ascension alone. Similarly, the dependency of oblique ascension alone can be
excluded.
This consequence has been written out in this detail, since it rules out many histor-
ical attempts of domification as non-spherical, insufficient for building a concept
of mundane position from it. They don’t solve the task of dividing the sky. Of
3 PTOLEMY 11

course, the primitive methods mentioned in section 2 are of this kind, since they are
working with the ecliptical longitude alone. But also a slightly more sophisticated
method like that of Porphyrios is insufficient for this purpose. Later methods
like Rhetorios, Alcabitius (based on right ascensions) have to be excluded as
well as the modern method of Koch based on oblique ascensions of fixed polar
elevation.11
The definition II implies that the set D Φ of celestial points subject to rising and
culmination can be decomposed as a disjoint union of the sets µ −1 Φ (γ) with γ ∈
[0◦ , 360◦ [. These sets will get a special name:

Definition IV The set µ−1 Φ (γ) is called the position curve for the mundane posi-
tion γ. It is the set of all points of the sphere having mundane position γ.

With these sets, we finaly arrived at the Ptolemy quotation from the beginning
of this section: The position curve is nothing else than the set of points having a
”similar position” in the sphere with respect to Meridian and Horizon.
Since the four special position curves mentioned in the definition II are all semicir-
cles, it is a natural question to ask for mundane position concepts based completely
on semicircles, meaning that all position curves are semicircles. What can be said
about such systems?

Consequence 3 A position curve that is a semicircle will be called position circle.

A position circle must necessarily end in the intersection points of Horizon and
Meridian, the North point and South point of the Horizon.

Indeed – if this wouldn’t be the case, the position circle would intersect the Horizon
in a point inside of DΦ which already would force it to coincide with the Horizon,
according to condition II.2.
The position circles have been introduced into the astrological calculations by Al
Biruni. Later, they were used for the house systems of Haly Abenragel and
Regiomontanus.
It is interesting to notice that a concept of mundane position based on position
circles can easily be extended from D Φ to the complete sphere: Any point, except
South and North point of the Horizon (the axes of the position circles) have a
unique mundane position, since they lie on a unique position circle. It is not
clear whether this can be valuated as an advantage of these systems: Since the
circumpolar points do not rise nor set, assigning a mundane position to them is a
meaningless operation.

3.3 Directions
Let’s come back to the chapter III.10 On the Length of Life of Ptolemy. It
gives a complicated algorithm to determine the time of death of a native. For our
11
However, there is a modification of the Alcabitius system ascribed to Abenragel, which is
spherical (see section 5.3).
3 PTOLEMY 12

purposes, this chapter is particularly important because it is the only place in the
whole book where the method of directions is explained and used. Roughly, the
algorithm goes as follows:
In the first place, the astrologer has to determine the prorogator ( "

-
, ,
called Hyleg in the later Arabic literature), which is usually the Sun by day and the
moon by night, if they are not placed in ’bad houses’, i.e. the eighth and twelfth
house. In that case, there is a list of alternative choices for the prorogator, ending
in the    , the ascending point of the Ecliptic.
When the prorogator has been distinguished, there are two different methods of
determination of lifetime, the horimaea (  -
   ) and the projection of rays or
actinobolia ( "--
2 -  ). The latter is the only in which we are interested. It
applies if the prorogator is located in quadrant IV: Above the horizon, but eastern
of the meridian.
In this case, the motion of the maleficient planets F and L and their aspects
due to the motus primi mobilis is considered. The first of these points reaching
the mundane position which the prorogator had at the time of the birth will kill
the native. More precisely, the time when this happens is measured in equatorial
degrees, and the number of degrees equals to the number of years which the stars
allot to the native.12
With the notions about mundane position concepts of the preceding section, we
may define in accordance with Ptolemy:

Definition V Directing a point of the sphere, the Promissor to another point of

the sphere, the Significator means to rotate the sphere around its polar axis until
the Promissor aquires the mundane position of the Significator. The rotation angle
to make this happen is called the directional arc.

Already Regiomontanus extracted this definition from the text: 13

Dirigere non est aliud quam movere spheram

donec locus secundus traducitur ad situm primi.

Directing is nothing else than rotating the sphere until the second place
is transferred to the position of the first.

Figure 5 demonstrates the procedure: It shows a point S of the Horoscope located

on its mundane position curve (the bold line). Another point P may have a different
mundane position – but, by the rotation of the sphere, it will eventually reach the
mundane position of S. As the figure shows, the Promissor will in general not meet
the place of the Significator itself, since it may have a different declination. The
problem therefore requires a concept of mundane position: Using such a concept, it
12
Of course, there are some exceptions of this rule. For example, the beneficient planets Venus or
Jupiter could interfer and modify the expected lifetime. But we are not interested in the details, since
we want to understand the method of directions.
13
[30], problem 25
3 PTOLEMY 13

P
P’

α
α’

Figure 5: Directions

is clear from requirement II.3 that there is exactly one point of intersection of the
Significator’s position curve with the Promissor’s parallel circle. The directional
arc ∆ will be the difference in right ascension between the Promissor and this
intersection point:
∆ = α − α0

The question now is: What was the concept of mundane positions that Ptolemy
used? In Tetrabiblos III.10, Ptolemy describes the concept in words, and addi-
tionally provides a computational example for it. Although only one of the exam-
ples is of worth for our question, we can be lucky to have it, since it finishes some
inappropriate interpretations of the text brought up by Regiomontanians. 14
Ptolemy observes that it is easy to determine the directional arc if the prorogator
is ascending or culminating: In that case, one just has to determine at what time
the destructing aspect point will ascend or culminate and then to compute the
difference of these two times, measured in equatorial degrees. However, he sees
that it is insufficient to work with ascension or culmination times if the prorogator
is located somewhere between rising and culmination. For the intermediate points,
he defines the notion of mundane position, as quoted and discussed in section 3.2.
The next statement informs us about Ptolemy’s choice:

  & 
   2-$ 2 
   

$ 


 -     
$*
/   
 
'

2 -

! 
-/.  0 
      '%
!

 '& 
 
-
  

 3
This is most nearly true of those [points] which lie upon one of those
semicircles which are described through the intersections of the meridian
14
The misunderstanding is going back to Regiomontanus himself, but has been propagated in the
last century by G. Schwickert, see [31], p. 38 and 123-124.
3 PTOLEMY 14

and the horizon, each of which at the same position makes nearly the
same temporal hour.

Thus, in the first place, we are told which concept of mundane position Ptolemy
prefers: Two planets are in the same mundane position if they make the same
temporal hour. A temporal hour is the twelveth part of the daily arc, i.e. of the
time from rise to set. This means, more precisely, that according to Ptolemy,
two points are similar if the times which passed since the points rised, are the same
when measured as proportional parts of the arc from rising to culmination. This
is a good concept, because the meridian and the horizon are naturally embedded
into it: they define lines of equal mundane position (the proportion makes 1 and
0 respectively in these cases). Let’s call this concept of mundane the position the
temporal concept and let us undertake the effort to spell it out in detail:

Definition VI If

M Dd := α − t
M Dn := α − (t + 180◦ )

denote the meridional distances of day and night, and

SAd/n := 90◦ ± arcsin(tan δ tan Φ)

denote the semi-arc of day or night, the temporal mundane position is defined by
 !

 M Dd



+ 3 · 90◦ , for points above the horizon

 SAd
µ=

 !

 M Dn

 + 1 · 90◦ , for points below the horizon

SAn

The formal requirements on a concept of mundane position listed in definition II

are verified easily.
In the second place, the quotation claims that the position curves belonging to the
temporal system can be approximated by semicircles. It is important to observe
that Ptolemy mentions the position circles only as an approximation for the
temporal concept. He doesn’t want to base directional calculations on position
circles unless as an approximation for the temporal system. This is stressed by the
fact, that he uses the word & 
  (approximately) at two places in this quotation.
This becomes even more obvious by his example following in the text. He directs
the ecliptical point with longitude 60 ◦ (= 0◦ c) to the vernal point (0◦ a) assum-
ing the latter in four different mundane positions: Rising, culminating, setting, and
in an intermediate position. If we want to find out whether Ptolemy worked with
the temporal mundane position or with position circles, we need his computations
of the directional arc when the prorogator is at an intermediate position – since the
position curves of the four angles ASC, DSC, MC and IC coincide for any system
3 PTOLEMY 15

of mundane position.- In his intermediate example, 18 ◦ b is in mid-heaven. His

computation of the directional arc until 0 ◦ c reaches the mundane position of the
vernal point results in a value of 64◦ . Using, as Ptolemy does in his example, the
latitude which yields a maximal daily arc of 14 hours (according to the Almagest,
this is the climate of Lower Egypt, 30 ◦ 220 N15 ) and the value 23◦ 510 2000 for the
eclipical obliquity,16 we obtain an arc of 64.1◦ in the temporal concept and an arc
of 66.7◦ when working with position circles. This proves that Ptolemy used the
temporal mundane positions for his directional calculations.

Meridian

δ
Equator OMD ϕ
a
RAMC AD
OA
MD

Figure 6: Parts related to the mundane position curve

There is a common nomenclature for different geometrical parts which are con-
nected with a concept of mundane position. First of all, by analogy to the horizon,
the position curve – and each point on it – has a polar elevation ϕ, which is the
complement to its angle with the Equator, 17 and an oblique ascension (OA), which
is the distance of the curve from the Vernal point, measured on the Equator. 18 In
general, each point of the sphere has a right ascension (RA). The difference of its
right and oblique ascension is called ascensional difference (AD). Finally, there is,
as we already know, a meridional distance (M D) of a star from its culmination,
and an oblique meridional distance (OM D), defined as distance of the position
curve from the meridian, as measured on the Equator. The mundane position can
be regarded as a function of M D, δ and Φ.
It is interesting to understand the difference between the position circle and the
temporal position curve by formulae. For this, it is convenient to keep the OM D
15
[29], p.95
16
[29], p.78
17
For the Horizon, the polar elevation coincides with the geographic latitude and is indeed the elevation
(i.e. the altitude) of the celestial pole above the Horizon.
18
This quantity is called ascension by pure analogy to the horizon. Of course, nothing is ascending
(rising) on an intermediate mundane position curve.
3 PTOLEMY 16

Meridian

δ
OMD
a
AD′ Equator
AD

Figure 7: Comparing mundane Positions

arbitrary but fixed and to look at the differences of the AD values that result for
different declinations on the two position curves having that fixed OMD (see figure
7).
Let’s assume for simplicity that the mundane position in question is somewhere
between rising and culmination, i.e. 0 ◦ ≤ M D ≤ 90◦ . We denote with ADp and
ADt the two ascensional differences derived from the position circle concept and
from the temporal concept, respectively. Then an easy calculation shows
MD
ADt = · arcsin (tan δ tan Φ) ,
90◦
ADp = arcsin (sin(M D) tan δ tan Φ) .

We see that ADt can be transformed into ADp by applying the following two
operations:

• Commute the factor M D/90◦ with arcsin, and then

• replace the linear function M D/90◦ by sin(M D)

Both operations are approximations: For small arguments, a factor can be com-
muted with arcsin, and the linear function is an interpolation of the sine (on the
interval [0, 90] that we are considering here). This shows that for small values of δ
or Φ and for points near the meridian, the two concepts of mundane position are
very close.
4 AFTER PTOLEMY 17

4 After Ptolemy
4.1 Division of ecliptical longitudes
In section 2, I already exposed the two oldest house systems based on the longitude
of the Ascendant alone. From the beginning, astrology had its focus on the Ecliptic.
If we leave aside specialities like the parallel aspects, the planetary latitudes were
ignored in astrological considerations – thus, not the body itself but its projection
on the Ecliptic is the astrologically relevant object. For a renaissance astrologer
like Pegius, the Ecliptic plays the role of a mediator, he consistently calls it
Scheinbrecherin in his textbook Geburtsstundenbuch [23]. 19
The development went on from the Ascendant to the four Kentra, a division into
four quadrants of 90◦ each, and in the first centuries C.E., astrology arrived at the
equal house system with a span of 30◦ for each house, starting with the Ascendant.
As soon as the ’true’ or ’mathematical’ Medium Coeli and Imum Coeli, the in-
tersections of Meridian and Ecliptic, had been introduced in the Horoscopes, the
question arose how to integrate these new points into the system of houses. It
was clear to the more reasonable astrologers that only one point could be the MC:
either the mathematical MC or the point being in right square aspect to the As-
cendant. If it was the first, then how could the quadrants defined by Ascendant,
Descendant, MC and IC be trisected in a natural way?
The first methods of this kind were based on ecliptical longitudes. Koch and
Knappich ([13], p.73) mention the Horoscope of the emperor Hadrian which
was most likely erected by Antigonos of Nikea in about 160 C.E. This seems
to be the first domification known to us that works with the mathematical M.C.
This same method has also been used later by the mathematician and astrologer
Eutokios of Askalon (born 497 C.E.) whose own Horoscope has been preserved
in various manuscripts.20
The Ptolemeic shift, his postulate that each house should start 5 ◦ before its asso-
ciated cusp (Tetrabiblos III.10, see section 3.1), led to some modifications of the
method:
Pancharios (about 200 CE) had an interesting interpretation of the Ptolemeic
shift: He claimed them to be not a fixed difference in longitude but always the sixth
part of the actual house length that comes out with the division method. This
means: He subtracts 5◦ in mundane position, not in ecliptical longitude. Por-
phyrios (about 250 CE) instead, worked with a fixed 5 ◦ shift in longitude.
It is interesting to observe that this domification method survived until the late
medieval times. The Astrolabium planum by Peter of Abano (copied in the
19
which could be translated to ’reflector of the rays’ – in the sense that all the aspects, the ’planetary
rays’ are mediated through the Ecliptic. This does not prevent him to base his directional computa-
tions on both longitude and latitude, as was done in his reference work, the tabulae directionum
profectionumque of Regiomontanus.
20
[13], p.70
4 AFTER PTOLEMY 18

well-known ”Heidelberger Schicksalsbuch” [1]) contains for each ascending degree

of the Ecliptic an empty Horoscope scheme with the house cusps. The house cusps
are computed by division of the Ecliptic, i.e. according to the Eutokios method.
In all cases, the lines of mundane position would be the semicircles of fixed eclip-
tical longitude, which, as was argued in Consequence 2, cannot be the base for a
spherical house system.

4.2 The Equator as reference

A manuscript by the byzantine astrologer Rhetorios (about 600 CE) How to
find on the degree the beginning of the twelve house cusps has been published by
Cumont in CCAG VIII.1, see also Neugebauer[22], No. L 428. The method de-
scribed marks the important transition from Ecliptic to Equator as more adequate
to the problem of mundane position, since the Equator is the reference plane for
the daily motion of the Sphere.
Rhetorios subtracts the Ptolemeic shift of 5 ◦ from the given Ascendant of his
sample Horoscope. This defines the beginning of his first house. Then he computes
the daily semi-arc of this Ecliptic point and divides it into three parts. Starting
with the RA for house I, he gets the RA for the beginning of houses XII, XI and
X by subtraction of the third parts. RA X, in his system, therefore is not the
true MC of this Horoscope but a calculated MC which belongs to his beginning
of house I. For houses IX, VIII and VII, he continues with the nightly semi-arc of
house I (the complement to 180◦ of the daily semi-arc).
The system itself is clearly of intermediate nature. But it shows that the ecliptical
division was felt insufficient. As an ex post definition of the houses, it does not
really derive the cusps in a constructive way. It only takes the four Kentra as
given, it does not look at the spherical construction or astronomical significance
of the Kentra. The method of Rhetorios uses the temporal hours of a certain
parallel circle instead. From an abstract and astronomically meaningless division
of the Ecliptic, it tries to proceed to a division of ascensional times.
This same method, but with cusp I as Ascendant (i.e. without applying the Ptole-
meic shift of 5◦ to it), the method has been popagated by Alcabitius (about 980
C.E.), and probably it is also identical with the method ascribed to Albategnius
(880-929), although his domification description is somewhat unclear.
Since the method is defined by meridians – it computes the cusps as Ecliptic points
with a given RA – it cannot be extended to the sphere, due to Consequence 2 above.
5 SPHERICAL HOUSE SYSTEMS 19

5 Spherical House Systems

5.1 The projection method
To illustrate some important spherical house systems, we will use the venerable
stereographic projection which has been used for sky maps at least since the time of
Eudoxos of Knidos. In contrast of the usual choice for astrolabes, however, we
will choose a different projection center than the geographic pole. For our purposes,
it is more convenient to project from the northern intersection of Horizon and
Meridian.21 The following picture shows a transversal section of the sphere along
the Meridian which therefore forms the circumference of the section. Orthogonal to
the drawing plane, we have the Prime Vertical (V1), and the Horizon. The points
of the sphere are mapped to the plane Π which is parallel to V1, containing the
South Point of the Horizon. The dashed line mapping the point P to P 0 illustrates
the projection method.

P’
Z
P NP

V1
S N
Horizon

Meridian
SP
Π

Figure 8: Projecting from N

21
Without loss of generality, we assume in the following that the birth place is located in the northern
hemisphere. For southern latitudes, North and South simply switch there roles.
5 SPHERICAL HOUSE SYSTEMS 20

Let’s recall the basic features of the stereographic projection: It preserves angles
(but cannot preserve distances, like any map of the sphere to the plane), and it
maps circles and lines to circles or lines. In our situation, this implies:

• If the plane is suitably normalized, the prime vertical is mapped to the unit
circle, and the South Point of the Horizon (S) is mapped to the origin.
• Circles on the sphere are mapped to circles on the plane, if they do not pass
through N. If they pass through N, they are mapped to straight lines.
• Position circles are circles passing through North and South point of the
Horizon. They therefore correspond to the straight lines passing through the
origin.
• The northern circumpolar region is mapped to an upper halfplane bounded
by a line parallel to the image of the Horizon.
• The southern circumpolar region is mapped to a circle which is symmetric to
the Meridian, tangent to the Horizon at its South Point, and lying completely
below it (i.e. in the invisible part of the sky).

The most important of these properties for our considerations is the identification
of the position circles with the straight lines passing through the origin. This
allows us to compare the methods based on position circles, and to illustrate the
differences of their construction methods.
5 SPHERICAL HOUSE SYSTEMS 21

5.2 Campanus

Z
V1
d
c e

f
b

g H
S
a
h

i
l

j Ecl.
k
Aeq.

Figure 9: Campanus mundane position

Position circles, defined as circles passing through the intersection points of Horizon
and Meridian, may have been used already by the classical spherical geometers
(Autolykos of Pitane, Kleomedes). But they were first introduced into the
astrological problem of houses by Al Biruni (978-1046) in his Book of Instruction
in the Elements of the Art of Astrology (1029), where he uses them to describe
a new method of domification.22 The method is nowadays called the method of
Campanus, since it has been described later by Campanus of Navarra (1239-
1296).
22
see [12], p.27-28.
5 SPHERICAL HOUSE SYSTEMS 22

In this method, the Prime Vertical – the circle passing through Zenit and East point
of the Horizon – is used for measuring the mundane position. The uniquely deter-
mined position circle passing through the point with altitude 30 ◦ on the Prime Ver-
tical is the boundary for house XII. Similarly, altitude 60 ◦ determines the boundary
of house XI, and the Zenit determines the meridian which is the position circle for
house X. Continuing this manner, one obtains a completely symmetric division of
the sphere into twelve equal parts, spherical diangles of angle 30 ◦ .
The associated mundane position can be determined for an arbitrary point of the
sphere as follows: Find the position circle passing through this point, and intersect
it with the Prime Vertical. The distance of that intersection point from the Horizon
(i.e. its altitude) gives the mundane position.
The Campanus method is the only spherical domification method with this high
degree of symmetry.23 The price is, however, that the reference to the motus
primi mobilis is lost almost completely. The Prime Vertical, the base circle for
this division method, is not related to the daily motion at all. This seems to be
the main criticism of Regiomontanus who decided to skate over this method,
”because it is against the minds of the ancients, and futile, because the Prime
Vertical is an imaginary circle, not based on anything with effect.” 24
23
The so-called Azimut houses, however, which divide the sky around the Zenit using the Azimut, also
divide the complete sphere harmonically into twelve segments. They are no true domification method,
however, since the cusp I in this system is not the Ascendant but the East point of the Ecliptic.
24
[30], 14th problem: Modus [...] ille quia alienus sit a mentibus antiquorum et
quia futilis quam circulo verticali imaginario ac nihil virtutis habenti innititur,
silentio pretereundum censemus.
5 SPHERICAL HOUSE SYSTEMS 23

5.3 Haly Abenragel

Z
V1
d
c e

f
b

g H
S
a
h

i
l

j
Ecl.
k
Aeq.

Figure 10: Abenragel mundane position

In his ”Eight Books on Astrology”, Haly Abenragel (about 1050) describes

an extension of the method of Alcabitius: Like Alcabitius, he divides the
quadrants formed by Meridian and Horizon on the parallel circle of the Ascendant
into twelve pieces. But the transfer of the division points to the Ecliptic is afforded
by the uniquely defined position circles passing through the division points.
Although even Regiomontanus mentions Abenragel as using position circles
(in problem 22 of his tabulae directionum), the method of Abenragel has been
forgotten in the course of the time, and was re-invented in our times by the Dutch
”Workcommunity of astrologers”, founded by the Th. J. J. Ram (1884-1961),
Thierens and Leo Knegt ([32],[14]). Ram called it the ”Ascendant-Parallel-
5 SPHERICAL HOUSE SYSTEMS 24

Circle system”. Thanks to this group, the Abenragel house method is still alive
in modern Dutch astrology.
The associated mundane position can be determined for an arbitrary point of
the sphere as follows: Find the position circle passing through this point, and
intersect it with the parallel circle of the Ascendant. The mundane position can
now be determined from the proportion in which the intersection point divides the
quadrant between the next intersection points of the parallel circle with Meridian
and Horizon.
There is a conceptional inconsistency in this method: If a position circle divides the
parallel of the Ascendant in a house proportion, then the second intersection point
of that position circle with the parallel will in general not realize the corresponding
proportion for the opposite house. If one would take the system seriously, then a
consquence would be that intermediate house cusps like II/VIII will not be opposite
to each other! Usually, this inconsistency is ignored: One constructs the cusps XI,
XII, II and III, and takes the oppposition places of these cusps as definition for the
remaining cusps. V, VI, VIII and IX. Figure 10 demonstrates the construction:
The dashed circle is the parallel circle of the Ascendant. The solid dots on it mark
its proportional division. The house boundaries are position circles and therefore
straight lines passing through these division points. It can be seen from the figure
that the opposite intersection points with the parallel do not meet the division
points that should define the house cusp.
5 SPHERICAL HOUSE SYSTEMS 25

5.4 Regiomontanus

Z
V1
d
c e

f
b

g H
S
a
h

i
l

j Ecl.
k
Aeq.

Figure 11: Regiomontanus mundane position

The Jewish astrologer Ibn Esra (born 1090) may have the priority for the domi-
fication that is nowadays called the Regiomontanus method. It was known that
Ibn Esra had worked out a domification method, but this work seems to be lost.
Surely, Regiomontanus contributed much to the success of this house system
by computing and printing tables for it and adding it, among with Campanus
tables, to his tabulae directionum profectionumque. It should be pointed out
that Regiomontanus never claimed the priority for this method. He preferred,
somewhat suggestively, to call this domification the rational method.
Starting with the East Point of the Horizon (which always intersects the Equator),
the Equator is divided into twelve equal sections of 30 ◦ . The position circles
5 SPHERICAL HOUSE SYSTEMS 26

passing through these points are the boundaries of the twelve houses.
The associated mundane position can be determined for an arbitrary point of the
sphere as follows: Find the position circle passing through this point, and intersect
it with the Equator. The distance of that intersection point from the Horizon,
measured on the Equator, gives the mundane position.
Regimontanus was consequent enough to base not only the house system but also
the calculation of directions on position circles. This shows that he had identified
the mundane position as link between house systems and direction methods.
It is interesting to notice that – after the purely geometrical method of Cam-
panus, which is classified by Regiomontanus as ”imaginary”, the Regiomon-
tanus method puts the focus on the Equator as the reference plane of the daily
motion again.
5 SPHERICAL HOUSE SYSTEMS 27

5.5 Placidus de Titis

Z
V1
c
b d

a
e Ecl.

l S H

f
k

j
g

i
h
Aeq.

Figure 12: Placidus mundane position

With the method used by Maginus (1555 - 1617) 25 and explicitly described by
Placidus de Titis (1603 - 1668), we are leaving the realm of position circles.
The house system is based consequently on the division of times. In contrast to
the other methods presented, there is not one singular reference circle on which the
division is performed and then projected to the Ecliptic. Instead, the division is
performed for each point of DΦ according to its own motion through the quadrants.
In contrast to Alcabitius and Abenragel, Placidus does not consider the semi-
arc of the Ascendant as base of the division, but each point on D Φ has its own
25
Maginus published tables of houses [19], but he kept his method secret. The table values clearly
show that he discovered the method of Placidus
5 SPHERICAL HOUSE SYSTEMS 28

parallel circle and therefore its own rising times, its own ”temporal hours”, its own
semi-arcs – depending on the declination of the point. These proper semi-arcs are
used to determine the mundane position.
For directions, this is a well-known technique, as it is identical to the method
exposed by Ptolemy (see section 3.3). The new aspect is that the method is now
consequently applied to the domification problem. In the Placidus domification,
the cusp of house XII is the uniquely determined point of the Ecliptic that has
made precisely 1/3 of the time it needs from rising to culmination.
This is again a spherical system of mundane position, although it is not based on
position circles. Contrary to the methods based on position circles, it is not possible
to assign a Placidus mundane position to the points of the circumpolar region.
But, as mentioned, this is not really a problem, since the concept of mundane
position makes sense only for points that pass the four angles (rising, setting and
the two culminations).

90°

0°
N E S W N

Figure 13: Division of daily arcs (Φ = 50◦ ).

Figure 13 shows the daily arcs of the visible celestial hemisphere in an Azimut/Al-
titude diagram. The dashed curves refer to the motion of points in the circumpolar
region, they are in fact concentric circles around the celestial pole. The solid curves
are the daily arcs of the points subject to rising and setting. All these daily arcs
are nothing else than the declination parallels inside of the domain D Φ , as seen
from an observer. The bold curves, crossing the daily arcs transversally, represent
the house boundaries according to Ptolemy/Placidus/Maginus, obtained by
dividing these daily arcs in the proportions 16 , 26 , . . . Again, the figure shows that
an extension of these mundane position curves into circumpolar region does not
make sense, since there is no arc that could be divided proportionally. 26
26
O. A. Ludwig, however, uses Meridians to join the end points of the mundane position curves with
the celestial north pole. This way, he achieves a continuous extension of the curves to the complete
hemisphere. But besides of producing non-smooth transitions, the choice of Meridians is arbitrary.
His claim that this extension of the mundane position curves goes back to Placidus himself seems
improbable to me, since Placidus always worked inside DΦ .
5 SPHERICAL HOUSE SYSTEMS 29

Placidus enriched the theory of directions by some own additions: In his attempt
to harmonize the two contradictory passages about house cusps in Tetrabiblos,
III.10 (see aboce, section 3.1), he introduced the notion of mundane aspects: Two
planets are in a mundane aspects if the difference of their mundane positions results
in the aspect angle. Therefore, the MC always is in mundane U with ASC, in T
with cusp XII, and so on. Building on these new mundane aspects, Placidus
went further and developed a new theory of mundane directions. In its elementary
form, one does not use the position curve for the position µ of the Significator
itself, but for the position µ + A, where A is the aspect angle. Further extensions
are the usage of crepusculine arcs for directions to the Sun, the notion of mundane
parallels (directions to equal distances from the Meridian, measured with mundane
positions), and, published by his followers, the so-called rapt parallels, where both
aspect partners are considered moving until they reach equal distances from the
Meridian (again in terms of mundane position).
From his examples in [24], it can be seen that in the majority of cases, Placidus
worked with temporal mundane position (definition VI), sinces he determines the
directional arc ∆ of a Significator S to a promissor P using the proportion
SAp
∆ = |M Ds · − M Dp |, (1)
SAs
where he calls the first term of the right-hand side the ”secondary distance” of P .
In rare cases, however, he proceeds differently: 27 He constructs the great circle
passing through the Significator and the Equator point with the same mundane
position. This great circle will in general not be a position circle, since it may not
contain the South and North point of the Horizon. Nevertheless, he sometimes
uses these great circles as an alternative to the proportional formula 1. Aiming
a precision of 1◦ for the final result, the directional arc, he did not bother much
about the difference between these two methods.
He admitted the use of these great circles as approximation of the mundane po-
sition curve, but he vehemently refused the usage of position circles for house
constructions. In Canon XII of his Primum Mobile, he explains how to determine
the ”polar elevation of a Significator” which determines the great circles which
he considers allowed for directional computations. ”I have no idea of circles of
position which are directed through the common sections of horizon and meridian,
but those that are described by the proportional distances of the stars towards the
angles; and we may, by means of a very easy method, know the Pole’s elevation
upon the Ptolemaic circle of any star whatever...” 28 The method then presented
simply is an interpolation of the (given) polar elevations of the surrounding house
cusps (per regulam auream).
It has often been complained that, due to the inequality of the daily and nightly
semi-arc, the Placidus method seems to produce a discontinuity since propor-
tional sections of the Ecliptic change their size at the Horizon. But it has been
27
see [24], p.197 for an arbitrary example
28
[24], Canon XII, p. 47
6 AFTER PLACIDUS 30

overseen that the Placidus method is not based on sizes but on motion. He
criticizes those methods that try to divide the sky into equal spaces as ”geometric
methods”, irrelevant to the domification problem. Placidus himself knew the
argument of inequal daily and nightly house sizes. In Physiomathica, he answers
that he is regarding the quality, not the quantity of the motion. Although a star
may need different times over and under the Horizon to pass through a house,
this is only an irrelevant quantitative difference – important is the quality, being
expressed as a proportion of quantities, relating the point to the four angles. 29

6 After Placidus
The Placidian works had been forbidden by the censors of the catholic church in
1687, but they found friendly asylum in Great Britain, and his main books were
translated by Manoah Sibly (1789) and John Cooper (1814). The Placidian
Table of Houses, published by Rafael I. in his Ephemeris, and in a textbook
Dictionnary of Astrology by J. Wilson (1819), contributed a lot to the success of
this house system in Great Britain. Zadkiel in his Grammar of Astrology (1849,
[34]), Simmonite with the Complete Arcana of Astral Philosophy (1840), A. J.
Pearce with his Textbook of Astrology (1911), and Alan Leo (The progressed
horoscope, 1906) gave instructions on how to do directions according to Placidus,
shifting the focus towards the use of polar elevation instead of temporal position
curves. Some authors, like Zadkiel in [34], completely abandoned the mundane
position curves in favor of the polar elevation method. It seems that Zadkiel was
led to this method by practical considerations, since it is clear from the text that
he did not fully understand the underlying concept. 30
At the time when the astrology renaissance reached Germany, in the first decades
of the 20th century, and astrology was propagated by the publications of Karl
Brandler-Pracht, A. Kniepf and A. Bethor, the situation began to change:
Many new astrological systems were created (not only in Germany); some of them
were completely new (like the Witte method), others were going back to almost
forgotten older astrological sources, claiming to reinstall the ”true astrological
tradition”. It was realized that many of the methods delivered by tradition, being
applied like recipes of a cookery-book, were lacking theoretical foundations. In the
sequence, many authors contributed to work out these foundations, among them
E. C. Kuehr, O. A. Ludwig, W. Noesselt, and Z. Wassilko.
29
Physiomathematica, p.189, quoted in [11], thesis 29.
30
Example: On page 456, he defines the term circles of position as ”small circles bearing the same
relation to the Meridian circle which the parallels of latitude do to the Equator” – which is completely
wrong.
6 AFTER PLACIDUS 31

6.1 Kuehr
The german astrologer Regiomontanus was rediscovered in this time – this was
clearly favoured by the nationalist spirit of the time. Regiomontanus was praised
for the elegance, clarity and purely geometric nature of ”his” house system. Some
astrologers began to study spherical trigonometry and were fascinated of the ge-
ometrical concepts. After the shape of the Placidus mundane position curves
had been worked out by O. A. Ludwig,31 many astrologers shared the disdainful
judgement of Zoe Wassilko: ”I think I am speaking in the sense of many oth-
ers, when judging the placidian temporal hour curves as of a very affected nature.
Compared to the regiomontanian circles of position, uniting in complete harmony
with Meridian and Horizon, they give the impression of artificially constructued
curves.”32
The fashion of that time was against the temporal position curves, favouring any
alternative that was based on great circle instead. It was highly en vogue to know
about polar elevations, position circles etc. and to make use of these items in
a directional system. E. C. Kuehr, although belonging to the placidian camp,
could not resist this fashion – and came back to Zadkiel, propagating that all
directions should be computed on the base of polar elevations. Other specialities
of his method, as the Naibod key instead of the true time key used by Placidus,
his rejection of mundane directions, and the mixed usage of ecliptical latitudes
(only the Significators are computed with their own ecliptical latitude, whereas
the Promissors are considered as sensitive points of the Ecliptic itself, having no
latitude) are not relevant to our topic.
Like W. A. Koch, E. C. Kuehr feels supported in his view by some passages in
Tetrabiblos stating that the planets of the horoscope should be made to Ascendants
of auxiliary horoscopes.33 For him, as for later followers like W. Lang, this means
that every house cusp and every planet of the horoscope is associated with an
individual ”Horizon” which is in contact with it and which is constructed with
an appropriate polar elevation and oblique ascension, in correspondence with its
mundane position.34 Of course, nothing is really rising at these ”Horizons”: they
are nothing more than great circles in a general position. The word is chosen only
to demonstrate the construction idea of intermediate mundane position circles, 35
which is carried out in analogy to the real Horizon.
As it turned out, Kuehr’s direction method is not new, since it goes back to
Placidus himself, who sometimes, and to 19th century astrologers like Zadkiel,
31
[18], p. 218
32
talk held on the XII. astrology congress in Munich, 1934, [33] p. 301
33
[28], III.4, p.249: ”For the rest, in carrying out these particular inquiries, it would be fitting and
consistent to set up the paternal or maternal place of the sect as a horoscope and investigating the
remaining topics as though it were a nativity of the parents themselves.” Similarly at the end of III.5,
p. 255, and some other places.
34
[16], IV.3, p. 122p, and [17], III.6.b, p. 189p
35
which are not circles of position in the sense of consequence 3, since they may not meet the north
and south point of the Horizon
6 AFTER PLACIDUS 32

who always used it. It is clear that the set of these ”Horizons” does not constitute a
system of mundane positions – they are not even disjoint to each other. 36 Moreover,
such a ”Horizon” does not comprise the points of equal mundane position: In
general, the only points with the required mundane position are, by construction,
the intersection point with the Equator and the planet itself.
Another problem is that the method is not well-defined for points on the equator,
since there the two points defining the ”Horizon” coincide. One could overcome
this problem by defining the ”Horizon” to be the tangent circle to the mundane
position curve starting at this point, resulting in a polar elevation given by
MD
tan ϕ = · tan Φ. (2)
SA
To my knowledge, neither Kuehr nor the British authors of this method mention
this necessary extension of the method. 37

6.2 Koch Houses

The Koch system38 is a product of the same time and of the same spirit: It
was the enthusiasm for the spherical geometry, particularly for everything based
on great circles and polar elevations which made this new house system appear.
Koch, formerly a Regiomontanian, takes the mentioned quotations of Ptolemy
literally that each planet of the nativity ”can be made a Horizon”. The Horizon,
however, has one and only one polar elevation: The geographical latitude of the
birthplace. Therefore, Koch postulates that the house construction has to be
based entirely on great circles with the polar elevation of the birthplace 39 For
this goal, he translates the Horizon along the Equator until it contains the MC
and considers this translated Horizon, not the defining Meridian, as constituent
position curve for the MC. By trisections of the semi-arc of the MC, he obtains
the oblique ascensions of the intermediate house cusps (which are then determined
as intersections of the great circle with this oblique ascension and with the polar
elevation the latitude of the birthplace.
Similarly, he postulates for directions that the Significator always has to obtain
the polar elevation of the birth place. 40
The fixed idea of Koch was that mundane position curves always have to be
local horizons, i.e. great circles with the polar elevation Φ of the birth place. A
36
It follows also from consequence 3, since the ”Horizons” are great circles but not position circles in
general.
37
It may be remarked that formula (2) has been used as the base for another modern variant of the
Placidus house system, the so-called topocentric houses[27]. The authors Polich and Page use (2)
as definition of the polar elevation, together with the OA of the Placidus system, yielding in slightly
different house cusps. They use the same circles also for doing directions. This is an even stranger
method: The planet that defines the mundane position curve is itself not even contained in this curve!
38
which in fact was a byproduct of geometrical investigations of Zanzinger and Specht and was
never intended by the authors to be published
39
This explains, by the way, the name ”Geburtsort-Häusersystem”.
40
[12], p.108
7 DISCUSSION 33

critical statement of Wassilko should finish this section: ”One cannot glue several
different Ascendants together and use them as house boundaries for a time to which
they are not related at all.”41

7 Discussion
If we overview the evolution of the house systems, leaving out some inessential
variations, the following main stages of development can be made out:

• While the first methods worked on the Ecliptic alone, the introduction of the
mathematical MC and the increased geometrical skills made it clear that the
reference system for the houses, as a partition based on the daily rotation of
the sphere, could not be the Ecliptic.
• Beginning with Rhetorios, astrologers switched to the Equator as reference
plane. Since the degrees of the Equator all ascend with the same velocity,
the Equator can be regarded as a big celestial clock. For a division of the
sky according to its daily rotation it is therefore natural to base it on the
Equator.
• The more stress was laid on the connection of house division and directional
techniques, the more it became apparent that the house division should be
extensible to the celestial sphere. The search for spherical house systems
began.
• The Campanus/Al Biruni method was geometrically most satisfying, as
the houses could be derived from a global division of the sphere by position
circles into twelve perfectly equal diangles.
• Haly Abenragel and Regiomontanus corrected this tendency to the
other direction and put the Equator back in his rights for this problem.
• Maginus and Placidus come back to the house division as division of times
– more precisely: as division of the daily arcs described by the stars.

The system of mundane position that had been outlined by Ptolemy in his famous
chapter III.10, has been used for primary directions by many astrologers. It is a
strange fact that his concept has been used for domification not before the 17th
century. How can this delayed evolution be explained? Why did we have to wait
almost 15 centuries, until the house system has been worked out by Placidus
and Maginus which belongs organically to the Ptolemeic concept of mundane
positions?
I think, there is a number of reasons for this.

1. Lack of mathematical understanding.- First of all, of course, the understand-

ing of the celestial geometry had to increase in order to understand the inti-
mate connection between house systems and directional systems. This level of
41
[33], p.84
7 DISCUSSION 34

understanding had been reached with the time of the spherical house systems
(Abenragel, Campanus, Regiomontanus). In Tetrabiblos itself, houses
and directions appear yet as more or less disparate subjects.
2. Misunderstanding Ptolemy.- The misleading hint of Ptolemy that his
mundane position curves are ”most nearly position circles” caused many at-
tempts for a purely geometric domification, defining the house cusps by inter-
sections of circles. It had been overseen that Ptolemy mentions the position
circles only as an approximation for the real mundane position curve. Instead,
astrologers like Regiomontanus were attracted by constructive solutions
that were as clear or ”rational” as a construction in elementary geometry.
3. Ideal that astronomy should work with circles only.- Apart from the ideal of
a ”rational”, constructive solution to spherical problems, there was another
postulate of astronomy requiring that all problems of the celestial sphere can
be described or solved using circles and spheres. The circle and the sphere
with their perfect symmetry appeared to be the only shapes adequate to
the dignity of the celestial spheres. It seems no accident to me that this
postulate has been rejected almost simultaneously in astronomy by Kepler
and in astrology by Maginus and Placidus. The beginning of the mod-
ern times made it possible to break this strong tradition. Some courage was
necessary to accept a position curve instead of a great circle as house bound-
ary. This step was as weightful as the transition from the epicycle model to
the acknowledgment of non-cyclic planetary orbits in astronomy (Kepler’s
Astronomia Nova).
4. Baroque emphasis on harmonic proportions.- In my opinion, the baroque zeit-
geist was the midwife and the necessary background for the development of
the Placidus houses. One of its main traits is to view the world as ”arranged
by measure, number and height”42 This can, for example, be read off from the
baroque theory of music which tried to define the musical harmonies from
certain geometric or arithmetic proportions. This aspect of music, although
common since the Pythagorean monochord experiments, has been exag-
gerated to an extreme degree in baroque theory of music. It sometimes goes
so far to define music as a branch of mathematics. musica est scientia
mathematica subalternata comprimis arithmeticae... 43 The creator
of musical harmonies bases his work on the same harmonies that he finds
ubiquitous in the universe, a testimony of God’s creation. In this sense,
Kepler was a child of his time when working on the Harmonice Mundi –
like A. Kircher, R. Fludd and others. Kepler, favoured by his times,
continued a work to which already Ptolemy had contributed with his book
on Harmonies. In his Harmonice mundi, Kepler demonstrates that the pro-
portions defining musical harmonies are indeed omnipresent in the world; in
particular, the power of astrological aspects can be derived from these propor-
tions. With this almost exclusive focus on harmonic proportions, it became
42
(Wisdom of Salomon, 11:20), which was a frequently quoted (apocryphal) bible passage to justify
the claim of omnipresent harmonic proportions.
43
Lippius, quoted in [6], p.11
7 DISCUSSION 35

possible, if not even necessary, to base mundane position on harmonic propor-

tions too. For the part of directions, this was already afforded by Ptolemy.
The way was now free for astrologers to work out the underlying mundane
position concept. Indeed, if according to Kepler the aspects as effective
relations of a nativity can be based on proportions, then this should hold
also for houses and directions.
5. Turn from cosmos to individual.- In contrast to other house systems, the
Placidus system lays a special emphasis on the subjective perception. For
Placidus, the division of the sky is not merely a geometrical problem. It is
the human soul who listens to the celestial harmonies and realizes the propor-
tions and is affected by them. This makes the house cusps and the mundane
aspects relevant. This again is an affinity to Kepler for whom the power of
proportions is based on a hidden affectibility of the soul who is stimulated
when the sensually percepted proportions coincide with the harmonic ideals:
”Eine geeignete Proportion in den Sinnesdingen auffinden heißt: die Ähn-
lichkeit der Proportion in den Sinnesdingen mit einem bestimmten, innen im
Geist vorhandenen Urbild einer echten und wahren Harmonie aufdecken, er-
fassen und ans Licht bringen. So findet der Geist Ordnung und Proportion in
den Tönen und Strahlen, daß aber diese Proportion harmonisch ist, bewirkt
die Seele durch die Vergleichung mit ihrem Urbild. Die Proportion könnte
nicht harmonisch genannt werden, sie besäße keinerlei Kraft, die Gemüter
zu erregen, wenn dieses Urbild nicht wäre.” 44
6. The dawn of the physical age.- As Max Caspar points out, Kepler’s finding
of the elliptic shape of the planetary orbits was inspired by his vision of
joining astronomy and physics45 Similarly, Placidus calls astrology physio-
mathematics and tries to found the complete astrological building on light
and motion.46
7. Increasing individualism.- The Placidus house system carries the signature
of an age of individualism, since each point of the sky is considered with
its proper mundane coordinate system. This has been characterized in an
article by Dieter Koch (translation by me): ”One cannot say that the
Placidus system follows a particular measure of time, as the system of Re-
giomontanus does – instead, each point of the sky has its own measure of
time. Everything is relative, and each point follows its own individual way.
Placidus conforms to our zeitgeist, which is characterized by relativism and
perspectivism.”47 This observation is not a mere reflection on the house
system. Turning it around, the new individualism that started with the re-
naissance times can be seen as a factor which favoured the development of
this house system.

44
[9], p. 205
45
[5], p. 156p.
46
[11], §§1-2
47
[15], p.16
7 DISCUSSION 36

8. Increasing routine in solving iterative equations.- The analogy between Kep-

ler and Placidus is closer, even on a purely formal level. They both had,
in a central place of their theories, to deal with a basic equation that can be
solved only iteratively. Kepler’s second law leads to the equation

E − e sin E = M, (3)

where M and e are known and the excentric anomaly E is to be determined. If

E is known, the heliocentric position of the planet can be computed directly.
Equation 3 is a transcendental equation that can be solved approximately
(for example, by iterating En+1 = M + e sin En ).
On the other hand, Placidus and Maginus are looking for Ecliptic points
that divide their own diurnal or nocturnal semi-arc in a certain fixed propor-
tion a. If ε denotes the obliquity of the Ecliptic, δ the declination and α the
right ascension of the Ecliptic point, this produces the equation

α = t + a · (90◦ + arcsin(tan ε tan ϕ sin α)) ,

which can be simplified by setting C := tan ε tan ϕ and α 0 := a · 90◦ + t, to

an equation that we may call the Placidus-Maginus equation:

α − a arcsin(C · sin α) = α0 . (4)

This again is a transcendental equation that can only be solved iteratively.

Moreover, for small C and working with radians instead of degrees, we can
use the approximation arcsin x ≈ x, and (4) reduces to

α − a · C · sin α = α0 ,

i.e. to the Kepler equation (3). This means, the Kepler equation is a first
order simplification of the Placidus-Maginus equation!

In the course of this paper, we have seen that the historical manifestation of an
idea is neither a unique event nor a linear, straight process. There are many
sideways, there are necessary stages to reach first, there are phases of stagnation
and even fallbacks. But the example shows that in due time the idea will be fully
established.
REFERENCES 37

References
[1] Peter von Abano. Astrolabium Planum. Cod. pal. germ 832, Bayern, >1491.
[2] Franz Boll, Carl Bezold, and Wilhelm Gundel. Sternglaube und Sterndeutung.
Teubner, Leipzig, 4th edition, 1931.
[3] A. Bouché-Leclerq. L’astrologie greque. Leroux, Paris, 1899.
[4] Rafael Gil Brand. Lehrbuch der klassischen Astrologie. Chiron Verlag, Mössin-
gen, 2000.
[5] Max Caspar. Johannes Kepler. Kohlhammer, Stuttgart, 2nd edition, 1950.
[6] Rolf Dammann. Der Musikbegriff im deutschen Barock. Am Volk Verlag,
Köln, 1967.
[7] Ralph William Holden. Astrologische Häusersysteme. Chiron Verlag, Mössin-
gen, 1998.
[8] E. S. Kennedy and D. Pingree. The astrological history of Māshā’allāh. Har-
vard University Press, Cambridge, 1971.
[9] Johannes Kepler. Weltharmonik. Oldenbourg Verlag, München, 1982.
Übersetzt und eingeleitet von Max Caspar.
[10] Wilhelm Knappich. Ptolemäus und die Entwicklung der Häusertheorien.
Zenit, I(8):270–277, 1930.
[11] Wilhelm Knappich. Placido de Titi’s Leben und Lehre. Zenit, VI(7-11), 1935.
[12] Wilhelm Knappich. Entwicklung der Horoskoptechnik vom Altertum bis zur
Gegenwart. Number 38/39 in Qualität der Zeit. Österreichische Astrologische
Gesellschaft, Wien, 1978.
[13] Wilhelm Knappich and Walter Koch. Horoskop und Himmelshäuser. Sir-
iusverlag, Göppingen, 1959.
[14] Leo Knegt. Astrologie, Wetenschappelijke Techniek. Amsterdam, 1928.
[15] Dieter Koch. Wie wähle ich zwischen den verschiedenen Häusersystemen?
Meridian, 7/8:11–20, 1994.
[16] Erich Carl Kühr. Berechnung der Ereigniszeiten. Rudolf Cerny, Wien, w/o y.
[17] Walter Lang. Die Astrologie im heutigen Weltbild. Arkana Verlag, Heidelberg,
1986.
[18] Otto A. Ludwig. Ein Beitrag zum Häuserproblem. Zenit, I(6-7), 1930.
[19] Maginus. Tabulae primi mobilis. Venice, 1604.
[20] Marcus Manilius. Astronomicon Libri V. Reclam, Stuttgart, 1990.
[21] Varaha Mihira. Brat Dschataka - Das Buch der Nativitäten. W. Wulff (pub-
lisher). Hamburg, 1925.
[22] Otto Neugebauer and H. B. van Hoesen. Greek Horoscopes. American Philo-
sophical Society, Philadelphia, 1959.
REFERENCES 38

[23] Martin Pegius. Geburtsstundenbuch. Basel, 1570.

[24] Placidus de Titis. Primum Mobile, translated by John Cooper. London, 1983.
[25] Rüdiger Plantiko. Primärdirektionen - eine Darstellung ihrer Technik. Chiron
Verlag, Mössingen, 1996.
[26] Rüdiger Plantiko. Algunos comentarios sobre el concepto ptolomeico de Pri-
marias. Mercurio-3, 27:54–64, 2000. Ed. Jaume Martin, Barcelona.
[27] Wendel Polich and Nelson A. P. Page. The topocentric system of houses.
Spica, 3(3):3–10, 1964.
[28] Ptolemaios. Tetrabiblos. Translated by F. E. Robbins. Harvard University
Press, Cambridge, 1953.
[29] Ptolemaios. Handbuch der Astronomie (Almagest). Translated by K. Manitius,
volume I. Teubner, Leipzig, 1963.
[30] Regiomontanus. Tabulae directionum profectionumque. 1490.
[31] Gustav Schwickert. Die Direktionslehre. Number 45/46 in Qualität der Zeit.
Österreichische Astrologische Gesellschaft, Wien, 1983.
[32] A.E. Thierens. Astrologische Berekeningen. J.F. Duwaer, Amsterdam, 1932.
[33] Zoë Wassilko. Eine Dokumentation. Number 48-51 in Qualität der Zeit.
Österreichische Astrologische Gesellschaft, Wien, 1987. Zusammengestellt von
Sándor Belcsák.
[34] Zadkiel. Grammar of Astrology. London, 1849. Reprint by Ascella Publica-
tions, Mansfield Notts.