ROMAN CLOTHING AND

FASHION

ALEXANDRA CROOM
 This edition published 2010.
This electronic edition published 2012.

Amberley Publishing
The Hill, Stroud, Gloucestershire GL5 4EP

www.amberley-books.com

Copyright © Alexandra Croom 2010, 2012

The right of Alexandra Croom to be identified as the Author of this work

has been asserted in accordance with the Copyrights, Designs and Patents
Act 1988.

All rights reserved.

No part of this book may be reprinted or reproduced or utilised in any
form or by any electronic, mechanical or other means, now known or
hereafter invented, including photocopying and recording, or in any
information storage or retrieval system, without the permission in writing
from the Publishers.

British Library Cataloguing in Publication Data.

A catalogue record for this book is available from the British Library.

ISBN 978-1-84868-977-0 (PRINT)

ISBN 978-1-4456-1244-7 (e-BOOK)
 CONTENTS

List of Illustrations
Acknowledgements
1 - Introduction
2 - Cloths and Colour
3 - Men’s Clothing
4 - Women’s Clothing
5 - Children’s Clothing
6 - Beauty
7 - Provincial Clothing
8 - Conclusions
Pictures Section
Glossary
References
Weaving Terminology
Bibliography
 LIST OF ILLUSTRATIONS

TEXT FIGURES

1 - The costume of goddesses

2 - Woman spinning
3 - Tunic forms
4 - A clothes press
5 - Tunics of the first and second centuries
6 - Tunics of the third and fourth centuries
7 - Tunic decorations
8 - Portrait of Stilicho
9 - Tunics
10 - Togas of the first to fourth centuries
11 - The ‘Brothers’ sarcophagus
12 - Togas of the fifth and sixth centuries
13 - Short togas
14 - Mantles and cloaks
15 - Cloaks and capes
16 - Tombstone of Ares
17 - Leggings
18 - Puttee
19 - Loincloths
20 - Knitted sock
21 - Footwear
22 - Footwear
23 - Hairstyles of the first to third centuries
24 - Hairstyles of the third to sixth centuries
25 - Headwear
26 - Religious costumes
27 - Amulet
28 - Intaglios and finger-rings
29 - Stola
30 - Tombstone showing woman in a gap-sleeved tunic
31 - Gap-sleeved tunic fastenings
32 - Gap-sleeved tunic fastenings
33 - Short-sleeved tunics of the first and second centuries
34 - Short-sleeved tunics of the third century
35 - Dalmatics
36 - Tunic decorations
37 - Late forms of tunic
38 - Slaves with belted tunics
39 - Portrait of Serena and Eucharius
40 - Tunic with ornamented belt
41 - Mantle
42 - Fourth-century decorated mantle
43 - Venus wearing a breastband.
44 - Breastbands and briefs
45 - Socks
46 - Hairstyles of the first and second centuries
47 - Hairstyles of the second and third centuries
48 - Hairstyles of the third to sixth centuries
49 - Hats and headcoverings
50 - Freedwoman wearing cap
51 - Parasols and fans
52 - Sandal
53 - Sandals
54 - Wooden clogs
55 - Bridal and religious costumes
56 - Finger-rings and bracelets
57 - Necklace
58 - Babies in swaddling clothes
59 - Children on the Ara Pacis, Rome
60 - Tunics of young girls
61 - Hairstyle of young girl
62 - Tombstone with native and Roman costume
63 - Greek costume
64 - Syrian costume
65 - Judaean costume
66 - Egyptian costume
67 - North African costume
68 - Spanish costume
69 - The Gallic coat
70 - Romano-Gallic costume
71 - Romano-British costume
72 - Costume of the Rhine Valley
73 - Costume of Lower Germany
74 - Woollen clothing from a bog deposit
75 - Dacian costume
76 - Pannonian and Norican costume
77 - Pannonian and Norican hats

COLOUR PLATES

1 - The Emperor Justinian and his court

2 - The Empress Theodora and her court
3 - Men at a bakery
4 - Fourth-century hunters
5 - The method of putting on a third-century toga
6 - Reconstruction of a cape
7 - Reconstruction of the cape with added gusset
8 - Procession to baths, Piazza Armerina, Sicily
9 - Dead or wounded hunter
10 - Coin portraits showing the hairstyles of Emperors
11 - Reconstruction of a gap-sleeved tunic and stola
12 - Reconstruction of a gap-sleeved tunic
13 - Woman having her hair dressed
14 - Portrait of woman from Egypt
15 - Painted shroud from Egypt
16 - Family portrait in gold and glass
17 - Reconstruction of leather briefs
18 - Women exercising in briefs and breastbands
19 - Coin showing the hairstyles of Imperial women
20 - Jewellery of the first to third centuries
21 - Child’s tunic from Egypt
22 - Tombstone of Regina
23 - Reconstruction of the tombstone of Regina
24 - Provincial male costume
25 - Provincial female costume
26 - Undyed wool
27 - Dyed wool
28 - First-century tunic
29 - First-century tunic
30 - Third-century tunic
31 - Third- to fourth-century tunic
32 - Fourth- to sixth-century tunic
33 - Decorated cloak
34 - Shoulder cape
35 - Hose
36 - Puttee
37 - Puttee
38 - Pannonian hat
39 - Gap-sleeved tunic
40 - Short-sleeved tunic
41 - Dalmatic
42 - Fourth- to sixth-century tunic
43 - Mantle
44 - Women with fan and parasol
45 - Tube-dress
46 - Brooch fastening the tube-dress
47 - Tube-dress
48 - Gallic coat
 ACKNOWLEDGEMENTS

Thanks are due to Karen Dixon and Richard Underwood, without whom
this book would never have been attempted. I must also thank the
members of Cohors V Gallorum, and in particular Dominique Leslie and
Philip Clark, for their help during research for this book, as well as all re-
enactors and work colleagues with whom I have discussed costume over
the years, whether they wanted to or not. Katherine Croom has most kindly
used her expertise to produce the index, while Roger Oram provided me
with some of the illustrations as well as useful advice. Thanks are also due
to Christoph Reichmann of the Museum Burg Linn, Jenny Hall of the
Museum of London and Caroline Imlah of Tyne and Wear Museums, for
their help with information and illustrations. Last, but not least, I must
thank Paul Bidwell and William Griffiths, who both kindly agreed to look
over the text, and who provided me with both suggestions and corrections;
any faults or mistakes remaining are the responsibility of the author.
In addition I need to thank the following for all their help with the
second edition: Phil Clark, Peter Coppack, Kathy Croom, Karen Griffiths,
Bill Griffiths, Aitor Iriate, Hannah-Jane Taskis and Jane Thompson, plus
the re-enactment groups Cohors Quinta Gallorum and Cohors Prima
Gallica. Many thanks!
1
 INTRODUCTION

The most common image of Roman costume is of a clean-shaven man,

wearing a brilliant white toga long enough to sweep the ground. The image
is correct – for a small proportion of men, for a short length of time, in a
distinct area of the Empire. In fact, Roman costume varied from country to
country, and also invariably changed over time. It is a huge topic to study,
and there are still many areas and subjects that need much more in-depth
research or clarification. This book does not attempt to answer all the
questions, but is intended to provide at least an introduction to this
fascinating and complex topic.
This study will look mainly at the clothing worn during the years of
the Roman Empire. As Rome was traditionally founded in the eighth
century BC, there were at least 250 years of rule by the Kings and 500
years of the Republic before the Empire even came into existence. The
Empire itself lasted for nearly 500 years in the West, and for even longer
in the East, where it was gradually transformed into the Byzantine Empire
of the medieval world.
The evidence is taken from three main sources: art, literary evidence
and the surviving textiles themselves. ‘Art’ covers a wide range of objects,
from portrait statues, tombstones, official reliefs, mosaics and wall
paintings, to the smaller decorative arts such as coins, small pipeclay
statuettes, and even graffiti. Although art often seems to provide detailed
information on costume, the evidence must still be treated with some
caution. Only Romans with money could afford to commission a life-sized
statue, bust or figured tombstone, and therefore the clothing depicted is
that of the rich or middle classes rather than the lower classes. Just as
early photographs in the nineteenth century were expensive, and showed
the sitters carefully posed and wearing their best clothes, so Roman
portraiture showed people in their formal clothing, and not necessarily in
what they would have worn every day. However, the Romans also liked
images of everyday life, including scenes of rural activities or domestic
scenes featuring the lower classes and slaves, which give at least some
insight into what poorer people wore.
 The misleading nature of the artistic record can be seen in the
example of the toga. The one item of dress that everyone associates with
the Romans was hardly ever worn in the countryside in Italy, and was in
decline in everyday use in an urban setting from the early second century
AD according to literary evidence, and yet is still found depicted in art
until the fourth century. The toga developed a symbolic importance that
made it the definitive male dress. As such, it was used on formal
ceremonial occasions, thus appearing frequently in art as these formal
ceremonial occasions were precisely those most likely to be
commemorated. Art does not, therefore, necessarily reflect everyday life.
Similarly, people today recognise a picture of a man in a kilt as a
Scotsman, yet anyone travelling in Scotland would be sadly disappointed
by the number of men actually wearing one.
Different media will also influence the depiction of costume.
Sculptors, for example, were particularly interested in the play of light and
shadow on the folds of cloth, and in mainstream art always show plenty of
heavy folds, realistic or not, in the cloth. They were not so interested in
detail, and so rarely show jewellery or hairpins, while details of patterns or
borders may have been left to be painted on the finished statue, most of
which were originally brightly coloured. On the other hand,
representations on mosaics frequently simplified complex elements such
as the folds in the cloth, belt fittings, and decorative patches on tunics.
Wall paintings could show both small detail and colour, but are fairly
limited in number. Egyptian mummy portraits are more numerous, but the
great majority are only head and shoulders.
Documentary evidence is the second major source of information
about Roman clothing, but this too can be misleading. Many of the authors
quoted belong to the first century AD, and therefore reflect the clothing
and terminology used during one short period. The Empire lasted for a
very long time, and what held for the first century is unlikely to hold for
the sixth century. For a start, it is difficult to tell if words still in use
changed their meaning over time. For example, in medieval England the
word ‘coat’ simply meant a garment over underclothes worn by men; 600
years later, it means a heavy garment with buttons up the front, worn by
both men and women over their indoor clothing for outside use.
 There is also the problem of defining what the words actually mean.
It would be difficult to explain the exact difference between the terms
sweater, jumper, pullover and woolly, or the subtle differences between a
jersey and a guernsey. The last two, of course, were originally applied to
garments coming from the islands where they were first made, but at some
time the terms changed and just meant sweaters made in the style of those
made on the islands. Most people would have no idea, and probably little
interest, in when exactly this change came about. The word jersey can thus
be equated with ‘jumper’, and yet in a different context ‘jersey’ can also
be used of the shirts worn by footballers while ‘jumper’ cannot. This
naming of garments after the place where they were made also happened
in the Roman period, raising similar problems in deciding whether they
were made in the named places, or made elsewhere copying that style.
The final source of evidence is the garments themselves, a number of
which have survived. Organic material survives in either very wet or very
dry conditions, so while very few have been found in Italy itself, many
have been found in the dry conditions of Egypt, and a smaller number
from wet sites in France and Britain and, outside the Empire, in Denmark.
These garments can show us exactly what was being worn, but of course
they will reflect local fashions and manufacturing methods that may not
represent practice elsewhere in the Empire. Other drawbacks must also be
borne in mind, such as the poor dating of many of the Egyptian pieces, or
the fact that linen does not usually survive in bog-conditions, leaving a
biased picture of the clothing available.
As far as possible, evidence for Roman clothing has not been taken
from depictions of deities or heroes. Roman art did not come with
captions, and therefore figures had to be easily recognisable by their faces,
clothes and attributes, in exactly the same manner as medieval saints.
Pompeii, for example, has a large number of wall-paintings taken from the
stories of legendary heroes such as Perseus and Hercules. Some of these
pictures may be copies of Hellenistic originals and therefore show
Hellenistic-style costume, while others that are not direct copies may well
use Hellenistic costume to give a historical feel to the picture (similar to
the way modern illustrators of fairy tales dress the figures in mock
medieval or eighteenth-century dress). The costume of true deities needs
to be treated with equal caution. For a start they are simply not human, so
there is no requirement for them to wear human costume. It is quite clear
that goddesses in particular are shown wearing very specific forms of
clothes that reflect Greek prototypes rather than contemporary clothing.
The deities were special creatures, outside everyday life, so there was no
reason for the Romans to believe they wore contemporary dress, any more
than Victorian artists show the Virgin Mary or Jesus in full evening dress,
let alone informal daywear. Like us, the Romans usually showed their gods
in vaguely ‘historical’ dress, and little reliance can be put on their
relevance to contemporary costume.
The goddess Victory, for example, is usually shown wearing a tube-
dress without sleeves, usually with a long overfold, fastened with a single
brooch on each shoulder and belted under the bust (1.1). An assortment of
other goddesses are shown wearing a tube-dress belted on the hips, with a
hip-length overfold, which is an ancient style of dress seen worn by
figures on the Parthenon c. 400 BC (1.2). Others, including Roma, Flora
and Venus, wear a tube-dress without sleeves, belted either on the waist or
low on the hips and fastened by a single brooch on each shoulder. On a
statue of Flora (1.3), one brooch has slipped down the arm, but instead of
the cloth forming a natural straight line from brooch to brooch, the cloth is
shown forming an unrealistic dogleg so that the breast is fully covered.
The tunic slipping off the shoulder is in fact a very common motif for
goddesses and mythological characters. It is occasionally found on
contemporary female portraits, such as sarcophagus portraits, but it seems
likely that it is used to indicate that the woman has joined the realm of the
supernatural, rather than reflecting genuine fashion.
1 The costume of goddesses. 1 Victory, relief in copper alloy, Caerleon; 2 Statue restored as
Hygeia, Museo Capitolino, Rome; 3 Flora, statue in Museo Archeologico Nazionale, Napoli; 4
Diana, statue in Vatican Museum.
There was also a tradition for empresses in portrait statues to be
shown in the guise of a goddess, while in the Antonine period in particular
there was a fashion for non-Imperial couples to be shown as Venus and
Mars (Kleiner 1992, figs 248-9). The woman could be half-naked or
dressed in a tube-dress, while the man was usually naked, but armed. In
some examples, the man is shown wearing normal dress (such as a toga),
while the woman is dressed as Venus – a potentially confusing mixture of
contemporary and historical dress that does not mean the woman ever
wore that outfit in real life.
There is some evidence, particularly in the fourth century, of
goddesses with contemporary jewellery and hairstyles, and historical
soldiers shown in contemporary dress, but that is no guarantee that the
whole of their costume is correct for the period; the artist, after all, was
under no obligation to be authentic. When scenes from the Bible became a
popular motif, it does seem that figures are actually shown in
contemporary clothing. However, information from the figures should still
be used with caution, as the artists are depicting historical scenes, and may
have introduced what they saw as ‘historical’ details. Despite this caveat
against relying on the costume of deities, it will be noted that some
examples in this study are taken from gods and heroes, where they provide
good instances of costume known from other sources showing humans.
2
 CLOTHS AND COLOUR

SPINNING

Thread was spun entirely by hand, using a spindle (2). The raw wool or
linen was tied to a stick called a distaff, and then teased out by hand, with
a weighted stick (the spindle) used to twist the raw material into thread
(Wild 1970, 31-40; Wild 2002, 8-10). The spun thread was also stored
temporarily on the spindle for convenience, as can be seen on
contemporary depictions of the spindle. Spinning was such a typical
activity that the distaff and spindle became a symbol of womanhood, and
were often shown on tombstones. Despite this, however, there are
remarkably few depictions of women actually spinning, rather than simply
holding distaff and spindle, and even fewer of them weaving (Cottica
2007, figs 36.1-2, 36.5-7). In spite of the symbolic value of clothes-
making, many women passed the work on to slaves if they could, and only
supervised the work. Even in the first half of the first century, the author
Columella complained about women who:
so abandon themselves to luxury and idleness that they do not
deign to undertake even the superintendence of wool-making,
and there is a distaste for home-made garments and their
perverse desire can only be satisfied by clothing purchased for
large sums and almost the whole of their husbands’ income (On
Farming 12, preface, 9).

LOOMS

There were two forms of loom used during the Roman period: the warp-
weighted and the two-beam loom (for terminology, see p.181). The size of
loom helped to dictate the form of the clothes made on it; rather than
make a long length of cloth and then cut it up into shorter lengths, the
Romans preferred to weave the required shape on the loom, which avoided
wasting excess cloth. Thus, in Egypt, extremely wide looms were
sometimes used, where tunics were woven from the wrist of one sleeve,
expanding to the full width for the front and back, and then narrowing
back down for the second sleeve, ending at the wrist (3.3), or else the body
and sleeves were woven to shape separately and then sewn together
(Carroll 1988, fig. 12). This way of weaving explains the preference for
the simple shapes of the basic range of Roman clothes.

2 Woman spinning. Sixth-century mosaic, Tabarka, Tunisia.

The warp-weighted loom rested at an angle against a wall (Wild
1970, fig. 53, pl. X; Hoffmann 1974). The threads hanging vertically
(called the warp) had clay weights tied to one end to keep them taut during
weaving. The width of the loom dictated the width of the cloth, but the
length could be as long or short as required. When the threads were set up
on the loom, they were cut to the required length (plus a little extra), with
the excess carefully tied to the weights to keep them out of the way until
required. The cloth was woven from the top downwards, until the heddle
rod was reached. The finished cloth was wound round the pole at the top,
and the weights retied further down the excess threads. This method
wasted a small amount of thread, as the warp could not be woven right
down to the weights, but these threads could be used to make long fringes
if required.
3 Tunic forms, showing seams (diagonal shading) and openings (black ovals). 1-2 Tunics made
from rectangles of cloth; 3-4 Long-sleeved tunics, showing layout of cloth on the loom; 5 Tube-
dress as made from rectangle of cloth. Drawn by R. Oram.
The second form of loom was the two-beam loom (also called a
tubular loom), which seems to have been the most popular form in the
Roman period (Wild 1970, fig. 57, pl. XI). Again, the width of the loom,
which was free-standing, limited the width of the cloth. In this case, the
warp was looped round two beams and was worked from the bottom up,
allowing the weaver to sit down while working. Depending on the way the
warp was set up, this type of loom could either produce a rectangle of
cloth, like the warp-weighted loom, or a continuous tube of cloth. If woven
as a tube, there was no waste of warp thread as with the warp-weighted
loom.
Clothes could be made from very simple pieces of cloth, such as two
rectangles of cloth sewn together, leaving gaps for neck- and armholes
(3.1). The openings did not even have to be hemmed, since they were
made from an uncut piece of cloth, and would have selvage edges or
starting and finishing borders on all four sides. Another form of tunic,
used by women in a number of different ancient cultures, could be made
by weaving a tube of cloth on a two-beam loom, or sewing a large
rectangle of cloth from a warp-weighted loom into a tube. The opening at
one end could then be part-sewn or held together by brooches to form
neck- and armholes (a tube-dress, or Greek peplos; 3.5). Once again, as
there was no need to cut the cloth, there was no need for hems. A surviving
example of a tube-dress from Iron Age Denmark (Walton Rogers 2007,
fig. 5.8) was made of a tubular piece of cloth 1.7 m (5 ft 7 in) long and 2.7
m (9 ft) in diameter (the great diameter being required as no armholes are
cut in this form of tunic, and the cloth has to reach from the shoulder, right
under the arm and back to the shoulder again, on both sides). More
elaborate forms of clothes, such as the long-sleeved tunic, the toga, or
cloaks with a curved edge, could also be woven to shape on the loom,
resulting in proper selvage edges, which are stronger than a sewn hem.
However, some surviving tunics do have their edges folded back and sewn
which, if it is not simply re-use of cloth, suggests some webs of cloth were
cut up for use.

CLOTH

The most commonly used fabrics were wool and linen, but other materials
were known, such as silk for the rich, goat-hair cloth and cotton (Wild
1970, 4-30). Non-woven materials used included felt, leather and fur, as
well as knotted fabrics such as sprang-work, knitting and netting. Pliny
mentions a number of plants used for producing cloth in different parts of
the Empire, such as esparto grass in Spain, as well as a very special form
of linen that would not burn (asbestos cloth: Natural Histories, 19.4.19).
The Romans could produce a wide range of different weave patterns
in their cloth, and weave cloth to whatever thickness was required. Martial
describes thick tunics from Padua made of ‘triple twill [that] uses up many
a fleece and only a saw can cut the thick tunics’ (Epigrams, 14.143), while
other authors refer to cloth so thin it was transparent.
Wool was the most commonly used material, closely followed by
linen. Pliny’s discussion of linen from different areas suggests that white
was the preferred colour, the more brilliant the better (Natural Histories,
19.2). Wool was considered to be softer than linen, and the best linen is
compared to wool: ‘no flax is more brilliantly white or more closely
resembles wool [than Pelignian flax]’ (ibid., 19.2.13). Silk was popular
because it could be woven into a very fine, lightweight cloth. It was worn
by both women and men, although it was always seen as rather decadent. It
was hugely expensive, since it had to be imported into the Empire from
China; during the late third century ‘a pound of silk was worth a pound of
gold’, and was thus literally worth its weight in gold (SHA Aurelianus
45.5). Cloth was frequently made of half silk and half linen or wool to cut
down the cost (Sebesta 1994a, 71).

LUXURY GARMENTS

There are repeated references to garments woven with gold thread from
the time of the Republic onwards. Nero, for example, was laid out in
‘white clothes woven with gold’ (Suetonius, Nero, 6.50), while the clothes
of the Emperor Commodus sold after his death included:
clothes of silk woven with gold thread of remarkable
workmanship, tunics, capes, cloaks and dalmatic tunics with
long sleeves and fringed military cloaks and purple Greek cloaks
made for service in the army camp. Also Bardaean hooded
shoulder capes and the toga and arms of a gladiator finished in
gold and jewels’ (SHA Pertinax, 8.2-4).
The thieves in the novel The Golden Ass steal silk clothes woven with gold
thread as well as coins and plate (4.7). Such silk and gold clothes were the
extreme version of luxury clothing, combining two expensive materials,
and would have been restricted to the very richest strata of society,
including the Imperial family. In AD 169, Marcus Aurelius even raised
money to pay for a war by selling ‘his wife’s silk and gold clothes’ (SHA
Marcus Aurelius, 17.4).
Some Emperors also wore jewel-sewn clothes. The Emperor Severus
Alexander was praised because ‘he removed from the imperial footwear
and clothes all the jewels that had been used by Elagabalus, and he wore a
plain white robe without any gold … and ordinary clothes and togas’ (SHA
Severus Alexander, 4). Elagabalus had worn clothes such as ‘a tunic made
wholly from cloth of gold, or one made from purple, or a Persian one
studded with jewels … He even wore jewels on his shoes, sometimes
engraved ones – a practice which aroused the derision of all’ (SHA
Elagabalus, 23.3-4). Caligula had worn capes sewn with jewels
(Suetonius, Caligula, 52), and Agrippina, the wife of Claudius, had a
military cloak of gold cloth (Pliny, Natural Histories, 33.19.63).
From the first to third centuries, such extravagant clothing, even for
the Imperial family, was castigated in no uncertain terms. Unpopular
Emperors may have been described as wearing such clothes, whether they
ever had or not, as an easy way to discredit them. During the fourth
century, however, there was a major change in attitude to conspicuous
status-display in costume, as well as an increased interest in the use of
pattern and decoration adopted from the east, and such rich clothing
became acceptable for the Emperor. Claudian describes the dress of the
Emperor Honorius on the occasion of being made consul for the fourth
time (AD 398):
Jewels of India stud your clothes, rows of emeralds make the
threads green; there gleams the amethyst and the glint of
Spanish gold makes the dark-blue sapphire look duller with its
hidden fire. Nor in the wearing of such a robe was unadorned
beauty enough; the work of the needle increases its value,
pictures traced in precious metals enliven the work; many a
jasper adorns the ornament and pearls twisted in varied
patterns... Who could sew precious stones on purple? Who
mingled the fire of the Red Sea and of Tyre? Tyre lent her dyes,
China her silks and Hydaspes his jewels (On the Fourth
Consulship of Honorius, 585-92, 599-600).
This elaboration in dress is an element of the emerging tendency for the
Emperor to stress his ‘apartness’ from the majority of people by the use of
costume, titles, regalia and a hierarchical court. Diadems, once avoided by
Emperors because of their association with kings, became commonplace
for fourth- and fifth-century Emperors, and eventually developed into true
crowns. The costume of both Justinian and Theodora in the San Vitale
mosaics show quite clearly how different their dress is from those of their
courtiers, from pearl-hung head-dresses to decorated shoes (colour plates
1-2).
The rich did, however, have to suffer for their fashion. Gold thread
could make the clothes feel uncomfortable against the skin. The Emperor
Severus Alexander:
was always eager to get good linen, without any purple in it, for
he used to say “if these garments are made of linen in order to
prevent their being rough, what is the use of having purple
[wool] in it?” And as for inserting gold threads, he thought it
was madness, since in addition to being rough they also made
the garment stiff (SHA Severus Alexander, 40.10-1).
Gold also added extra weight to the cloth. Claudian refers to a consul’s
toga ‘stiff and heavy with gold’, and refers to a man ‘straining his sapless
limbs beneath the weight of the consul’s toga, borne down by the wearing
of it (On the Consulship of Stilicho, 340; Against Eutropius, 299-304).

SLAVE CLOTHES

The term ‘slave’ covers everything from those with a considerable fortune
in their own right to manacled farm-workers treated little better than
animals. Households could have one general-purpose slave, or an army of
several hundred, each of whom had their own strict role. The quality of
clothing given to slaves would therefore vary greatly, to reflect the status
of the person involved. Slaves invested with power, such as farm bailiffs
or town-house stewards, and those slaves who dealt directly with the
family, would have been given better quality clothes than behind-the-
scenes slaves and farm workers. The majority would have worn clothes of
a similar quality to those of poor free people, and in the Price Edict of
Diocletian of AD 301, the lowest quality of clothing is always described as
suitable for ‘common people or slaves’.
Most slaves would own few clothes. In the second century BC, Cato
listed the suggested clothing allowance for male farm slaves: ‘a tunic
three and half feet long, and a thick cloak (sagum) every other year. When
you issue the tunic or cloak, first take up the old one and have patchwork
made of it’ (On Farming, 59). Columella suggests that the clothes for farm
slaves should be chosen for their ‘usefulness rather than appearance,
taking care to keep the [slaves] fortified against wind, cold, and rain, all of
which are warded off with long-sleeved leather garments, garments of
patchwork, or thick shoulder capes’ (On Farming, 1.8.9).
SEASONAL CLOTHING

For those who could afford it, there could be a distinction between winter
and summer clothing. Long-sleeved tunics or multiple tunics were worn in
winter, while Juvenal mentions ‘keeping off the East wind with skins
turned inwards’ (Satires, 14.185-8). A will from Egypt refers to bequests
of ‘clothes, both winter and summer’ (Oxyrhynchus Papyri, 16, 1901). A
Jewish man had to provide his wife with new clothes in the winter, and she
could then ‘cover herself with the worn-out ones in summer’ (The
Mishnah, Ketuboth, 5:8).

CLEANING AND MENDING

Professional fullers, who prepared newly woven cloth by washing it, and
raising and cutting the nap, would also clean and re-dye garments that had
faded. Because of the cost, clothes would be patched and mended by
everyone other than the very rich, and in the later period the decorative
roundels and stripes were often removed from one tunic for use on another.
Pliny refers to a certain type of wool in Egypt ‘which is used for darning
clothes worn by use and making them last again for a long period’
(Natural Histories, 8.73.191), but when the clothes became too worn or
torn, they were cut up and the better pieces used to make patchwork
clothes for the poor or slaves. A number of rags cut into squares or
rectangles about ten by fifteen centimetres in size that have been found at
Mons Claudianus in Egypt have been suggested as possible pieces
prepared for patchwork, while a well-made and carefully patched tunic
was made of re-used pieces of mantles, without taking into account the
final position of the woven decorations on the original garments
(Mannering 2000, 283, 286).
4 Clothes press. First-century wall painting from a fuller’s shop in Pompeii.

STORAGE

All clothes get creased when worn, and linen in particular creases very
easily, but it is not clear how the Romans dealt with this. Fullers, when
finishing off new cloth, would sprinkle it with water by taking a mouthful
of water, and spraying it over the cloth, and would then rub it smooth. It is
possible that clothes were ironed in the same way, perhaps using large flat
stones similar to the large glass discs used in the Viking period. It is not
known if such smoothing was also done in domestic surroundings on a
more regular basis. Clothes were also compressed in large clothes presses,
between boards that could be screwed down (4). The simple design of
most Roman garments meant that they could at least be folded away
easily, and most were stored in chests (Croom 2007, 140).

BAGS AND POUCHES

Roman clothing did not have pockets in the modern way. Objects were
carried in folds of the clothes themselves, in small pouches or in shoulder
bags, while the rich had an accompanying slave to carry anything
necessary. Aulus Gellius mentions a scroll carried in the overfold (sinus)
of a toga, and Suetonius a lover’s slipper between toga and tunic (5.18.9;
Vitellius 2). Martial observes one of the dangers of this method of carrying
items when he describes a key dropping out of the folds of a garment at an
embarrassing moment (Epigrams, 5.35.7).
Money was kept in a pouch (marsuppium) that could be held in the
hand or hung from a belt, although this left it in danger of thieves. In
Plautus’ play Epidicus, a character has ‘a good sharp knife to disembowel
the old man’s pouch’ (2.2.183). A larger pouch (crumina) was usually
worn round the neck. In The Comedy of Asses, the character Leonida takes
on a role as a slave: ‘I’ll act as porter myself; as for you, you walk on
ahead as a master should, empty-handed ... Why don’t you hand the bag
over and let it crush my shoulder?’ (657-663). In scenes of processions to
bath-houses, the family members go empty-handed, accompanied by
slaves carrying the items they needed, including in one case a slave with a
shoulder bag slung across her shoulder (colour plate 8).

COLOUR

Although modern images of Romans are influenced by the hundreds of

surviving statues in white marble, or by modern paintings of men in togas,
the Romans were in fact lovers of bright colours, and used them whenever
they could (Sebesta 1994a, 70). However, the ancient authors are
frequently scornful of people wearing too bright a colour, which was often
seen as ‘un-Roman’; the Romans had a wonderful nostalgia for a past
where men were austere and women modest, and no-one had yet been
corrupted by luxury. Juvenal comments that ‘it is purple clothing,
whatever it be, foreign and unknown to us, that leads to crime and
wickedness’ (Satires, 14.185-8), and Martial describes a man who ‘loves
sad-coloured cloaks and Baetic wool and ash-grey colours, who thinks that
scarlet clothes are unmanly, and calls amethyst clothes the dress of
women. He praised natural shades, yet although he is always in dark
colours, his morals sport a different colour’ (Epigrams, 1.96.4-7). The
early Christians were just as bad; Tertullian wrote that if God had wanted
humans to wear coloured clothing, he would have made sheep with blue
wool (Croom 2005, 120). Colour was much more important in the past for
showing status than it is now. In fact, the colour and quality of cloth has
always been important in indicating status until extremely recently, and
during the medieval period there were sumptuary laws forbidding people
of certain classes to wear particular colours or cloths.
Wool comes in a number of natural shades, often kept as the colour of
the finished garment. Pliny mentions white fleece from Italy, black fleece
from Spain, golden-red from Asia, tawny from Conossa and a dark-greyish
shade from Tarento (Natural Histories, 8.73.191). White wool could, of
course, be dyed different colours. Animal thread, like wool, takes dye
colours very well, while plant thread, like linen, takes it very poorly. Silk
and cotton lie somewhere in between. Although the Romans could dye
linen, it seems usually to have been left undyed, or was bleached white. In
coloured images of clothing, garments shown as white could therefore be
either of wool or linen, while coloured clothes are almost certainly
woollen. A pound-weight of wool needs a pound of dye-stuff to dye it; for
dark shades, the wool needs to be dyed several times, and there are
references to cloths described as ‘double-dyed’ (eg Horace, Odes, 2.16.36;
Martial, Epigrams, 4.4.6). It was therefore expensive to produce rich, deep
colours, so dying clothes was very much a symbol of wealth. Poor people
wore blacks, browns, greys and creams which were, in many cases, the
natural colour of wool (colour plate 3). These colours were also worn by
mourners or people awaiting trial, as a sign that they did not care about
their appearance at such times. Publius Scipio Africanus surprised people
because he did not follow the usual traditions while awaiting trial, for:
‘although he was under accusation he neither ceased to shave his beard and
to wear brilliant-white clothes nor appeared in the usual dress of those
under accusation’ (Aulus Gellius, 3.4.1). A number of colours held
particular significance.

Purple
Purple was one of the most popular colours during the Roman period, and
very early came to represent wealth and decadence. Later it became
symbolic of the Emperor himself. The most famous purple dye was
produced by shellfish from Tyre, but other purples came from a wide
number of sources. Shellfish off the coast of Africa also produced a purple
dye, while Pliny mentions that ‘Transalpine Gaul [could] produce with
vegetable dyes Tyrian purple, oyster purple and all other colours’,
including whortleberry, which was used ‘to supply purple dye for slaves’
clothes’ (Natural Histories, 22.3.3; 16.31.77). Diocletian’s Edict of AD
301 refers to simple or once-dyed purple, best genuine Milesian purple,
second quality Milesian purple, archil-lichen purple (of four different
qualities), light purple and bright Tyrian purple (Graser 1975, section 29).
These dyes would have produced different shades of purple. Pliny
refers to the changing fashions in colour during his own times, from violet
(violaceus), to Tarentum red (rubra), to double-dyed Tyrian purple
(Natural Histories, 9.63.137), and he uses two separate words for purple
itself; ‘robes of shellfish purple (conchylia) and purple (purpura) are worn
away by every hour of use’ (9.60.124). Nero banned the use of amethyst
(amethystinus) and purple (Tyrii color), and closed down the dealers
(Suetonius, Nero 32). Ancient authors refer to people wearing cloaks,
mantles and tunics of purple, and from the third century almost all men’s
tunics were decorated with purple and white motifs. Diocletian’s Edict
refers to tunics, tunics with hoods, women’s shawls and face cloths, all
with purple stripes.
The most expensive purple was that from Tyre. Ovid commented,
‘What madness to carry whole incomes on one’s body’ (Art of Love, 3.171-
2), and it can still be bought today at huge expense: 10,000 shellfish are
required for one gram of the dyestuff (information from A. P. Fitzpatrick
Fine Art Materials). The demand for it was huge, and Pliny laments that
from being a famous nation state, ‘the entire renown of Tyre now consists
in a shellfish and a purple dye’ (Natural Histories, 5.17.76). Although it
produced a rich purple colour, the dye was equally famous for its terrible
smell, so that while wearers might look impressive, people could smell
them coming from a long way off. Martial refers to a woman who wore
purple clothes day and night, so that the smell of the dye concealed her
own body odour (Epigrams, 9.62.1-4; see also 2.16.3, 4.4.6).
For much of the time, anyone who could afford purple cloth could
buy it, but once in a while Emperors tried to restrict certain shades of dye
and all-purple clothes to their own use. When Nero saw a matron wearing
purple at one of his recitals, during the period when he had banned anyone
other than himself from wearing it, he had her dragged out ‘and stripped
on the spot, not only of her garment, but also her property’ (Suetonius,
Nero, 32). In the late Roman period, a concept of ‘imperial robes’
developed, whereby certain designs of clothes, probably made of silk dyed
with Tyrian purple, could only be worn by the Emperor (Reinhold 1970,
62-3). Hence, in the mid-fourth century a man was ruined because one of
his guests had pointed out how he could use the wide purple edges of the
table- and couch-covers to make a purple robe, as worn only by the
Emperor, and the host was thus suspected of having Imperial ambitions
(Ammianus Marcellinus, Roman History, 26.8.8). Government dying and
weaving workshops were set up in Tyre, and laws were repeatedly passed
to restrict certain grades of purple clothes to Imperial use only. Possession
of ‘Imperial’ garments was considered to be high treason, and could, and
did, result in death (Reinhold 1970, 64).
It is clear, however, that throughout the Empire, it was only ever
certain shades of purple, or forms of purple clothes, that were forbidden,
and that there was never a blanket ban on all purple. Some Emperors
realised that purple clothes did not make an Emperor, and the Emperor
Julian not only dismissed the case of a man who had had imperial robes
made for himself, but even sent him some purple shoes to show that he did
not feel threatened by purple clothes (Ammianus Marcellinus, Roman
History, 22.9.11). The Byzantine Emperor Leo later declared:
I do not know for what reason former Emperors, who were all
clad in purple, established the rule that nothing of this kind
should be sold and did not even allow anyone to either purchase
or sell stuffs of this colour … For what evil could result, even if
everyone wore purple, and the distinction of Majesty was
thereby, to a certain extent, impaired? … We hereby decree that
the sale of small fragments and scraps of purple cloth, which
afford means of ostentation to our subjects, shall not be
prohibited (Constitution, 80).

Scarlet
Although some translations freely use purple and scarlet interchangeably,
it is clear that the Romans considered them to be separate colours, even if
scarlet was considered to be almost as decadent as purple itself. Martial
refers to both scarlet and purple when talking about extravagant clothing
(Epigrams, 4.28.1, 5.23.5-6), and it is clear it also had imperial
significance, no doubt because of its cost. Fronto, writing to Marcus
Aurelius, refers to him and his adoptive father as being ‘bound to wear
purple and scarlet’, while the Emperor Commodus gave the title of Caesar
to Albinus, and wrote: ‘[so] that you may receive some definite symbol of
an Emperor’s majesty, I authorize you to wear both at the present time and
at my court the scarlet mantle’ (SHA Clodius Albinus 2.5). Coccum,
considered to be a berry by the Romans, but now known to have been the
insect kermes, was used in the east to produce the scarlet colour for
generals’ cloaks (Pliny, Natural Histories, 22.3.3; Sebesta 1994a, 69, 71).
Perhaps because of this use, red was the colour traditionally associated
with the military, although the evidence suggests that the soldiers
themselves more often wore white tunics and a yellow-brown cloak
(Sumner 2009, 115).

White
The word for ‘candidate’ (candidatus) came from the word candidus,
meaning the dazzling snow-white toga which a politician would wear to
stress his purity. Such white clothing was worn whenever the wearer
wanted to impress, as well as at ceremonies and holidays such as weddings
and birthday meals (Horace, Satires, 2.2.60-2; Claudian, Epithalamium,
295). Another word, albus, was used for a less brilliant white. As well as
bleaching, fullers used both rootlet juice (possibly the plant called dyer’s
rocket) and sulphur to make woollen cloth both brilliant-white and
particularly soft (Pliny, Natural Histories, 19.18.48; 35.50.175). Cloth was
draped over a frame and the sulphur burnt underneath it, although this
method did have a habit of turning the cloth yellow after a while, and
could hardly have smelt much better than the shellfish purple.

Black
There were also two terms for black: ater, a dull matt black and niger, a
glossy black. Blacks, greys and dark colours generally were the colour of
mourning and misfortune. The fifth-century writer Prudentius describes
people trying to show remorse to an angry Christ: ‘the matron, taking off
her necklaces, puts on dark clothes, and instead of jewels and silk, covers
her flowing hair with foul ashes. The fathers wear the dark clothes of
mourning, all unbelted; the common crowd put on coarse haircloth in
lamentation; the maids, with unkempt hair shaggy like beasts, cover their
face with black veils. The king himself, pulling away the brooch, tore his
cloak of Coan purple into pieces, and took off his bright jewels’ (The
Daily Round, 7).
PRICES

Clothing in the Roman world and, indeed, throughout history, up until the
second half of the twentieth century, was expensive in a way difficult to
understand now. Producing enough thread to make even a simple tunic
using nothing more than a drop spindle was a major undertaking, and the
finished garments were correspondingly expensive. Clothing was seen as
something of value, and as such is found mentioned in marriage
settlements, contracts and wills, some numbers of which have been found
in Egypt. The sixth-century will of Falvius Pousi, for example, includes a
number of clothes:
I also wish the said mistress Manna to have the third part of all
my clothes, both winter and summer, and the other two-thirds of
the said clothes I wish to be given to Sambas and Iulianus ...
because of their devotion to me. [To] Cyria ... the articles of
female clothing found in my house ... [To] Manna the third share
of my clothes, both winter and summer (Oxyrhynchus Papyri 16,
1901).
A marriage settlement dated AD 260 describes a dowry made up solely of
gold jewellery and clothing, consisting of ‘a silvery striped Dalmatian veil
worth 260 drachmae, a white, single, tasselled striped frock worth 160
drachmae, a turquoise-coloured Dalmatian veil worth 100 drachmae,
another white Dalmatian veil with a purple border worth 100 drachmae’
(Oxyrhynchus Papyri 10, 1273). A first-century apprenticeship contract
with a weaver includes clothing to be given as payment: ‘the boy being
maintained and clothed by the teacher Seuthes ... after the period he is to
give the boy on his departure a tunic worth 12 drachmae, or the 12
drachmae themselves’ (Oxyrhynchus Papyri 41, 2971).
Due to the cost of items, there would have been a flourishing second-
hand market, and clothes were therefore a prime target for thieves. The
poet Tibullus talks about walking through Rome after dark, afraid of
someone attacking him with a weapon or of being mugged, not for any
money he might be carrying, but for the clothes on his back (1.2.26), and
in the story of the Good Samaritan the robbed man is left naked because
the thieves have taken his clothes (Luke, 10.30). Baths were a favoured
location for thieves taking garments unguarded by their owners. Some of
the curse tablets from Bath refer to stolen clothing. One Docilianus had
lost his hooded cape:
I curse him who has stolen my hooded cape, whether man or
woman, whether slave or free, that … the goddess Sulis inflict
death upon [him] … and not allow him sleep or children now
and in the future, until he has brought my hooded cape to the
temple of her divinity (Tomlin 1988, no. 10).
Richer clients would therefore take a slave to the baths with them to watch
over their clothes. Clothes could also be hired (Juvenal, Satires, 6.352).

Diocletian’s Edict
Of particular interest is the Edict of the Emperor Diocletian, published in
AD 301 (Graser 1975). This lists the maximum price that could be charged
for a large number of items, from a single needle to a donkey-load of
firewood, and including many different types of clothes. The price given is
always the maximum price, and it is not clear how strictly or for how long
the Edict was followed, but it gives an insight into comparative prices at
the time it was written. In the section on clothes, a range of prices is given
for certain items, according to where the garment was made and its
quality. Therefore, a woman’s unmarked (undecorated?) dalmatic tunic ‘of
the first quality’ could cost up to 11,000 denarii if it came from
Scythopolis, but as little as 7,000 if it came from Tarsus. The same 7,000
denarii would only buy a third quality tunic from Scythopolis. Mantles
range from 7,500 denarii for first quality from Scythopolis through second
quality, third quality and three different levels of ‘inferior to third quality’
to ‘coarse linen for common people or slaves’, the cheapest of which cost
only 500 denarii, one fifteenth of the price of the best.
The prices for clothes range from 55,000 denarii for a woman’s shawl
(mafortia) decorated with vertical stripes using one pound of purple dye,
to 500 denarii for a third quality slave’s coarse linen tunic or mantle, and
200 denarii for a third quality slave’s loin cloth. These prices can be
compared to the wages of certain craftsmen mentioned in the Edict. A
farm worker earned 25 denarii, and a carpenter 50 denarii per day. A
barber was paid two denarii per person, while a teacher of public speaking
earned 250 denarii per pupil per month. The prices can be also be
compared with other objects in the list. The sum of 3,000 denarii would
buy a Numidian hooded cape, a woman’s third quality dalmatic tunic from
Tarsus, a first quality face-cloth from Tarsus – or a four-wheeled carriage
(minus the ironwork).

MALE AND FEMALE CLOTHING

There were distinct clothes for men and women, best summed up in
Roman law:
Clothing is either intended for the use of men, women, or
children, or is common to both sexes, or is used by slaves. That
peculiar to men is such as is designed for the use of the head of
the household, for instance togas, male tunics, small mantles,
shaggy coverlets, bed coverings, coarse wool cloaks, and other
things of this description. Garments peculiar to children are such
as are used for no other purpose, as for example, the toga
praetexta, short tunics, Greek-style cloaks, and mantles, such as
we purchase for our offspring. Women’s clothing is that intended
for the use of the mother of the family, and which a man cannot
readily wear without censure: as for example, stolae, mantles,
female tunics, caps, belts, turbans which are designed rather to
protect the head than for the purpose of ornament, coverings, and
travelling capes. Those are common to both sexes which both
men and women use indiscriminately, such as travelling capes
and mantles and other garments of this kind, which either a man
or his wife can wear without rendering themselves liable to
unfavourable comment. The garments of slaves are such as are
intended to clothe them, for example, coarse wool cloaks, tunics,
travelling capes, linen clothes, bed coverings, and other articles
of this description (Digest 34.2.23).
It is clear from this description that men could sometimes wear women’s
clothes, but they were generally considered effeminate and derided if they
did so. Women could not, of course, wear men’s clothing. It should also be
noticed that coverlets and bed clothes were considered as clothing by the
Romans. Cloaks, after all, were often no more than a large rectangle of
thick woollen cloth that could easily be used as a blanket. In a papyrus
document from Egypt that records an order for various textiles, four
cloaks are the same size as a blanket mentioned in the same list (2.66m by
1.77m), although not as heavy in weight (Sumner 2009, 81). The poet
Martial describes a poor man’s room as containing ‘a sad hearth
unwarmed by fire, and a mat, and a bug, and a bare bed-frame, and a short
toga worn day and night’ and mentions an old man who used his mantle as
a blanket (Epigrams, 11.56.4-7; 4.53.5).
3
 MALE CLOTHING

Male costume will be studied garment by garment, starting with the tunic
and continuing with the toga, outer coverings and leg coverings, including
shoes. These, along with hairstyles, will be discussed with their changes
over time. Other items, which either changed little or which lack good
evidence for major changes, such as religious clothing, will also be
considered.

THE TUNIC
The tunic was the most basic item of male clothing, worn by everyone
from slaves to Emperors. The length of the tunic could vary, but it
generally came to the knees, and could therefore be worn by itself without
any form of leg covering, a point which distinguishes it from any later
concept of ‘shirt’. In modern western society trousers are the main item of
male clothing, so that while it is acceptable in some contexts for a man
wishing to cool down to take off his shirt, he would get strange looks if he
kept his shirt on and took his trousers off.
In its simplest form, a tunic is two rectangles sewn together, or a
rectangle of cloth folded over and sewn to form a tube, with slits left for
head and arms (3.1). If the body of the tunic is baggy, it will be too wide
for the shoulders, and will naturally form short sleeves (5.1). Very often,
however, the width of the cloth was widened towards the armholes, to
make the sleeves slightly longer without making the body of the tunic
wider. True sleeves could also be woven or sewn onto the tunic.

CONSTRUCTION

Some first-century statues show probable seam lines running down the
arm, from the neck to the end of the sleeve, suggesting the tunic was made
from two rectangles of cloth sewn together or, just possibly, from a
rectangle of cloth folded in half vertically and sewn along the shoulders
(3.1-2). Where possible, the elements of the tunic were woven to shape on
the loom (Granger-Taylor 2000, figs 13-4). The weaving of many Egyptian
tunics was done ‘sidewards’, starting from the wrist-end of one sleeve.
When this section was finished, extra warps had to be added to either side
so that the next section of cloth, representing the full length of the tunic,
front and back, could be woven, with a slit left for the neck. The weaving
was then reduced in width again to form the second sleeve. When taken
from the loom, it was simply sewn up under the sleeves and down the
sides (3.3). A tunic woven in this manner would have needed huge looms,
up to 2.7 m (9 ft) wide.
5 Tunics of the first and second centuries. 1 Belted tunic, second-century bas relief re-used on
the Arch of Constantine, Rome; 2 Tunic with stripe, wall painting from Pompeii; 3 Unbelted tunic
worn by a knife-seller’s assistant, bas relief from Rome; 4 Unbelted tunic worn by man in tavern,
wall painting from Pompeii.
Another form of tunic weaving required a much narrower loom; the
weaving was again ‘sidewards’, but without the sleeves, and instead of the
front and back of the tunic being connected along the short sides, they
were joined along a long side, so the loom had only to be as wide as the
final length of the tunic (a typical width was approximately 1.5 m: Carroll
1998, 34). The sleeves were woven separately, sewn up, and added to the
body, which was then sewn up on one side and along the shoulders (3.4).
Any form of decoration, from simple stripes to complicated motifs, was
usually woven in while the cloth was on the loom, and not added later.

FIRST AND SECOND CENTURIES

Belted tunics
The most popular colour during this period was bleached white, with the
only form of decoration consisting of simple vertical stripes down front
and back (5.2). Originally, these were restricted to certain classes of
people: a wide stripe (laticlavus) for senators and their adult sons, and a
narrower stripe (angusticlavus) for knights (Cleland et al 2007, 35). The
senators and the knights (also called equestrians) were two social ranks
which had a monetary qualification; in the first century, a senator had to
have at least 1,000,000 sesterces, and a knight 400,000 sesterces to
become members of the orders. The width of the stripes was the important
element of this social display, as Roman art shows that men of all social
statuses wore tunics with stripes (5.4; Coarelli 2002, 146, 147). However,
a study of the width of 240 surviving stripes on tunics found at a quarry
site in Egypt, where there would have been no senators or knights present,
revealed a wide range of widths, from 1-4 cm wide, to over 7 cm wide.
The majority were between 1 cm and 4 cm wide, and most were purple,
although examples in blue, red, brown, green and multi-coloured examples
were also found (Mannering 2000, table 2; 285).
The tunic was fastened at the waist or hips with a belt, although
usually the details of the belt are obscured by the folds of the tunic.
Leather belts with copper alloy buckles had long ends that hung down after
passing through the buckle loop, rather than being held in loops as in
modern belts. These long, dangling ends are rarely visible on images of
civilians, and it may be that such leather belts were restricted to the
military. The belt was a particularly important item for soldiers, and there
were a number of forms that were only worn by them. Even when dressed
in civilian clothing, their characteristic belts made soldiers recognisable.
Disgraced soldiers could be deprived of their belts: Augustus is recorded
as punishing centurions by making them stand all day outside the general’s
tent ‘in unbelted tunics’ (Suetonius, Augustus, 24) Male civilians,
therefore, may have used cord belts without metal fittings, similar to the
female version.
During this period, the tunics were generally very wide and without
sleeves, as can be seen from a number of surviving examples, but the great
width of the tunic usually created the effect of elbow-length sleeves
(Granger-Taylor 2000, colour plate 2, no. 4; Sumner 2009, fig. 14; fig. 26).
A surviving tunic from Nubia in Egypt was 1.40 m wide and 1.27 m long;
another from Khirbet Qazone, Jordan was very similar in size, 1.42 m
wide and 1.17 m long (ibid., fig. 14). Some long-sleeved tunics were
known at this period; Pliny the Younger commented that: ‘in winter [his
uncle’s] hands were protected by long sleeves, so that even bitter weather
should not rob him of a working hour’ (Letters, 3.5.15). Thick tunics could
also be worn in winter as protection against the cold, or else several layers
of tunic: the Emperor Augustus ‘protected himself with four tunics and a
heavy toga, besides an undertunic, a woollen chest protector and wraps for
his thighs and shins’ (Suetonius, Augustus, 82). The length of the tunic
depended to a certain extent on whether it was belted, and how much it
was pouched up over the belt, but the most usual length was approximately
down to the knees, although soldiers wore shorter tunics. The upper
classes would not have appeared in public wearing their tunics unbelted,
unless attending a funeral or a court case, or some other occasion when
their disturbed emotional state could be suitably reflected in a lack of
attention to their dress; Suetonius called the Emperor Nero ‘utterly
shameless’, not only because he appeared in public wearing dining-
clothes, but because he appeared ‘with a handkerchief bound round his
neck, unbelted and without shoes’ (Nero, 6.51). Inside the privacy of their
own home, or during such times as the Saturnalia festival, when there was
a general laxity in conventions, men could wear their tunic unbelted if
they so wished.

Unbelted tunics
The lower classes are often shown wearing unbelted tunics during this
period. In Pompeii, first-century painted scenes not only show men
indoors, such as drinkers and game-players in taverns, wearing unbelted
tunics, but also men out in the street, such as men buying bread, or looking
after stalls in the forum (Nappo 1998, fig. 46). Stone carvings from Rome
and Ostia show a knife-seller’s assistant, furnishing salesmen and people
at a grocer’s shop in unbelted tunics (5.3). Wall paintings show that these
tunics were often dark in colour, for example brown or yellow-orange,
perhaps reflecting the unbleached, natural colours of wool, but even so
they could be decorated; a quarrelling game-player in a Pompeian wall
painting has two thin, dark stripes down his tunic (5.4).

THIRD AND FOURTH CENTURIES

In the late second or early third century, a new form of tunic was
introduced. It had tight-fitting sleeves down to the wrist, and often
decoration that developed, over time, into multi-coloured, elaborate
tapestry-woven bands and roundels (6.1). These decorative elements on
tunic front, back and upper shoulder were not only circular, but could be
square, rectangular, or star-shaped; they will all be called roundels
hereafter for convenience. As the sleeves were often woven separately, the
bodies of the tunics were also often narrower than before (when the width
had helped form the sleeves).

Belted tunics
In the first half of the third century, decoration was restricted to a number
of half lines over the shoulders ending in decorative terminals such as
arrowheads. Fragments of tunics from the Syrian city of Dura-Europos,
destroyed in AD 256, show only these simple forms of decoration,
suggesting that in Syria at least, more elaborate decoration did not come
into general use until the second half of the century (Pfister and Bellinger
1945, fig. 1, nos 3-14). A fragment of a tunic from the city of Palmyra has
a simple purple shoulder roundel with a white interlace pattern (7.1;
Trilling 1982, fig. 1). Such decorations were almost always made of purple
wool, even if the body of the tunic was linen, as wool takes dyes better and
gives brighter colours. Although embroidery was known in the Roman
world (for an example, see Bender Jørgensen 2000, fig. 8), almost all of
the decoration, however complex, was woven, rather than added by needle
afterwards.
An estimated 20,000 textiles of all types have been recovered from
Egypt, many of them being the decorative elements cut away from
‘boring’ plain cloth (Carroll 1988, 1). Very few are securely dated, but it is
usually considered that the decorations started simple, and grew more and
more complex over the centuries (7.2). This, however, does not take into
account the fact that good and mediocre weavers can be working at the
same time, or that they could be producing clothes of different quality, and
for different prices. The dating of these tunics must remain uncertain until
more securely dated pieces can be recovered.

6 Tunics of the third and fourth centuries. 1 Man carrying belt, wall painting, Silistra, Bulgaria;
2 Man carrying candlestick and wreath, wall painting from the Tomb of Aelia Arisuth, Gargaresh,
near Tripoli; 3 Unbelted tunics worn by men receiving largesse from the Emperor, bas relief from
the Arch of Constantine, Rome.
By the first half of the fourth century, when the mosaics in the
Sicilian villa at Piazza Armerina were being laid, the decoration on tunics
included square or circular roundels on the upper arm, and on the lower
front and back of the tunic. The hunters in the Piazza Armerina mosaics
show a range of decorations; soldiers in white tunics have two or more
bands round the wrists, full length stripes with roundels on arm and body,
short stripes with or without roundels, and tunics with an inverted U-
shaped band under the neck, a U-shaped band following the hem at the
bottom, and the slits up the side of the tunic. These slits were rarely
necessary in more baggy tunics, and during the first and second centuries
were generally only seen on barbarian tunics. From the fourth century, the
slits became more popular. A narrow, knee-length tunic had the advantage
of using less cloth than a baggy one, but it could easily restrict a man’s
stride, particularly if he was active, as when hunting. Tunics with short
slits up either side kept the narrow look, but avoided hampering the
wearer. Alternatively, the tunic could be hitched up at the belt at the point
above each leg, a style of wearing a tunic most frequently seen in images
of hunters.
7 Tunic decorations in purple and white. 1 Tunic from Palmyra, Syria (restored), third century; 2
Tunic from Egypt, possibly sixth century.

Unbelted tunics
The lower classes are still shown wearing unbelted tunics in scenes such as
Constantine addressing the public on his Arch in Rome (6.3). From the
late third century, unbelted tunics became acceptable for other classes:
fourth-century catacomb paintings show a number of men wearing
unbelted, mid-calf length tunics decorated with simple stripes (for
example Ferrua 1991, figs 67-8). More elaborately decorated tunics were
also worn unbelted (6.2). In the Piazza Armerina mosaics, both soldiers
and civilians are shown wearing unbelted tunics with decoration very
similar in that seen on the belted tunics. The civilian hunters have
unusually short unbelted tunics, reaching to above the knee, but this may
be a more specialised form specifically for hunters who might need to run
quickly. Generally, the tunics of this period were longer, reaching to below
the knee. Those of the civilian hunters and some of the soldiers, both
belted and unbelted, have a wider range of colours than those of the
soldiers, including red, yellow, orange, blue and green, with decoration in
black, red, green, brown, yellow and white (colour plate 4). Some of the
hunters are clearly wearing undertunics, visible only at the neckline.

Dalmatic
This type of tunic originated in Dalmatia and became common during the
third century, and was distinguished by its long, wide sleeves. It was worn
unbelted, and in combination with the cloak rather than the mantle
(Granger-Taylor 2006, 162-4). It was decorated with stripes on the body
and round the wrists, but could also have roundels on the shoulder and
body (ibid., no. 116; Pritchard and Verhecken-Lammens 2001, figs 3.1,
3.4-5). A full-length version of this tunic was adopted by early Christian
priests (colour plate 1, three figures to the right), and continues to be
worn to the current day.

FIFTH TO SEVENTH CENTURIES

The same style of decorated tunic continued through the following

centuries. Decoration seems to have moved away from geometrical
patterns to a range of human figures, animals and plants. Some tunics were
still just purple and white, but others were more colourful. One tunic,
possibly dating to the sixth or seventh century, has a mermaid and a fish in
a jewel-inlay roundel in brown, cream, black, yellow, light and dark green
and red, while another has four birds in a plant, with a square border of
hooks in blue, red, pink, green and off-white (Carroll 1998, pls 36, 40). A
tunic stripe fragment has a double jewel-inlay pattern framing nude
dancers, a Cupid and a winged dog in pink, black, green, grey, beige and
yellow (ibid., no. 41). By the fifth century, men were also wearing tunics
with all-over decoration. Stilicho, for example, has a tunic with shoulder
roundels, wrist and hem decoration and an all-over pattern of circles (8).
8 Portrait of General Stilicho. Late fourth-century ivory diptych. © Monza Cathedral Treasury.

TUNICS UNDER TOGAS

The tunic under the toga always had short sleeves, and would be white like
its outer garment. Quintilian, when discussing the correct dress for an
orator in the late first century AD, describes this type of tunic:
the speaker who does not have the right to wear the broad stripe
[of the senators] will wear it belted in such a way that the front
edge of the tunic falls a little below his knees, while the rear
edge reaches to the middle of the back of the knee. For below is
for women, above is for centurions. If the purple stripe is worn,
it takes little care to make it hang straight; negligence in this is
noted. It is the fashion of those with the broad stripe to let it
hang a little lower than the belted [tunic] (Institutes of Oratory,
11.3.138-9).
Although Quintilian describes the tunic under a toga as being knee-length,
it is clear from art that it usually reached to mid-calf. A normal tunic,
knee-length when belted, would be calf-length when unbelted, but it is
clear from Quintilian, and from a comment from Tertullian, that the tunic
under a toga was usually worn belted. It would therefore appear that men
needed a special, extra-long tunic to wear under the toga.
9 Tunics. 1 Man wearing two tunics, bas relief from the third-century ‘Annona’ sarcophagus,
Museo Nazionale Romano, Rome; 2 Consul’s decorated tunic, ivory diptych of Basilius, AD 480,
Museo Nazionale del Bargello, Firenze; 3 Boy in tunic worn off the shoulder, bas relief from the
Ara Pacis, Rome; 4 Knotted tunic worn by olive harvester, second-century bas relief, Museo
Arqueológico Provincial, Cordova.
This style of tunic continued for a long while, despite the variations
in the form of the toga over it. Then, in the late third or early fourth
century, a fashion of wearing two tunics under the toga began. The man on
the Annona sarcophagus of the late third century is wearing two tunics,
with the hem of one partially visible below the other (9). By the fourth
century, the two tunics were even more distinct: the inner was mid-calf or
ankle length, and had wrist-length, tight-fitting sleeves. The outer tunic
was slightly shorter, with elbow length, baggy sleeves (10.6).
Two tunics were also worn under the highly decorated togas worn by
consuls in the fifth and sixth centuries. The inner one, generally ankle-
length, often had decoration round the wrist, hem and neckline, but was
otherwise plain. The outer one, mid-calf in length, always had all-over
decoration to complement the toga worn over it, the most commonly
recorded being flower motifs in circles, squares, or diamonds, presumably
woven at the same time as the toga to make it a set. A figured marble
pavement (dated c. AD 330-50) of the consul Junius Bassus shows him
wearing an ornate toga in brown, blue, red and orange, with the outer tunic
in matching brown and blue, and the tight-fitting inner tunic in white.
The cloth used for the tunic often having the same pattern as that of
the toga, it is difficult to say whether the wide band on the right shoulder
really belongs to the toga or the tunic. One example at least shows a
different design on the right-shoulder band, with a figure in a two-horse
chariot instead of the flowers in squares and circles shown on the toga
band (9.2). If this is indeed part of the tunic, this tunic must have a V-
shaped neckline, and the folds of cloth seen at the neck are those of the
inner tunic (9.2). In the sixth century (and very occasionally before), these
folds at the neck are replaced by a decorated circular neckline (12.2).

TUNICS WORN OFF ONE SHOULDER

Farm workers and labourers, including blacksmiths, are often shown in

short tunics with the right arm taken out of its sleeve, leaving it and the
right breast bare (White 1977, pls 13d, 20, 24-5). Sometimes it looks as if
the tunic was designed to have only one armhole, but frequently the empty
right sleeve is visible in the folds near the waist (9.3). In such cases, the
tunic must have been made in the normal way with two sleeves, but with a
very large neck-hole, through which the right arm could be passed. To
wear a tunic with such a large neck-hole with both arms in the sleeves, a
knot would have had to be tied in the cloth at the nape of the neck to stop
it slipping off both shoulders (9.4; Fuentes 1987, figs 1-8).
It is difficult to know just how common this type of tunic was, as it is
most frequently seen in depictions of rural idylls, or worn by gods such as
the blacksmith Vulcanus, or Silvanus, the god of wild places, and there are
numerous images of farm workers wearing normal tunics. However, there
are also scenes of everyday life, including farm-workers and a rope-maker,
showing men wearing the tunic worn off one shoulder, while a number of
first- and second-century depictions show soldiers wearing the knotted
tunic (ibid., figs 4-8). The knot itself requires a lot of cloth, and it is
unlikely that the narrower, less baggy tunics of the third century were
worn in this way.

DINING-CLOTHES

Men would change into a different set of clothing for the main afternoon
meal, but unlike Edwardian evening dress, Roman dining-clothes seem to
have been intended to be less formal and more comfortable than daywear.
They were worn when dining out with friends, or with guests in one’s own
home, but it is not clear if they were worn when men dined in private with
their own family. Martial, however, refers enviously to a man living in the
country who only wore his toga on a couple of days every month, and
whose ‘one and only dining robe set has gone through ten summers’,
suggesting that it was not commonly worn in the relaxed surroundings of
country estates (Epigrams, 4.66). Two words are used to describe dining-
clothes: cenatoria (related to the words for dinner and dining), and
synthesis (used in other circumstances to mean a set of objects). A set
probably consisted of a tunic (perhaps worn unbelted) and a mantle, but it
is not clear exactly how they differed from normal tunics or mantles,
although it is possible that they were generally of the same colour.
Worshippers of the goddess Bona Dea ate meals together, wearing dining-
clothes of white (Acta Fratrum Arvalium, 27 May 218 and 17 May 241).
Martial describes a chest which was full of dining-clothes in numerous
colours (Epigrams, 2.46). They were certainly used as an excuse for
display by some people:
During the course of a single meal you got up 11 times, Zoilus,
and 11 times you changed your dining-clothes, in case sweat
 retained in the damp clothes should cling and a draught harm
your delicate skin. Why don’t I sweat when I am dining with
you, Zoilus? A single set of dining-clothes makes a man very
cool’ (Epigrams, 5.79).
Dining-clothes could be worn in the privacy of the house at times other
than at meal-times, but no respectable man would wear them outside. The
only occasion when this was acceptable was during the five-day festival of
Saturnalia, when the world turned topsy-turvy and all sorts of license were
allowed. Slaves were the equal of their masters, gambling was legal, little
work was done and gifts given; the toga could be set aside for five full
days, and informal dining-clothes could be worn outside the house.
Martial mentions even knights and senators ‘rejoicing in dining-clothes
and the wearing of the cap of liberty’ during the festivities (Epigrams,
1.1.1-2).

UNDERTUNICS

The tunic was originally considered to be the undergarment of the main

item of male dress, the mantle, but in time came to be the main garment
itself. It is not clear whether, when this change came about, it was
considered necessary to have an undergarment for the tunic. In the first
century, there are a few references to an undertunic (subucula) worn under
the usual tunic; Horace, for example, complains about being laughed at ‘if
by chance I have a worn-out undertunic under my brand-new tunic, or if
my toga sits badly’ (Letters, 1.1.95-6), and in the novel The Golden Ass by
Apuleius, a man takes off his ‘outer tunic’ to cover the naked hero after he
is transformed back into a human (11.14). In art, however, there are no
examples of visible undertunics in this period, and if one was commonly
worn (and it may be that it was only worn occasionally for extra warmth)
it was not a garment that should be seen.
An undertunic visible at the neckline was acceptable in the north-west
provinces by the second century. It also appears on portraits from Egypt
from the third century onwards, and seems to have become a feature of
more mainstream fashion from the late third or early fourth century,
although it was still not universal. Where colour is shown, the undertunic
appears to be white, which implies that it was made of linen.
 The long-sleeved inner tunic worn under a short outer tunic was not
an undertunic in the true sense, as it was clearly meant to be seen, and was
usually decorated. The two tunics were worn under the toga, and also seem
to have been worn under the mantle as well, but generally this appears to
have been a fashion for the rich. A number of figures in the sixth-century
Ravenna mosaics, admittedly from scenes depicting biblical stories, are
shown wearing the two tunics (Bustacchini 1988, 108, 110, 114). In these
cases, the two tunics are always of the same colour, and were clearly made
as a set to be worn together. Junius Bassius, however, wears an inner tunic
of white under a multi-coloured tunic, and an actor on a mosaic from
Sousse, Tunisia, wears a cream- or buff-coloured inner tunic under a dark
red tunic (Blancard-Lemée 1996, fig. 165).

THE TOGA
The toga is the most well-known item of Roman clothing, and even during
the Roman period it was considered the national costume. It could be worn
only by a Roman citizen, but it was never everyday wear throughout the
Empire. In the first century, the poets liked to compare the joys of country
life with the horrors of city life, and again and again they observe that in
the country, the toga is rarely worn: Martial points out that in the Italian
countryside, ‘on the odd Ides or Calends you might take out your dusty
toga and give it a shake’ (Epigrams, 4.66.1-4), while Juvenal remarked
that, ‘there are many parts of Italy, to tell the truth, in which no man puts
on a toga until he is dead’ (Satires, 3.172.177-8). The toga started off as a
practical outer garment, but was transformed into a symbol of Roman
citizenship, and became more and more a ceremonial costume. As with
most clothing that has symbolic importance rather than practical use, the
people who had to wear it were not too fond of it. In the early third
century, the writer Tertullian wrote a book in praise of the Greek mantle,
comparing it to the horrors of the toga. The mantle was simple in design,
took little re-arranging, and:
there is no necessity for any craftsman to shape its fold right
from the beginning of the day before … and to commit the
whole formation of the umbo to the care of the forceps; and then,
at dawn, to first shorten with a belt the tunic that would have
 been better if it had been woven to a more moderate length, and
next to examine the turned-back umbo and re-arrange it, if out of
line; of that curved edge out of which the sinus grows, to draw
back the loosened folds from the shoulder and, excluding the
right shoulder, heap it up on the left shoulder, with equal folds
devoted to the back: in this way is a man clothed with a burden.
In short I will ask your consciences in turn, what is your first
thought when wearing a toga? Are you dressed, or laden down?
Are you wearing clothes, or carrying a burden? If you should
answer negatively, I will follow you home and I will see what
you hasten to do immediately you cross the threshold’ (On the
Mantle, 5.1-2).
It is quite clear that in its developed form, the toga was an uncomfortable
and unpleasant garment to wear, but that it continued in use for important
occasions because of its symbolic value as the Roman national costume.
The toga was generally made of white wool, and would have been
very expensive to buy because of the large amount of cloth required; it has
been estimated a first-century toga could require 5 m of cloth about 2.5 m
wide (Cleland et al 2007, 190). When Martial talks about buying some
costly items, the three things he lists are ‘a slave, a brand-new toga or
three or four pounds of silver plate’ (Epigrams, 2.44.1-4). The toga was
worn solely by men; the only women who could wear it were prostitutes.
Boys wore a toga with a purple edging (toga praetexta), exchanging this
for a pure white one when they came of age. High priests and officials
such as consuls and magistrates also wore the toga praetexta, with its
distinguishing broad purple border (5.2). Later versions of the toga worn
by Emperors or high officials could be heavily decorated, and even be pure
purple in colour.
The toga is a development of a simple type of outer garment, the
mantle, which was draped round the body rather than fastened by a brooch.
The left arm was always used to carry some of the drapery, so that the toga
could never give the wearer the freedom of a cloak. Although during the
early Republic, togas were simple enough to be worn during farmwork
such as ploughing, or even when dancing, the later versions were too
voluminous and complex for such activities.
 Because of the amount of cloth involved, it was important to arrange
it with great care, and to keep re-arranging it as necessary. Ovid mentions
artistically arranged loose folds catching the eye (Remedies of Love, 680),
although anyone who spent too much care and attention on arranging the
folds was considered foppish.

REPUBLICAN

The toga started off as the main garment for men and women, worn
without a tunic underneath, but by the second century BC it was being
worn with the tunic, and only by men (Stone 1994, 13). It was similar to
the mantle used by Greek men, but unlike the rectangular mantle, the toga
always seems to have had a curved lower edge (Wilson 1924, fig. 7). The
simplest method of arranging the toga was to drape it over the left
shoulder, with a length of it hanging down the front of the body, and over
the left arm. The cloth was then taken round the back of the body, under
the right arm, and the other end was flung back over the left shoulder. The
left arm remained covered in cloth, and was generally held across the
body, while the right arm was free and unencumbered. The section round
the back of the body could be pulled up to cover the right arm and head, if
necessary. The basic method of draping the toga can be seen on the bronze
statue known as the Arringatore, dated 100-75 BC (Stone 1994, fig. 1.1). It
is mid-calf length, and has only a slight roll of cloth on the section under
the right arm. The same simple form of toga can be seen on statues during
the first century BC, where they often show the cloth brought up to cover
the right arm (10.1).

LATE FIRST CENTURY BC TO EARLY SECOND CENTURY AD

During the late first century BC and early first century AD, the toga
developed into a much larger garment, using 5 m (18 ft) of cloth. It could
no longer be put on without help, thus emphasising the status of the wearer
(Stone 1994, 17). The shape of the cloth also changed from that used for
the early form, so that to get the roughly semi-circular shape necessary, it
now had to be folded in half first (Wilson 1924, fig. 27). As a result, a man
now wore at least two layers of wool (four layers over the left side of the
body, where the toga overlaps) in addition to a woollen tunic. On a hot
summer’s day, the cumbersome toga must have been unpleasant to wear,
and it is hardly surprising that the popularity of the garment for anything
other than ceremonial use declined from the early years of the Empire.
The draping of this form of toga differed slightly from the simpler
Republican version. The end section, draped over the left side of the body,
now hung right down so that it touched the ground. Folding the toga cloth
almost in half meant that the section that came across the body under the
right arm (called the sinus) now had two edges, which hung in two loops to
the knee; the top of this section was still sometimes slightly rolled, to keep
the folds in this area tidy (10.2). The end of the toga was again brought up
and over the left shoulder, with the end hanging down the back of the body.
Usually, the length of cloth that hung down the left side of the body,
originally unseen, was pulled up slightly, and a small loop of it allowed to
hang over the sinus (10.3); this section is called the boss (umbo) because
of its small, circular shape. Due to the huge size of this toga, there was a
lot of cloth to be kept up on the left shoulder. Perhaps because of this,
some of the lower layer of cloth on this shoulder was pulled out over the
top layers, to help weigh them down and keep them in place (10.3).
In the late first century, Quintilian describes the correct dress for an
orator, and even goes into details about the correct way of wearing the
toga:
There are also details of dress which are altered to some extent
by successive changes in fashion. The ancients, for example, did
not have the sinus, and those who followed them had it very
short. Consequently it follows that in view of the fact that their
arms were, like those of the Greeks, covered by the garment,
they must have employed a different form of gesture in the
introductory section of a speech from that which is in use now
… The toga itself, in my opinion, should be round, and cut to fit,
otherwise there are a number of ways in which it may be
unshapely. Its front edge should by preference reach to the
middle of the shins, while the back should be higher … The
sinus is most becoming if it fall to a point a little above the
lower edge of the tunic and should certainly never fall below it.
That bit of the toga that passes like a belt from under the right
shoulder to the left should be neither too tight nor too loose. The
portion of the toga which is the last to be arranged [the umbo,
although the word is not used here] should sit rather low, since it
will sit better thus and be kept in its place. A portion of the tunic
should also be drawn back in order that it may not fall over the
arm when we are pleading, and the sinus should be thrown over
the shoulder, while it will not be unbecoming if the edge be
turned back (Institutes of Oratory, 11.3.138-40).
10 Togas of the first to fourth centuries. 1 Unknown man, statue dated 30-13 BC, Rome; 2
Nonius Balbus, early first-century statue from Herculaneum; 3 Emperor Titus, statue dated AD
71-81, Vatican Museum; 4 Emperor Hadrian, bas relief dated AD 136-8, Arco di Portogallo,
Rome; 5 Unknown man, statue dated to AD 260s, Villa Doria Pamphili, Rome; 6 Unknown man,
fourth-century statue, Palazzo dei Conservatori, Rome.
As was usual, the state of one’s clothes was used to make a statement when
in court. It was important, Quintilian said, that at the start of the speech,
the toga was neat, and that if any of it slipped it should be corrected
immediately. In the middle of the speech, however, it was acceptable to
throw the end that had slipped down back over the shoulder without great
care, or to pluck the toga away from overheated throat and chest. By the
end of the speech, ‘practically everything is becoming; we may stream
with sweat, show signs of fatigue, and let our dress fall in careless disorder
and the toga slip loose from us on every side’ (Institutes of Oratory,
11.2.144-9). When talking to a judge in private (but never, he stresses,
when before the Emperor or a magistrate), he suggests pausing when
called upon to speak: ‘We must rise with deliberation. We shall then, to
make the toga more becoming, and to secure a moment for reflection,
devote a brief space to the arrangement of our toga, or even, if necessary,
to throwing it on afresh’ (ibid., 11.3.156). It is clear from a number of
comments that he expected the folds to the toga to need constant re-
arranging as it had a habit of slipping, which is presumably one of the
reasons it was so hated.

EARLY SECOND CENTURY TO EARLY THIRD CENTURY

During the second century, there were minor changes to the toga, with the
overfold being longer, and hanging down closer to calf length than the
first-century knee length. A new form of draping without the need for the
umbo came into use from at least AD 118-9, when it is shown on a
Trajanic state relief (Stone 1994, fig. 1.11). For this toga, the section under
the right arm was brought up higher under the armpit, and the top section
twisted more tightly to form a band, sometimes called the balteus or
‘belt’. The section hanging down the left hand side of the body was again
pulled out over the shoulder to lie over the top layers, as earlier, but now
even more obviously (10.4). It is this feature that develops most
throughout the second and third centuries, with the balteus becoming
increasingly longer and more visible across the front of the body.
11 The ‘Brothers’ sarcophagus, showing the deceased in four different poses. © Museo
Archeologico Nazionale, Napoli.

EARLY THIRD CENTURY TO LATE THIRD CENTURY

In the third century, the balteus was developed one step further. Instead of
letting the cloth form numerous small, uneven folds, it was now
concertina-folded to form a smooth band. The end section draped over the
left front of the body was folded to form a wide band (the end of which is
usually visible between the legs), which continued round the back and
under the right arm. The end section, instead of hanging down in a straight
line from the left shoulder, was brought over towards the right side of the
body and held there while the draping was arranged (colour plate 5).
When the sinus was brought tightly under the right arm, it was then tucked
under the wide band before being brought over the shoulder again. By
tucking the sinus section in firmly, the broad band could be brought down
until it no longer rested on the top of the shoulder but was actually over
the top of the arm. This same band continued round the back of the body,
and was the same section of cloth that was tucked under itself again, so it
sat quite firmly and was in no danger of dropping down the arm. In theory,
pulling the wide band down onto the upper arm should mean there was no
cloth at all on the actual shoulder, but it can clearly be seen in Figure 10.5
that there are indeed some folds of cloth over the shoulder. In practice,
tucking the sinus section in under the wide band produces a fold in the
cloth, forming the non-concertinaed section of the front end which
remains visible at the neck. The first full figure to the left in Figure 11
shows what the band looks like if the lower section of the toga is only
brought three quarters of the way across the body; the second full figure in
11 shows what it looks like if it is brought all the way across to the right
armpit.
Slight variations can have the lower edge of the top fold of the sinus
also concertinaed into a band, which continued over the shoulder and
remained in a wide band as it hung down the back of the body. A number
of statues showing this style of draping have twin parallel lines across
these wide bands, and it may well be that these represent lines of stitching
to keep the concertina folds in place for convenience (10.5; Stone 1994,
34).
During this period, a new style of holding the drapes of the toga came
into fashion. The left arm, already carrying several layers of cloth, and
with a restrictive band across the top of the arm, now also carried the
lower edge of the sinus, forming a pouch across the front of the body (9.1).
This method of holding the drapery often revealed the end of the front
section. The toga was now often shorter than the tunic worn underneath it,
and did not reach the floor as in the earlier styles.

EARLY FOURTH CENTURY

The style of toga with a wide, flat balteus continued into the fourth
century, but sometime in the early fourth century there were further
changes, with the sinus growing so long and baggy that if not held on the
left arm, it would sweep the ground (10.6). Once more, the appearance of
the balteus changed; it was no longer folded into one wide smooth band,
but was arranged in lots of small folds, and was created in a new way. The
start of the third-century balteus can quite clearly be seen on the front of
the body, in front of the right armpit, but in the fourth century it equally
clearly came from under the right arm. To create this effect, the end
section that hung between the legs went up the left side of the body, over
the shoulder, and was brought across the back and under the right arm,
where it was brought diagonally across the body in the balteus, expanding
towards the left to cover both shoulder and upper arm. It continued round
the back and expanded into the sinus, to be draped over the left arm.
Although this style of toga survives into the fifth century, during the fourth
century the formal dress of the Emperor and high officials came to be
based on military dress, and more and more often portraits depict them
wearing the military cloak instead of the toga.

Trabea
The trabea was a ceremonial garment of great antiquity, but in the very
late Roman period the term was used to describe a form of decorated toga
used as the ceremonial dress for Emperors and, in particular, consuls
(Cleland et al 2007, 197). The number of men who wore this form of toga
was therefore very small, and the best examples can be seen on the ivory
diptychs carved for consuls during the fifth and sixth centuries. The
earliest form of this toga was not very different from that of the standard
fourth-century toga, although generally less cloth seems to have been
used, and it looks more tailored, so that the front end section hanging
between the legs is now a simple rectangular band of cloth rather than one
part of a large semi-circle of cloth. In the fourth and fifth century, this
long strip went up the front of the body, over the left shoulder, round the
back and under the left arm; it then grew wider as it passed over the left
shoulder and upper arm. There was a small semi-circle of cloth that sat
rather precariously on the right shoulder, while the rest of the cloth
expands into the wide curve of the sinus, now rather apron-like, which
instead of being tucked under the balteus is simply carried round the front
of the body, and draped over the left arm (12.2).
The late fifth- and sixth-century design was very similar to the above,
but the flap of cloth sitting on the right shoulder had now gone, to be
replaced by a wide strip of decorated cloth running diagonally across the
body to mirror the band of the balteus (12.1). It is possible that this is part
of the trabea, with the band that hung between the legs going over the left
shoulder being taken round the back of the neck and brought over the right
shoulder, then folded back on itself to be taken under the right arm, in a
highly complicated form of draping. It may, however, have been a separate
band attached to the highly decorated outer tunic made out of the same, or
similar, decorated fabric as the toga. It is sometimes thought that these
clothes must have been embroidered, but it is clear from studying
surviving Roman textiles that they rarely employed true embroidery:
patterns were almost always created during the weaving of the cloth.
These togas and tunics were, however, extremely bright and elaborate
cloths, worn only by the very rich: Claudian refers to them as ‘jewel
studded’, ‘shining with gold’, ‘stiff and heavy with gold’, and ‘twice-dyed
in purple’ (On the Consulship of Probinus, 206; On the Consulship of
Stilicho, 3.198, 2.331-4, 339-40). The designs in the cloth are used to
define the different bands within the toga; the lower band hanging between
the legs, the vertical band running vertically up the body, the balteus band
across the chest and the band over the right shoulder are always shown
with a border round their edges (9.2, 12), making it a highly tailored
garment of some complexity. The toga had come a long way from the
simple mantle thrown round the body six hundred years previously.

12 Togas of the fifth and sixth centuries. 1 Consul Boethius, fifth-century ivory diptych, Museo
Romano, Brescia; 2 Unknown consul, fifth-century ivory diptych, Monza Cathedral Treasury.

Sacrifice
The toga was worn mainly for ceremonial occasions, the most common of
which were religious sacrifices. On such occasions it was necessary to
cover the head as a sign of respect, so it was important, with whichever
style of draping worn, that the section round the back of the body and
under the right arm could be pulled up to cover the head (5.2). Perhaps
because of this requirement, the first-century style of toga with the umbo
continued to be shown during sacrifices right up until the fourth century,
as it was one of the easiest styles of toga in which to pull up the folds of
cloth over the head (e.g. Kleiner 1992, fig. 385). Men depicted sacrificing
wearing the toga with the wide band balteus are shown with bare heads, as
there were no suitable folds of loose cloth to pull up over the head (e.g.
Strong 1976, fig. 220). The toga with the umbo is also seen on a few other
occasions in the second and third centuries, including weddings, where
again perhaps an old-fashioned style of toga was worn for the special
occasion. It also figures on a few tombstones, so it is clear that earlier
styles of togas continued to be worn, even when the style-leaders had
brought in new fashions in draping, but this is hardly surprising,
considering the cost of a new toga. Some sarcophagi show men wearing
three different styles at same time. The so-called ‘Brother’s Sarcophagus’
shows the same man in four different poses (11); first he is shown at his
wedding, dressed in a toga with an umbo (second figure from right), then
as a senator with a twisted balteus (early second to early third century;
fifth figure from right), then dressed in the guise of a philosopher wearing
simply a mantle (seventh figure from right), and finally as a consul
wearing a toga with a smooth band (early to late third century; tenth figure
from right). During his lifetime, of course, the man could quite easily have
worn the two later styles of the toga and, as suggested above, the toga for
his wedding may have been special for the occasion.

13 Short togas. Bas relief from the Arch of Marcus Aurelius, Rome, dated AD 176-80.
Small togas
Although the length of the fashionable toga changed over time, it is clear
that short togas were always worn by those unable to afford the full-width
toga. The changing fashion of the toga is based on the images of
Emperors, high officials, or the very rich, who could quite obviously
afford togas long enough to sweep the ground if they so wished; in the first
century, Horace talks of a rich ex-slave ‘parading from end to end of the
Sacred way in a toga three yards long’ to show off his new status and
wealth (Epodes, 4.8). Martial refers to short togas when referring to a poor
man, to a countryman who rarely wore a toga, and to the meagre gift
received from a patron, as well as the clients of a rich patron (Epigrams,
11.56.6; 4.66.3; 10.15.7; 11.24.11; 10.74.3). There are also a couple of
second-century scenes of the Emperor, dressed in a full-length toga,
talking or giving gifts to the general public, who are shown wearing knee-
length togas (13).

MANTLE, CLOAK AND CAPE

The Roman’s concept of ‘outer wear’ was not the same as the modern idea
of ‘outdoor wear’. When outside the house, Roman men (other than slaves
and the poor) would wear something over their tunic, not because they
necessarily needed protection against the weather, but because it was the
respectable thing to do. Even inside the house some form of outer-wear
was often worn, while the clothes worn especially for meals noticeably
consisted of both tunic and mantle. Historically, the toga and the mantle
had been the most important items of male clothing, worn with little on
underneath, and the tunic had only been developed as a major item of
dress at a later stage. Roman outer wear should, therefore, not be seen as
merely ‘outdoor wear’, but as an integral part of the costume, of equal
importance to the tunic.
14 Mantles and cloaks. 1 Emperor Hadrian wearing a mantle, AD 117, British Museum; 2 Man
carrying cloak with brooch, fourth-century tomb wall painting, Silistra, Bulgaria; 3 Men at the
circus wearing cloaks, third- or fourth-century mosaic, Cologne, Germany; 4 Hunter wearing
decorated cloak, fourth-century mosaic, Piazza Armerina, Sicily.
This outer wear came in three main forms; mantles, cloaks and capes.
Mantles (which originally included the toga) were simply draped round
the body without any form of fastening, and therefore needed occasional
re-arrangement; one arm was permanently occupied in keeping a mantle in
place. Cloaks were fastened with a brooch on one shoulder (usually the
right), so that the arm was left free of the folds. Capes, of varying lengths,
were sewn up the front to form a bell-shaped poncho, keeping both arms
covered. Each form was worn for different occasions.

MANTLES

The pallium was a Greek garment, which came to be regarded as the

Hellenic equivalent of the Roman toga as a form of national dress.
Because of the Greek association, it was considered the correct dress for
philosophers and scholars, or those who wished to appear cultured. It was
a large rectangle of cloth, draped over the left shoulder, round the back,
under the left arm and back over the left arm and shoulder in a similar
fashion to the early forms of toga (14.1). It was distinct from the toga,
however, because it always had straight edges. It was worn without a tunic
for those who wished to pose as a Greek philosopher (11, central figure),
but most Romans wore a tunic underneath that could apparently be
unbelted. Tertullian’s book in praise of the mantle lists its advantages; it
was very quick and easy to put on, there was no worry about the way the
folds fell, it was easy to re-arrange, the tunic underneath could be
comfortably unbelted and it could be worn with healthy openwork shoes
(On the Mantle, 5).
Another form of mantle was the abolla, possibly distinguished from
the pallium by being made of a thicker cloth (although it has been
suggested this is just another general term for a wrap), while the endromis
was a thick mantle worn by athletes after exercising that could also be
worn as protection against the weather; Martial refers to a gift of an
endromis that, although inelegant, was ‘not to be despised, [as] in cold
December … you will laugh at winds and rains, clad in this gift’
(Epigrams, 4.19.40).
Like the toga, the mantle involved the use of the arms to keep it
round the body, and was therefore not a practical garment for physical
work; it is occasionally seen worn round the waist, in the manner of a
sarong, by musicians and men involved in religious ceremonies, as well as
by a butcher on a tombstone from Rome, so that both hands were free (e.g.
Nappo 1998, fig. 87; Feder 1978, 129). The mantle, although an outer
garment to be worn over a tunic, was not only outdoor wear, but could be
worn inside, most frequently as part of the dining-clothes set, when it was
also often coloured (such as shown in the dining scene from Pompeii now
in Naples Museum: Feder 1978, 35).
Decorated white mantles appear in the east by the second half of the
first century (Adams and Crowfoot 2001, 35). The motifs were generally
purple in colour, and were L- or H-shapes set near the four corners,
although fourth-century catacomb paintings from Rome also show
swastikas (a symbol of good luck to the Romans), and double-ended
crosses (Adams and Crowfoot 2001, figs 4.3-4.5; Pfeister and Bellinger
1945, 11; fig. 1, nos 19, 24-32; Ferrua 1991, figs 82, 134).
Mantles were frequently, but not always white, especially when worn
with a white tunic. Suetonius tells the story of Caligula executing a man
because his purple mantle attracted too much attention in an amphitheatre
(Caligula, 35), while wall-paintings from Pompeii show people in the
forum in brown mantles and men in dining-robe sets with red and green
pallia (Nappo, 1998, 47).

CLOAKS

There are a number of different words for cloaks, and it can be difficult to
understand the exact difference between the forms. At its most basic, a
cloak was simply a rectangle of cloth fastened by a separate brooch,
almost always on the right shoulder, thus leaving both hands free. In the
Roman world, brooches were functional rather than purely decorative in
the manner of modern jewellery, and in mainstream Italian fashion they
were worn almost exclusively by men. To avoid misplacing the brooch,
they were often left fastened to the cloak, even when not being worn
(14.2). Cloaks could be of any length from thigh-length to ankle-length,
they could have a fringe on their shorter edges and they could have a
straight or curved lower edge (14.3-4).
The sagum was a thick, coarse cloak that probably originated in
north-west Europe, and was used particularly by soldiers and country
workers, and any-one who had to work outside in adverse weather
conditions. The sagum was so closely associated with the military that it
became the symbol of war, just as the toga was the symbol of peace.
Another cloak originally with military associations was the chlamys,
another garment of Greek origin. A higher quality form of military cloak
that was presumably distinguishable by its colour and the quality of the
cloth was the paludamentum, worn by generals and Emperors. Emperors in
particular are shown wearing it on marble busts, with a large circular
brooch and a thick fringe. Military cloaks other than the sagum often had a
curved edge, so that the cloth at the front fell in graceful folds. Civilian
cloaks seem more often to have had a straight edge, so that the front of the
cloak fell in a straight line from brooch to hem (14.3).
The lacerna and laena are two forms of civilian cloak, although the
exact difference between them is unclear. The lacerna was not as thick as a
sagum, but was used particularly as protection against rain; the elder Pliny
cynically refers to people who raised their price for cloaks as soon as a wet
winter was predicted (Natural Histories, 28.60.225), and there are
references to them being worn over togas at public spectacles, where they
were generally white in colour. The colour of cloaks, of course, depended
on the wealth of the owner, but in the first century Martial refers to
lacernae made of undyed wool, brown, black, red, scarlet, purple and
white (Epigrams, 14.133, 127; 4.2; 14.129, 131; 13.87; 14.137).
From the late third century, if not before, cloaks were frequently
decorated in the same way as mantles, sometimes with plain L- and H-
motifs or swastikas, and even patterned roundels on front and back (14.2,
14.4). By the fourth century, the costume of the soldier was that worn, in
modified form, by all high officials, up to and including the Emperor, and
portraits of the powerful begin to show them in a military cloak rather
than the toga (15.1). Originally, military cloaks were knee- or mid-calf in
length, but during the fourth century they become ankle-length and the
round disc brooch was replaced by a large crossbow brooch (colour plates
1-2). The crossbow brooch was no longer simply a functional object, but
had become symbolic of rank in itself, and examples have been found
made of precious materials, some with inscriptions suggesting they were
gifts from the Emperor himself (e.g. Gilles 2006, no. 73).
Some cloaks of the late Roman period were made out of patterned
cloth, such as that worn by the General Stilicho (8), reflecting the general
trend towards pattern in the late Empire. Others were plain, but decorated
with a tablion, a modern term used for an inset rectangle of a different
coloured cloth, sometimes decorated, set half way down the front of the
cloak against the straight edge, which was a symbol of the Imperial court
(Cleland et al 2007, 186). This can be seen in the Ravenna mosaic of
Justinian, where the court officials wear plain white cloaks with purple
insets, while the Emperor himself is set apart by having a dark purple
cloak with orange dots, and a tablion of gold cloth with blue birds within
red circles (colour plate 1).
15 Cloaks and capes. 1 Decorated cloak, mid-fourth-century relief in silver, Proiecta Casket,
British Museum; 2 Ploughman in cape, statuette from Trier, Germany; 3 Shepherd in skin
shoulder cape, third-century bas relief, Rome; 4 Men in capes at a shop, late second-century bas
relief, Museo Ostiense.

CAPES
In some ways the cape was similar to a cloak, but instead of being fastened
by a brooch it was sewn up the front to give better protection than a cloak,
which usually leaves the right side less well covered than the left because
of the way it was fastened (noticeable particularly in windy weather, when
the cloak has to held closed). The cape also usually had a hood. There are
three Latin terms for long capes: paenula, byrrus and caracalla, although
it is difficult to identify the exact differences between them, or match
them up with the artistic evidence. The neck of the cape is sometimes
shown with an extra form of fastening at the neck, such as thongs, flaps or
an added gusset, and it has been suggested this could be a distinguishing
feature of a type of long cape called the byrrus (colour plate 7; Wild
1963). The cape called the caracalla may have been the version worn in
the north-east provinces (adopted by the Emperor Marcus Aurelius
Antoninus, and providing him with the nickname he is more commonly
known by, Caracalla: Hendzsel 2008, 31).
If necessary, one or both sides could be rolled up onto the shoulder to
leave the arms free for work. The cape, when worn in this fashion, forms a
V-shape front and back (15.4, colour plate 3). Soldiers seem to have
favoured a cape that was closed only a short way up the front, either sewn
or fastened with toggles (Sumner 2009, 75). Civilian capes are usually
depicted with an opening or seam down the front, with the two edges
always touching and not gaping or opening up in any way (15.2),
frequently with spaced pairs of horizontal lines, as if depicting large
stitches or lacing. Some are very clearly sewn up right to the hem, but very
often a very short length is left unsewn, producing a characteristic ‘W’
shape to the lower edge (6.3, 15.4 figure to left). Soldiers, in particular,
and some civilians often either rolled up the cloth round the neck, or else
wore a large scarf tucked round their neck (15.4 central figure).
Capes were worn by all classes in very bad weather, or when
travelling, and by country workers who were out in all weathers. They
were made of thick wool, sometimes with the lanolin still in it to make it
waterproof, or of leather, while a letter from Egypt also mentions a cape of
goat-hair (Oxyrhynchus Papyri, 3871). Juvenal refers to ‘the rain pouring
in streams off the cape’ (Satires, 5.79), and Martial recommends a cape
since, ‘although you may set out on your journey when the sky is
continuously serene, keep a leather cape nearby in case of sudden showers’
(Epigrams, 14.130).
 As Romans very rarely wore hats, a hood was the usual head
protection for bad weather, either as part of a long cape, or as a shorter
type that generally only covered the shoulders (cucullus or bardocullus,
6.3). Martial refers to a hood ‘contaminating the purple clothes of town
with its grease’ (Epigrams, 1.53.5), suggesting it was made of unwashed,
undyed wool of the type worn by country folk, while scenes of shepherds
sometimes show them wearing shoulder capes made of sheepskin or fur
(15.3). Although frequently shown in rural settings, there are references to
them also being worn in town, as a way of hiding the face to avoid
recognition: Juvenal refers to a man stealing out for adultery, his ‘brow
concealed under a hood of Gallic wool’ (Satires, 8.145). They could also
be worn over a cloak, although it was sensible to make sure they were of
matching colours; Martial refers to the dye of a blue-green hood running
onto a white cloak (lacerna, Epigrams, 14.140(139)).
Waterproof capes or hoods made of undyed wool would have been
cream or dark brown in colour (colour plate 3). Other capes would have
been dyed; apart from the blue-green hood mentioned above, Martial
refers to a scarlet hood (Epigrams, 10.76) and Juvenal to a blue one
(Satires, 3.170), while one of the gift tags written by Martial describes a
thick frieze or felt cape: ‘such is my brilliant whiteness, such is the beauty
of my wool, you would choose to wear me even in the midst of harvest’
(Epigrams, 14.145). The sons in the bath procession mosaic at Piazza
Armerina are shown in cream/yellow and green capes (colour plate 6). In
the sixth-century Biblical scenes in the Basilica of San Appollinare
Nuovo, the men all wear white tunics and white mantles (apart from
Jesus), but when they are shown wearing capes they are allowed some
colour; light brown, yellowish, bright red, pale red and purple (Bustacchini
1988, 110, nos 3, 6-7; 114, nos 1-2; 117, no. 2).
Decoration seems to have been very restricted. The men receiving
bread depicted in a Pompeian wall painting have L-shapes on their hoods,
and one of the sons going to the baths in the Piazza Armerina mosaics has
short, straight lines on the front (colour plates 3, 8; Hendzsel 2008, fig.
3.1). A man getting undressed in the Piazza Armerina bath mosaics
appears to be wearing a white cape with red stripes, and an illustration
from a second-century document shows a man wearing a blue cape
decorated with purple stripes (Gentili 1959, pl. V; Sumner 2009, 129).
 Although the cape started as bad-weather clothing, it seems to have
been the most common form of outer wear for the lower classes, while in
the north-west provinces it took the role of the cloak. It was eventually
adapted by Christians as the dress of their priests (colour plate 1,
Archbishop Maximianus, to the right of the Emperor; Wild 1963, 196),
and as such has survived down to the present day.

16 Tombstone of Ares, shown dressed as soldier (left) and civilian. Late second-century bas
relief, British Museum.

LEG COVERINGS
TROUSERS

First and second centuries

In the first and second centuries, town-dwelling Romans would have
generally been bare-legged. Full-length trousers were considered the
costume of barbarians, and would not have been worn, even in cold
weather; instead, wrappings of cloth were used on both upper and lower
legs. The rural population, including rich men out hunting, had more need
of protection for their legs, not only from the weather but from
undergrowth and plants, and often had their lower legs covered with one
form of puttee or another, possibly including strips of cloth wrapped round
in the manner of a bandage. Such strips could also be worn on the upper
leg, either for warmth or protection. Juvenal, when describing a woman
training as a gladiator, comments: ‘see how thick the bandages wrapped
round her thighs are’ (Satires, 6.263). Soldiers, including cavalry, wore
breeches that extended just below the knee (5.1, 16), but it is not clear how
often civilian horsemen also wore them, as they are usually depicted
wearing only puttees.

Third and fourth centuries

The Emperor Severus Alexander is described as always wearing puttees
and choosing to wear white breeches (bracae) instead of the ‘scarlet ones,
as was the previous custom’ (SHA Severus Alexander, 40.11). The knee-
high lower leg coverings continued to be worn by those in the countryside
throughout this period, but the introduction of the long-sleeved, decorated
tunic was accompanied by a new form of leg covering. They were not
trousers in the modern sense, but were closer to medieval hose, being
tight-fitting, following the shape of the leg, and with integral feet. A
fourth-century wall-painting depicts a slave carrying a pair over his
shoulder, showing the belt threaded through loops and the fitted tailoring
of the calves (17.1).
It is not clear exactly how these leggings were made, but there are a
number of possibilities. Knitting was known during the Roman period, but
it is not known how widespread it was. More common was sprang-work, a
form of knotted fabric which has equal elasticity to knitting. Both these
methods would provide enough give to let the cloth stretch over the leg.
Alternatively, the leggings could have been cut from woollen cloth, as in
the medieval period. These have to be slightly baggy round the ankle to let
the foot through, and round the knee to allow the leg to bend. Some
representations show fold marks as if the leggings are not absolutely
figure-hugging, while others, particularly mosaics, show them following
the shape of the leg exactly, which might just be artistic
license.
 When the leggings are shown as skin-tight, it is not always clear they
are being worn, unless the image is in colour. One of the earliest certain
examples of leggings comes from a third-century wall painting at Dura-
Europos, where soldiers are shown wearing dark-brown coloured leggings
(Sumner 2009, pl. 19). The animal-collectors shown on early fourth-
century mosaics at Piazza Armerina wear leggings of dark grey, blue,
olive, or cream, and a hunter on a mosaic from Apamea in Syria wears
pink leggings (Capizzi and Francesco 1989, 50, 54-5, 58-9; King 1982,
123). Leggings are very rarely shown with any form of decoration, and
usually those that do are worn by mythological or historical characters and
are therefore unreliable, but the late fourth-century Lord Julius mosaic
from Carthage does show a hunter wearing hose decorated with a line of
dots down the front (Blanchard-Lemée 1996, fig. 121). It seems, however,
that decorated leggings were generally not very common.
In the fourth century, there are a number of depictions of huntsmen
and sportsmen combining upper and lower leg coverings. They are shown
with skin-tight leg coverings visible above the knee, puttees below the
knee, and a separate coloured section over the knee. The marble pavement
depicting Junius Bassus with four horsemen from the circus shows clearly
that this section is not just a patch of flesh visible between breeches and
puttees: the four wear off-white upper and lower leg coverings, and knee
pads of dark green or blue (17.2).

Fifth to seventh centuries

Hose continued to be worn until the end of the Roman period. A late
fourth-century mosaic of the bishop of Milan shows him wearing white
leggings, while the sixth-century mosaics of Justinian and Theodora show
that, at least at court, high officials, churchmen and soldiers were still
wearing white leggings, although those of the Emperor himself were
purple (colour plate 1). A lack of coloured images of the lower classes
makes it difficult to say if white leggings were worn generally.
17 Leggings. 1 Man carrying hose and shoes, fourth-century tomb wall painting, Silistra,
Bulgaria; 2 Horseman wearing puttees, fourth-century marble picture, Basilica of Junius Bassius,
Rome; 3 Hunter, mosaic from Henchir Toungar, Tunisia; 4 Farm-worker wearing cross gartering,
third-century mosaic from Cherchell, Algeria.
18 Puttee from Søgaards Mose, Skive Museum, Denmark. 1 As found on body; 2 Unfolded.

PUTTEES

In the country, rich and poor alike wore a knee-high form of puttee or sock
to protect their lower legs. In the first and second centuries, these lower
leg coverings are generally shown as being plain, with ties at the ankle and
below the knee (16). It is likely these take the form of a rectangle of cloth
wrapped round the leg and fastened with ties. A pair of such puttees was
found on a body recovered from a bog in Denmark (18, shown both as
worn when found and opened out). They were made from woollen cloth
with a single stripe (36 cm long by 31 cm wide) and thin woollen cord ties
(63 cm long), and may have originally been blue, as traces of woad were
found on one of them (Hald 1980, 34-6).
By the third century, puttees are shown most frequently with a bold
inverted-V or herringbone pattern in dark brown or black (colour plate 4),
probably depicting both ends of a long strip of cloth wrapped round the leg
at an angle so that they overlap. Alternatively, these puttees could have
taken the form of a true sock with a heavy weave or pattern. However they
were made, they were fastened with thin ties just under the knee (17.3).
Although the inverted-V-patterned puttee is the most common type
depicted in art, a few other styles are also known, for example cross-
gartering, where a narrow braid or thong is criss-crossed over an inner
sock or wrapped cloth (17.4). Some sportsmen, including circus riders and
gladiators, wear puttees, with one end of a strip wrapped round and round
the leg in overlapping, horizontal lines (17.2).
In the sixth century, there are depictions of puttees worn over the top
of other leg coverings, and tied with a garter under the knee. A painter in
an illustration in the Vienna Dioscurides wears white leggings and dark
brown puttees tied with a red garter (Weitzmann 1977, pl.17). Red garters
are also worn by a soldier in a picture in the Vienna Genesis, wearing dark
blue puttees over light blue leggings, and white ties are known in the
Asburnham Pentateuch (ibid., pl. 27; Weitzmann 1979, pl. XIII).

UNDERWEAR

Loincloths
There is evidence both for and against men wearing any form of briefs
under their tunics. The reason for doing so seems to have been modesty.
Cicero commented that an actor on stage wore a loincloth ‘for fear he
might make an improper exhibition, if by some accident certain parts of
his person should happen to become exposed’ (Offices, 1.35.129). A
mosaic from Rome also shows a dead or wounded hunter from a wild beast
display wearing a white loincloth under his tunic (colour plate 9), while
the loincloths sent to a soldier at Vindolanda along with socks and sandals
were probably intended as underwear (Bowman and Thomas 1983, no. 38).
On the other hand, there is evidence to show the practice was not
universal. Another hunter, on a mosaic from Antioch, is shown with his
tunic hitched up round his waist, and is clearly not wearing anything
underneath. The fifth-century writer Sulpicius Severus relates a story of a
Christian reprimanded for indecency because, while he was warming
himself in front of a hearth, ‘he brought a stool near and sat back with his
feet spread apart over the fire with his groin exposed’. Hugely
embarrassed by the reprimand, he ‘ran almost fainting to [his companions]
and acknowledged his shame’ (Dialogues, 3.14). It seems likely that
people wearing short tunics, such as soldiers, wore underwear, while those
in less danger of exposing themselves did not bother.
Loincloths were certainly known as outer-wear, but these were
usually the preserve of rural workers such as farmhands and fishermen, or
other low-ranking slaves such as mill workers (Olson 2003, 207). One
form of loincloth was probably folded in a very similar way to a nappy,
with the addition of two long ends, often fringed, that hung down below
the knot at the front, as shown on a simple carving of a quarry-worker
found in a Roman quarry in Germany (19.1). Another was simply a long
length of cloth passed between the legs, wrapped round the waist a number
of times and the two loose ends either tucked in or tied in a number of
different ways, the front end often hanging down as a form of apron
(White 1977, pls 12d, 18; Capizzi and Francesco 1989, 20, 40). A third
form, frequently shown being worn by gladiators, had roughly semi-
circular loops of cloth draped front and back, fastened by a belt at the
waist. This may in fact have been a length of cloth worn in the fashion of a
sarong or kilt, which was hitched up on either side (19.2). Sometimes
farmworkers worked naked when doing messy summer-time jobs, such as
treading grapes or threshing grain (White 1977, pls 12, 18).
19 Loincloths. 1 Quarryworker, carving from a quarry at Kruft, Germany; 2 Fisher, mosaic from
Lepcis Magna, Tripoli Museum.
20 Knitted sock from Egypt. © National Museum of Ireland.

Socks
Socks were widely worn in the Roman world. Some of the knee-high leg
coverings may have been socks (see above), but true ankle socks were also
known. Martial refers to socks made of goat’s hair cloth (Epigrams,
14.140), while a child’s sock from Vindolanda was made from woollen
cloth (as were the women’s socks from Les Martres-de-Veyre). The sock
was made from two pieces of diamond twill cloth cut to shape and roughly
tacked together (Wild 1993, 83, pl. XII, no. 2). Other socks surviving from
Egypt were made of sprang-work, a form of knotted fabric, or from knitted
wool (20). Egyptian socks in particular were made with a built-in slot for
the thong of the sandal. From the third century, separate socks were not so
necessary, as the leggings had integral feet (17.1).

Winter Wear
Most leg coverings were for use in the countryside, but even in town
wrappings for the legs and chest (worn under outer clothing), as well as
scarves, were acceptable for those who were sick, or felt the cold badly in
winter.

FOOTWEAR
In the Roman world, footwear said much about the person wearing it: there
were shoes for wearing inside the house and others for outside, some for
wearing with togas and some for mantles, some worn by the military and
some by civilians. There were even shoes that could only be worn by
certain ranks of men. It was important for a respectable man to wear the
correct form of shoe in the correct circumstances; closed boots, for
example, with the toga, but openwork shoes with the mantle. Different
types of shoes would also have been worn on different occasions for
different activities. The crew of a ship which sunk in the Rhine had both
one pair of closed shoes and a pair of sandals each, suggesting the need for
footwear for specific uses (van Driel-Murray 2001, 185). In the fourth
century, men are shown wearing sandals when they were bare-legged, such
as when they were inside the house (22.1, colour plate 8), while they wear
shoes with low sides with their hose.
There are a number of different words for shoes, boots, slippers and
socks, the exact form of which is not always easy to identify fully,
particularly when they are compared to surviving shoes from the
archaeological record or, indeed, the evidence from art. Therefore, the use
of the Latin terms for specific shoe-types in the following has to be taken
with some caution (Goldman 1994, 105).
In the fourth century, Diocletian’s Edict listed a number of shoes and
boots that illustrate some of the available range: there were boots for mule
drivers or farmworkers, boots for soldiers (both noted as being sold
without hobnails, suggesting that the nails had to be bought separately),
shoes for soldiers, Gallic sandals, Gallic sandals for farmworkers (double-
soled), Gallic sandals for runners, and socci (of the first quality).
Babylonian sandals, Babylonian socci, Phoenician socci, gilded ox-hide
shoes and wool-lined ox-hide sandals were also mentioned, without being
identified by the intended sex of the wearer. However, the archaeological
record has shown that on the whole men, women, and children tended to
wear the same design of footwear anyway (van Driel-Murray 2001, 194).
 There are three main manufacturing techniques; sandals, which at the
most basic are simply a sole held onto the foot by thin straps; one-piece
shoes where holes round the edges of a piece of leather are threaded
through and pulled together, and multi-piece shoes, where the uppers
could be made of several different pieces which were attached to soles
also made up of a number of different pieces. One-piece shoes were
technically very simple to make, but could still be cut in such a way as to
form very elegant shoes that were often asymmetrical in design (van
Driel-Murray 1987, figs 3-4). Multi-piece footwear could be made as both
shoes or boots, and included some sophisticated features, such as leather
scraps packed between the different layers of soles to avoid the inner sole
dipping (ibid., fig. 8). Surviving examples show the high level of
workmanship Roman cobblers could achieve (Goldman 1994, figs 6.14,
6.25; Waterer 1976, fig. 303).
All Roman shoes were flat, without any form of heel. Both sandals
and multi-piece shoes were often hobnailed to make them long-lasting, as
iron nails do not wear down as quickly as soft leather. Hobnails, however,
are not the best things to wear when on a hard surface such as paving, as
they cannot grip. Josephus tells the story of a Roman soldier called Julian
attacking Jews in the inner courtyard of the temple in Jerusalem, who ‘was
himself pursued by fate; for as he had shoes all full of thick and sharp
nails, as had every one of the other soldiers, so when he ran on the
pavement of the temple, he slipped, and fell down upon his back with a
very great noise, which was made by his armour. This made those that
were running away turn back’ and he was surrounded and eventually killed
(Jewish War, 6.1.8).

Leather
Most Roman shoes were made out of leather, but could also be made out
of vegetable matter such as straw, palm-fibre or reed, or have elements
made out of cork or wood. The Romans introduced vegetable-tanned
leather into the north-west provinces, where skins had previously been
smoked or treated with fats (van Driel-Murray 2001, 185). Vegetable-
tanned leather is brown, but could be stained, dyed or painted to produce
other colours, and could have gilded or stamped decoration. It could also
have a polished or dull surface, or a smooth or grained finish (Waterer
1976, 180). Alum-tawing (mineral tanning) produced a white leather, and
references to white shoes probably refer to items made from alum-tanned
leather, although this type was not waterproof.

SHOES WORN WITH THE TOGA

The Romans used the word calceus, ‘shoe’, to describe the form of
footwear worn with the toga, but it was more like an ankle boot, so the
term shoeboot will be used to describe it (Goldman 1994, 116). Men
wearing togas are shown wearing three different forms of shoeboot. The
first type was worn by common people, and was a plain shoeboot with a
characteristic ridge shown across the instep (16). It is not clear exactly
how it was put on and fastened; some of the figures on the Ara Pacis have
a flap across the front of the foot, making part of the characteristic ridge,
but no form of fastening ties or toggles are shown. They could be made,
like all the forms of shoeboot, of such fine leather that the shape of the
toes was visible (30).
The second form of shoeboot was another ankle-high boot that was
open down the front, often with a tongue; it was fastened by two straps
attached to the side of the shoe at the widest point, and which were crossed
over the top of the foot and wrapped a number of times round the ankle.
The straps were tied with a knot at the front, with the long ends left
dangling so that they almost touched the ground (21.1). The third form of
shoeboot had an extra set of straps that were tied half-way up the calf so
there were two knots at the front; the ends of the upper knot were often
even longer than those of the lower one, and were tucked into the lower
straps to stop them trailing on the ground (Goldman 1994, figs 6.22,
6.23b).
These shoeboots with straps tied round the ankle were for the
senatorial and equestrian classes. The exact distinction between the
shoeboot for the two ranks is unclear, and even the evidence for colour is
contradictory. There were three forms of shoe that could only be worn by
men of the right social rank, listed by Diocletian’s Edict: patrician (with a
price of 150 denarii), senatorial (100 denarii) and equestrian (70 denarii).
The patrician’s shoe is described as being mulleus, the colour of the red
mullet, a deep pink. Literary evidence also suggests that the senator’s
shoeboot had a crescent sewn onto it, but there is no recognised
representation of this in art.
 The senator’s shoeboot, or perhaps only the straps, might have been
black; in the first century Horace talks of a new senator wearing a broad
stripe on his tunic, and ‘[binding] the black leather [straps] halfway up his
leg’ (Satires, 1.6.27-8), while Juvenal mentions a man ‘sewing onto the
black soft leather the crescent of the senator’ (Satires, 7.191-2). Martial,
however, when referring to an ex-slave sitting in the senators’ seats in the
theatre mentions both crescent and shoe-strap, but also describes the man’s
‘unchafed foot … decked in scarlet soft leather’ (Epigrams, 2.29). The
lower classes may have had more choice of colour: another man, of
unknown rank, wore shoeboots ‘whiter than untrodden snow’ under the
trailing end of his toga (Martial, Epigrams, 7.33.1-2).

21 Footwear. 1 Shoeboot, statue, National Museum of Antiquities, Edinburgh; 2 Openwork shoe,

statue, Vatican Museum.
The shoeboot was worn with the toga throughout the Roman period,
with the characteristic crossed straps and knots with long ends clearly
visible on sixth- and seventh-century portraits of consuls (for example, see
9.2, 12).

BOOTS

The closed boot (pero) was worn throughout the Roman period, and may
have been very little different to the first form of shoeboot described
above. The men at the baker’s shop in the Pompeian wall painting wear
dark brown ankle boots (colour plate 3). In the early Empire, boots were
worn particularly by poor people or country folk, and Juvenal refers to ‘the
man who is not ashamed to wear high boots in time of frost’ (Satires,
14.185).

OPENWORK BOOTS AND SHOES

Openwork shoes, made by cutting leather into a number of straps, were

also worn by men (21.2). Openwork shoes in particular were worn with the
Greek mantle; Tertullian praises them for being healthier than the enclosed
shoe used with the toga (On the Mantle, 5.3). The simplest forms were
made out of one piece of leather (sewn up the heel), which could be very
basic in design, or have a highly sophisticated cutting pattern (Goldman
1994, figs 6.19-20). More complex shoes had separate uppers and multiple
layers of soles.

Military openwork footwear

In the early Roman period, there seems to have been certain types of
footwear worn only by the military, although in the later period there
seems to be no distinction between military and civilian. In the novel
Satyricon, a character trying to pass himself off as a soldier is let down
when a genuine soldier points out he is wearing very unmilitary white
Greek shoes (Petronius, Satyricon, 82). The most extreme form of
openwork boot is that worn by soldiers in the first century (traditionally
called caligae by modern scholars). Although perhaps the best-known type
of Roman footwear, archaeological evidence shows it went out of fashion
during the AD 80s (van Driel-Murray 2001, 190). It was replaced by other
boots where the fine openwork was net-like (ibid., fig. 1; Goldman 1994,
fig. 6.25). From about the middle of the second century, shoes with much
simpler openwork became standard for soldiers (van Driel-Murray 2001,
fig. 1). In the late second and third century, the design changed again to
enclosed boots or shoes, where the straps were integral to the shoes and
were laced through loops cut in the uppers (16; Sumner 2009, fig. 114; van
Driel-Murray 1987, fig. 8). This type of footwear was also worn by
civilians, and from this period on there appears to have been no distinctive
military-style footwear.

SHOES WITH LOW VAMPS

In the fourth century, a shoe with low vamps (the sides between the toe and
ankle sections) was worn with hose. These shoes were enclosed over the
toes and round the heel and tied at the ankle, and seem to be almost
exclusively black in colour (22.2). This style of shoe seems to have lasted
a long time, as the sixth-century mosaic depicting Justinian’s court,
soldiers, high officials and churchmen all wear the same type, made of
black leather. The Emperor himself wears purple hose and red shoes
decorated with pearls and blue stones set in gold (colour plate 1).

SANDALS

Sandals could be made out of leather, or plaited and sewn palm fibre or
other similar materials (Goldman 1994, 107; Walker and Bierbrier 1997,
212, nos 327-9). While some sandals would be have used for wearing
round the house, many surviving examples have multiple layers of soles
(up to seven layers) or hobnails, and were clearly substantial items
intended for use outside. In the first and second centuries, the sandal soles
have a natural foot-shape, sometimes with an indent emphasizing the
second toe (van Driel-Murray 2001, fig. 3). However, in the late second
century male sandals start to become broad at the front, one of the few
times male and female footwear fashions separated (ibid., 194). By the
late third century and into the fourth, this style has become so exaggerated
the soles are almost triangular in shape, which would have affected the
walk of the men using them (ibid., fig. 3).

HOUSE SHOES

There were separate forms of shoes or sandals that were worn inside the
house, and only rarely outside. When dining at a friend’s house, outside
shoes were exchanged for thinner house shoes; those rich enough brought
a slave to carry their shoes for them, and to look after their outdoor shoes
while they ate. Martial refers to a man at the baths hoping for an invitation
to dinner already wearing his house shoes (Epigrams, 12.82.6). These
house shoes are often described by the word soleas, which Aulus Gellius
described thus: ‘in general all kinds of footwear which cover only the
bottom of the soles, leaving the rest almost bare, and are bound on by
slender thongs are called soleae, or sometimes by the Greek word
crepidulae’ (Aulus Gellius, 13.22.2-8). Martial refers to house shoes lined
with wool (Epigrams, 14.65).

22 Footwear. 1 Sandal worn by slave, mosaic from Piazza Armerina, Sicily; 2 Shoe worn by a
bishop, mosaic from Milan, Italy; 3 Clogs worn by water-carrier, Hellenistic statuette, Vienna
Kunsthistorisches Museum.

SOCCI

It is difficult to distinguish in literature between cloth shoes worn as outer

wear, and socks worn in shoes and boots. Translations usually refer to the
soccus, worn by comedians in particular, as a cloth shoe, despite the fact
that a cloth shoe must have had a very short life. Although a
comparatively large number of Roman socks have been found, some
scholars in the past were very reluctant to believe that Romans commonly
wore socks in their shoes or sandals. However, it is likely that it was
intended that coloured cloth linings or socks would be visible through the
openwork design used for so many shoes (cf Bishop and Coulston 2006,
fig. 2c).

CLOGS AND PATTENS

Footwear could be made from woven palm and other material, but on the
whole was made of vegetable- or alum-tanned leather, and would not have
been very waterproof. Wet shoes grow soggy, as many a modern re-enactor
knows, and a wet shoe is sometimes used in literature as an example of
something loose or baggy (Martial, Epigrams, 11.21.4), while loose shoes
invited derision (Ovid, Art of Love, 1.514-6). People working out in the
countryside, especially within the winter in particular, needed some form
of footwear to keep out the wet and mud. Farmworkers wore wooden-soled
shoes, and Cato suggests slaves should be given a good pair every other
year (On Farming, 59.135.1). This type of clog is shown being worn by a
Hellenistic statue of a water-carrier in Vienna’s Kunsthistorisches Museum
(22.3). The shoes have a wooden sole with raised blocks under the heel and
the widest part of the foot, and are held on by straps of leather tied over
the foot and round the ankle. Large numbers of similar wooden soles have
been found on waterlogged sites, and some seem to have been slip-ons,
with a wide leather band over the foot (Goldman 1994, fig. 6.17). Some
were probably used in bath-houses. It is not known if these wooden-soled
shoes were always worn as clogs, or if they could also be used as pattens
(worn over other footwear to protect them from mud or wet).
In portrait statues, being barefoot was often a sign that the person was
divine, pious, or a hero (Goldman 1994, 105). However, many people
probably often also went unshod, as seen, for example, by the bare
footprints left by people walking over clay tiles left to dry before being
fired (Brodribb 1987, 125), or the people carrying out a number of
different tasks on a farm (White 1977, pls 12-3). People may have gone
barefoot only at specific times, for certain tasks where it was either more
practical to do so, or because it saved footwear from unnecessary damage,
but it may also have been their usual practice. In post-medieval Ireland,
for example, women and children living in rural areas frequently only
wore footwear on special occasions such as market days and church-going,
and some remained barefoot throughout the winter as well as the summer,
whatever the weather (Lucas 1956, 322, no. 101; 335, no. II; 337, no.
XXV).

HAIRSTYLES AND HATS

HAIR STYLES

Male fashions in hair followed that of the Emperor to a certain extent, but
not as closely as fashionable women followed the hairstyles of the
Imperial ladies. While many of the female styles would have required
false hairpieces, if not complete wigs, it was less acceptable for men to
wear them, and their hairstyles were therefore more dependent on nature.
Depictions of individual Emperors can often be recognised by the
arrangement of their hair and how it was cut, but it is not clear how closely
other men followed the style, particularly when it came to such matters as
the degree of curl, which for many men would not have come naturally.
Receding hairlines or bald patches were also unlikely to be imitated
intentionally!

23 Hairstyles of the first to third centuries. 1 30 BC – AD 117, Augustus; 2 30 BC – AD 117,

Nero; 3 AD 117 – 170; 4 AD 170 – 211; 5-6 third-century youths.
Suetonius says that the Emperor Otho wore a wig, and his portraits on
coins do indeed show someone with a full head of hair (Otho, 12). Only a
small number of people in the Empire would actually have seen the
Emperor in the flesh, so the majority of people were dependent on
portraits such as coins, official statues and commemorative works of art to
see what he looked like, and of course the Emperor had control over these
images. Perhaps the best example of this is Augustus, whose official
portraits continued to show the face of a young man up until his death at
the age of seventy-six (Kleiner 1992, 62). However, men could still copy
what they thought was the Emperor’s hairstyle, and there were a number of
broad trends throughout the Empire.

30 BC – AD 117

Augustus was clean-shaven, with short hair in comma-shaped curls,

brushed forward into a fringe at the front (23.1). The hair at the back of the
neck could be either short or long enough to brush forward round the neck
under the ears. This basic, simple style was copied by the Emperors
immediately following Augustus with only minor alterations. Nero is
sometimes shown with a short beard round the edge of his face, under his
chin (23.2). Two of the short-lived Emperors of the civil war period had
short hair with no curls, and when the new Flavian dynasty came into
power the Augustan style continued, with short hair with many curls
(colour plate 10).

AD 117 – AD 170

Hadrian was the first Emperor to have a full beard. From his reign
onwards, almost every Emperor (other than boy-Emperors) had a beard
until the fourth century, and it is clear that many non-Imperial men copied
the style. Hadrian was extremely fond of Greek culture, and as Greek
poets, philosophers and statesmen of the past were shown with beards, he
may have deliberately chosen to sport a beard (Kleiner 1992, 238). His
beard was short, following his jaw, and his hair remained short, with
comma-shaped curls (23.3, colour plate 10).

AD 170 – AD 211

Marcus Aurelius had a bulkier hairstyle, with more loop-shaped curls and
a longer, pointed beard (23.4). Septimius Severus had an even longer
beard, sometimes depicted as if it had small ringlets within it (colour
plate 10).

THIRD CENTURY

The fashion for beards continued throughout the third century. Generally,
but not exclusively, men’s hair was very short during this period, often
being little more than a crew-cut, with equally short beards. The third
century was a time of numerous Emperors, many with short reigns. There
was also a growing tendency to have co-rulers instead of a single Emperor,
with the more important Emperor referred to as the Augustus and the more
minor one, often the son and heir, as the Caesar. Boy-Emperors and sons
appointed as Caesars meant a number of youths were depicted on coins or
official portraits statues. Young emperors such as Caracalla (twenty-nine
when he died), Severus Alexander (twenty-seven) and Gordian III (about
nineteen) are often shown with long sideburns and moustache, or a narrow
beard that follows the jaw line with a separate moustache (23.5-6).
Severus Alexander, for example, was fourteen when he became Emperor,
and for the first few years his coins show him as a beardless youth. When
he was about eighteen, he is shown with short sideburns, and when
nineteen with a moustache and a beard that follows his jaw line. A few
coins dated to the last four years of his reign show him with a full beard
(colour plate 10). In the first century, some statues of Nero (seventeen
when first Emperor) also show long sideburns and a narrow beard, and it
may be these were always accepted as a young man’s hairstyle. The
moustache alone was not considered a very Roman hairstyle, and was
generally the preserve of Celts or northern barbarians when shown in art,
but in the third century at least, a moustache seems to have been
acceptable when worn by young men before they progressed to a full beard
(24.7).
24 Hairstyles of the third to sixth centuries. 7-8 third century; 9 fourth century; 10-11 fifth
century.

LATE THIRD TO EARLY FOURTH CENTURY

In the late third century, the hair is usually shown as a crew-cut, with an
equally short beard that often grows some way down the neck (24.8). In
AD 293 a new system of government known as the Tetrarchy (four man
rule) was introduced, in which the Empire was divided up so that each of
the four co-Emperors had separate areas of authority. To stress the
relationship between the four men, there was a tendency to suppress any
depiction of individuality in favour of a communal image for the four. In
AD 305, Constantine was appointed a Tetrarch, and at this stage in his life
was usually shown with a beard, like his fellow rulers. By AD 312,
however, there were only two Emperors instead of four, and in AD 324
Constantine got rid of his fellow Augustus and became the first sole
Emperor for forty years. There was a major change in the representation of
the official image of the Emperor after AD 312, and from this time on
Constantine was always shown clean-shaven (colour plate 10).
This alteration in hairstyle was a calculated move; the clean-shaven
face and short curls brushed forward recalled the hairstyle of Augustus and
his successors of 300 years earlier, and Constantine wanted to be
associated with the Augustan ideal and the glories of times past, not with
his predecessors, the Tetrarchs (Kleiner 1992, 436). The fourth-century
version of this hairstyle did not have short, tousled curls all over the head,
but instead the hair was straight and brushed forward from the back of the
head (24.9).

FOURTH TO SEVENTH CENTURY

This hairstyle remained in fashion until the seventh century at least,

although after the fourth century the hair was often slightly longer, with a
noticeable bulk over the ears (24.10-11). The principal exception to the
clean-shaven look was the Emperor Julian, who reigned AD 360-3.
Emperors had been Christian since Constantine, but Julian was a
determined pagan who wished to see a return to the old ways. Although
having a beard may simply have been a personal choice, as he had worn
one when young and had resumed it as soon as he became sole Emperor, it
did help him distinguish his image from that of his Christian predecessors,
and reflected the accepted image of the Greek philosophers he admired.
Julian wrote a book called the Beardhaters, in which he recalled the
ridicule he suffered for having a beard during a time when it was
unfashionable. After his short reign, the clean-shaven look returned,
although it is noticeable that during the fourth and fifth centuries it was
generally the usurpers to the throne who wore a beard, in contrast to the
generally clean-shaven image of the official rulers (such as Procopius, AD
365-6, and Eugenius, AD 392-4, who probably modelled their image on
that of Julian, and Johannes, AD 423-5, and Avitus, AD 455-6). It is likely
that many people followed Julian’s fashions as he was a legitimate ruler,
but it is possible that followers of the later usurpers deliberately wore
beards to show where their loyalties lay. The mosaic of Justinian’s court of
c. AD 546-8 shows the Emperor, his bodyguard and some of his high
officials with a variety of short hairstyles and generally clean-shaven
faces, while the Archbishop Maximianus is shown with a beard, which
seems to have become more and more common with priests.

HAIR OIL

It is clear from the literary evidence that some men resorted to curling
tongs to produce the perfect style when curls were in fashion (Ovid, Art of
Love, 1.505). They also used oil on their hair. Perfumed oils were used on
the head during feasts (Martial, Epigrams, 5.64; 12.17), but it is clear they
were also applied at other times, probably to keep the hair looking tidy
(Juvenal, Satires, 9.12; Horace, Satires, 2.3.125-6). In the time of the
Republic, Cicero referred to men ‘with their carefully combed hair,
dripping with oil’ (Second Speech Against Catiline, 2.22), while in the
first century, Ovid warned women against being swayed by men with hair
covered in nard (an aromatic ointment made from the spikenard plant),
and Pliny complains about soldiers who justified the use of hair-oil
because the military standards were anointed with perfumes on special
days (Art of Love, 3.443; Natural Histories, 13.4.23). Both spikenard and
myrrh, a resin used in incense, are mentioned as scents used in hair oil
(Ovid Metamorphoses, 5.52-3), and Martial complained about a man
‘whose oily hair can be smelt all over the theatre’ (Epigrams, 2.30). It is
not clear if the fashion continued into the fourth century, or whether it
came back into fashion then: Prudentius refers to men using spikenard on
their hair (Mansoul, 359-60), and hair surviving in a fourth-century male
grave at Poundbury, Dorset, still had oil or fat on it (Farwell and Molleson
1993, 205). The oil could also be used as a perfume: Pliny mentions a man
in hiding being discovered because of his powerful perfume, and says that
the Emperor Nero wore scent on the soles of his feet (Natural Histories,
13.4.22, 25).

HATS

In the Roman period hats were comparatively rare, generally being

functional and not worn simply as an item of fashion until the late third
century, when soldiers began to wear them. In the first century, Statius
mentions a hat, made of the cuttings from cloaks sewn together, as a
suitable cheap gift (4.9.24); an example of just such a patchwork hat partly
made from recycled garments has been found at Mons Claudianus in
Egypt (Sumner 2009, pls 22-3). A number of conical capes, made of
segments or quarters of different coloured cloths, have now been found in
Egypt, although some of them have been identified as caps to be worn
under helmets as padding (ibid., 166). These include a green hat made out
of felt, and one made up of triangles of red, green and yellow (Bender
Jørgensen 2000, fig. 4-5, colour plate 3, no. 5).
25 Headwear. 1 Priest, bas relief from Ara Pacis, Rome; 2 Soldier wearing pill-box hat, bas relief
from sarcophagus, Arles, Rome; 3 Scenes of slaves being freed, bas relief from Mariemonet,
Belgium.
Protective head coverings were used in particular by travellers and
labourers. The hoods of capes were used as protection against rain and bad
weather (15.2), while wide-brimmed straw hats were worn as protection
against the sun, particularly by fishermen and sailors (19.2). Augustus
hated the sun, and often wore such a hat, even in Rome, and similar hats
could be seen in the theatres when the wind was too high for the awnings
to be used (Martial, Epigrams, 14.29). A fold of toga or cloak could also
be used to protect the head, but was used particularly when a man was
sacrificing and covered his head out of respect (5.2). Brimless, tight-
fitting caps of fur or leather were worn by sportsmen and athletes such as
wrestlers and charioteers (Martial, Epigrams, 14.50, Aulus Gellius,
2.26.2), while caps made to a similar pattern, but much more elaborate,
were worn by some priests (25.1).
The Pannonian hat (pilleus Pannonicus) was a pill-box cap made of
felt, leather, sheepskin, or possibly fur that, as the name indicates,
originated in Pannonia (25.2). A sheepskin version in Diocletian’s Edict
cost 200 denarii, while Vegetius refers to some made of leather (Epitome
of Military Science, 1.20), and an example of green felt at least 2 mm
thick has been found at Mons Claudianus in Egypt (Sumner 2009, 165; fig.
107). The Egyptian example dates to the second century, but it is not until
the late third and early fourth century that they become common in art,
when they are usually shown being worn by soldiers. In the fourth century,
Ammianus Marcellinus refers to a soldier wearing a hat under his helmet,
probably to act as padding (29.8.8), and it may be that the pill-box hat
worn by soldiers, as well as being worn off-duty, had the same use.

Freedman’s cap
The most important hat with symbolic value was the freedman’s cap, a
brimless conical hat made of felt or cloth. It was worn by freed slaves, at
least in the first and second centuries, to show their new status,
symbolising liberty and freedom. Figure 25.3 shows the ceremonial
method of freeing slaves in the presence of a magistrate, with the two
freed men wearing the conical hat. It was otherwise a form of headgear
worn only on festive or ceremonial occasions, such as funeral processions,
and during the Saturnalia celebrations (Persius, 3.105, Martial, Epigrams,
9.6). After Nero’s death, ‘such was the public rejoicing that the people put
on liberty-caps and ran about all over the city’ (Suetonius, Nero, 67). It is
unlikely that the cap was worn much at any other time, as it would have
drawn attention to a man’s servile background.

SPECIAL COSTUMES
WEDDING

While some women wore a special outfit for their wedding, men had no
corresponding traditional costume and are usually shown in a toga (9.1,
11). Those men not entitled to wear the toga are likely to have worn a
white tunic, as white is frequently described as the colour to be worn for
celebrations (cf Horace, Satires, 2.2.60-2). In the play Casina by Plautus, a
man about to be married is described as ‘parading around with a garland
and white clothes on, all spick and span’ (767).

NIGHTWEAR

There is evidence for men both wearing and not wearing anything in bed
(Adkin 2000). This could be a matter of individual preference, the time of
year, or personal circumstances. A fourth-century educational glossary
records a boy’s day, starting from when he got up: ‘I awoke before dawn; I
arose from my bed; I sat down; I accepted socks and shoes; I put them on;
I requested water for my face; I washed my hands first and then my face; I
wiped them dry. I took off my night-clothes and put on my tunic; I did up
the belt’ (Colloquia Monacensia, trans. Shelton 1988, 113). In the early
seventh century, Isidore recorded that the word ‘nightshirt (camisia) is so
named because we sleep in these in our cots (cama), that is, in our beds’
(Etymologies, 29.32.27). On the other hand, a priest was reminded of the
occasion when, ‘a sudden earthquake in this province in the middle of the
night awoke us all out of our sleep, [and] you, the most prudent and the
wisest of men, began to pray without putting your clothes on’ (Jerome,
Against Vigilantius, 11; trans. Fremantle et al 1893).

RELIGIOUS

Most Roman religious activities were based on an offering of incense and

wine, or the sacrifice of animals and anyone, male or female, could
perform the ceremony. Those who could afford to own a toga usually wore
it for the sacrifice ceremony, pulling a fold up over the head to show
respect for the gods (5.2). Such sacrifices did not require a priest as such,
but when the sacrifice involved an animal, men were needed to kill the
beasts. They wore a fringed cloth round their waist sarong-fashion, which
reached to below the knees. It was held up with a belt, or band of cloth
wrapped several times round the waist, which also held the special
triangular-bladed knife used to deal with the animal (26.1).
Priests (and some priestesses) were found in both older, traditional
cults, and exotic new cults imported from the east, and they were more
likely to have distinctive items of costume (Cleland et al 2007, 154). In
some cases, there could be quite elaborate regalia, including decorated
wrist-clasps, body harnesses and crowns (for example, those worn by the
priests of Cybele: Jenkins 1991).

Priests of Jupiter
One of the oldest priesthoods was that of Jupiter, which had a number of
restrictions and taboos. Aulus Gellius describes the dress of one Jovian
priest: ‘he has not a knot in his head-dress, belt or any part of his dress ...
he must not be in the open air without his cap … he does not take off his
inner tunic except under cover … [and the Chief Priest] alone has a white
hat’ (10.15). The hat worn by the priests was a tight-fitting leather cap tied
under the chin (galerus), which had a wool-covered olive wood shaft
attached to the top of it, called the apex (25.1; Cleland et al 2007, 7, 77).

26 Religious costumes 1 Slaughterman for sacrifices, bas relief from Arch of Severus, Lepcis
Magna; 2 Priest of Isis, wall painting from Herculaneum.

Isis
One of the most important of the eastern religions was that of the Egyptian
goddess Isis. People connected to her worship, whether priests or
followers, wore linen. This was so well-known that they are almost always
referred to as ‘linen-clad’. Apuleius describes the priests in a passage in
his book The Golden Ass, where he goes into detail about the worship of
the cult: ‘the men’s heads were completely shaven … the leading priests
also clothed in brilliant-white linen drawn tight across their breast and
hanging down to their feet’ (11.10). A wall painting from Herculaneum
shows priests wearing this outfit (26.2).

MOURNING
There were a number of occasions when people were not supposed to take
care of their appearance, the most important of course being during
mourning. Men went unshaven, or with untrimmed beards, and wore old,
dark, or torn clothing. Ovid refers to ‘black clothes’ as the ‘clothing of
woe’ (Ibis, 1001-4), but it is clear any dull- or dark-coloured clothing
could be used; the clothes are described by words that can be translated as
dirty, mean, soiled, dark grey or undyed (sordidatus, pullatus). Similar
clothes were worn on other unhappy occasions, such as when Cicero went
into mourning in sympathy with a publicly disgraced relative (Letters,
5.1.2), although the most common reason was having to appear in a court.
In the fifth century, Sidonius illustrates the importance of dressing
correctly for court when describing the trial of his friend Arvandus:
Our man makes his way to the senate-house, having shortly
before been shaven and rubbed down, while his accusers, in half
mourning and unkempt, await the summons … having robbed
the accused of his due sympathy by their own intentional
squalor, and availing themselves of the indignation which the
sight of men in squalid clothes arouses … Because he had
marched into the presence of his judges adorned and elegant
whilst his accusers were dressed in black, the pitiable
circumstances in which he appeared only a little later evoked no
pity as he was dragged off to prison. For who would distress
himself greatly about the position of one whom he saw being
carried off to the quarries or the slave-prison so carefully
dressed and perfumed? (Letters, 1.7.9-11).
Such mistakes were not unimportant either, as the over-confident
Arvandus was in need of any mercy he could inspire, since he had been
found guilty of treason and sentenced to death, a sentence that was only
later changed to exile.

JEWELLERY
In the Mediterranean region men did not usually wear much jewellery,
since it was considered effeminate. Torcs, bracelets and ear-rings were
worn by men in other parts of the Empire, and the fashions occasionally
spread to Rome, but they were generally looked down upon as not being
‘Roman’ and the trends did not last long.

NECKLACES

Bullae
In the first century AD, young boys from rich families wore a hollow
circular pendant amulet (bulla) round their neck (27, 59), following an
earlier Etruscan habit. It was only worn by those of free birth, and was
given up at the same time as the toga praetexta (Cleland et al 2007, 26).
For the rich, the bulla was made of gold, and there is some evidence that
poorer boys wore a leather version (Juvenal, Satires, 5.165). The tradition
of wearing the amulet died out after the first century, although there are a
few late Roman gold-glass portraits of boys wearing one (Weitzmann
1979, no. 265; Pillinger 1984, Taf. 9, Abb. 79; Taf. 32, Abb. 246).

27 Amulet (bulla) from Pompeii.

Torcs
From the fourth century AD, there was a growing fashion for torcs. They
are commonly shown being worn by slaves and soldiers, and may have
been more popular in the eastern part of the Empire; although some are
shown in the mosaics of Piazza Armerina in Sicily, most are on
mythological beings, and only one on a male slave (Gentili 1959, pl. V).
They were not open at the front in the manner of some earlier Celtic torcs,
but could be hung with a pendant (6.1) or have a large central motif
(possibly jewelled; see Justinian’s bodyguard, colour plate 1).

FINGER-RINGS

The main item of jewellery worn by men was the finger-ring. During the
Republic and the early part of the Empire the type of ring worn was a
status symbol, and laws were passed setting down who could and could not
wear them. In the first century, Tiberius ruled that gold rings could only be
worn by a third-generation freeborn with at least 400,000 sesterces, but by
AD 197 Septimius Severus had extended this honour to all soldiers. Pliny’s
discussion of the wearing of finger-rings in Natural Histories says that at
first they were worn only on the ring-finger, then also on the fore-finger
and then on the little finger:
Gaul and Britain are said to have used the middle finger. At
present this is the only finger exempted, while all the others bear
the burden, and even each finger-joint has another smaller ring
of its own, while others wear only one ring even on that finger,
and use it to seal up their signet ring, which is kept stored away
as a rarity not deserving the insult of common use (33.6.24-5),
(28.4-5).
28 Intaglios and finger-rings. 1 Cornelian intaglio, possibly from Old Capua, Lewis collection,
Cambridge; 2 Impression of an intaglio, Caerleon; 3 Moulded paste intaglio, South Shields
Roman Fort; 4 Hand of man with finger-rings holding wreath, painted linen shroud from Egypt,
Staatliche Kuntsammlungen, Dresden, Germany; 5 Hand of young man with finger-rings holding
a wreath, painted linen shroud from Egypt, Musée du Louvre.
The signet ring was set with an intaglio (a semi-precious stone carved with
a small figure or scene), and was part decorative, part amulet and part
personal seal. Figures of deities were particularly popular because of the
role of the gem as a good luck charm. When Pliny the Younger sent a
nugget of gold in a letter to the Emperor, he ‘sealed it with [his] signet
ring, the chariot-and-four’ (Letters, 10.74.3), so that the Emperor could
see the letter had not been opened and resealed with some other
impression. This is also spelt out in a letter from Egypt, where a man asks
his father to send him weapons, tools and clothes, and adds: ‘if you are
going to send anything, put an address on everything and describe the seals
to me by letter lest any exchange be made en route’ (Sumner 2009, 102).
However, this would not always have been fool-proof as hundreds of seals
carved with the same motif, such as Fortuna, have been found, since there
were only so many themes available. A second-century jeweller’s hoard
found at Snettisham, Norfolk, contained 110 loose intaglios, and seventeen
more already set into silver rings, with Bonus Eventus found on 25 per
cent of them, Ceres on 20 per cent and Fortuna on 13 per cent (Johns 1996,
214, fig. A.2; pl. 3). The demand for intaglios was such that they were
produced in great numbers, with competent but not top quality
workmanship (28.1-2). For the poorer end of the market, there were cast
glass examples with extremely basic motifs (28.3).

BROOCHES

Brooches are not strictly speaking jewellery, as they were not purely
decorative but essential for fastening cloaks, but they could be exploited
for display without the man being considered effeminate. In the first to
third centuries, the most common form of brooch depicted in mainstream
art is a round disc, usually with a central stone, glass, or even cameo
setting (16). In the archaeological record, however, a huge variety of bow
brooches are also known (see Hattatt 2000 for a wide range of types). In
the fourth century, brooches were of the type now called a ‘cross-bow’,
large, inverted cross-shaped brooches that become a symbol of soldiers
and civil servants. Some very fine gold examples, inscribed with the
Emperor’s name, were given by the Emperor to high-ranking soldiers or
civilian officials on special occasions (Hartley et al 2006, no. 73). They
were probably not worn by civilians, but they appear frequently in art
precisely because it is usually soldiers or civil servants being depicted (8,
15.1). As few other fourth-century forms of brooches are known, the
implication is that almost all civilians used mantles or capes rather than
cloaks during this period (6.3).
4
 WOMEN’S CLOTHING

Many items of women’s clothing, such as the tunic and mantle, had the
same basic designs as men’s clothing, and were only distinguished by size
or length. Some items, such as socks, appear to be indistinguishable from
male versions, and indeed surviving examples can be described as male or
female only on the evidence of objects found with them. There were,
however, a few garments worn by men, but not in the general way used by
women, and a few items of distinctively female clothes. The importance of
women’s clothing as an indicator of social status can be seen in Roman
law:
If any-one accosts young girls who are dressed in the clothes of
slaves, his offence will be seen as minor, and even more so if the
women are dressed as prostitutes, and not as [respectable]
mothers of families. Therefore, if a woman is not dressed as a
matron and some-one calls out to her or entices away her
attendant, he will not be liable to the action for injury (Digest,
47.15.15).
In other words, matrons (respectable married women) had more protection
in law than lower class women, but only if they looked respectable.

TUNIC
There were a number of different forms of tunic worn by women, but the
study of them is complicated by the fact that portraits of women usually
show them respectably dressed in their mantles, large rectangles of
enveloping cloth that conceal many of the details of the clothes
underneath. The areas most likely to be hidden – the shoulders and arms –
are unfortunately those that identify the different forms of tunic. Luckily,
the mantle is often only shown covering one shoulder, while women
shown in indoor scenes do not always wear any outer garment.

THE STOLA
The stola is often seen as the symbol of female clothing in the same way
the toga was that of male clothing. It was worn only by matrons, as an
outer tunic over the top of the type of tunic worn by other women.
However, the approximately 100 surviving statues of women wearing the
stola are all dated to the Augustan and early Imperial period, and it is
likely that Augustus deliberately revived use of the garment in the same
way that he promoted the toga for men (Cleland et al 2007, 182). It was
not worn by girls or slaves, and in literature it held great significance as a
symbol of the Roman matron’s honour, her respectability and her moral
character (ibid., 182; Sebesta 1994b, 48-9). However, while the toga
continued to have a ceremonial use long after it ceased to be everyday
wear, the stola was quietly discarded, and fell out of use during the second
century. The stola continued to be used in literature as a symbol of female
modesty for many centuries afterwards, and therefore the term is still
widely known, but despite its fame it was never a long-lived or widespread
fashion during the Empire.
The two distinguishing features of the stola that are usually
mentioned in literature are its length (it always covered the feet), and an
element called the institia. This is still often described as a ‘flounce’, even
though flounces are outside the repertoire of Roman fashion and there is
not a single image of one in Roman art. Rather, the institia was probably a
band woven into, or sewn on, the hem of the tunic that would still be
visible when the rest of the gown was covered by the all-enveloping
mantle (Olson 2008, 30), although coloured borders are not restricted
purely to the stola. It is likely the stola was also distinguished by the
straps across the shoulders that connected the front and back of the dress
(Bieber 1977, 23; Cleland et al 2007, 96).
29 The stola; three portrait busts of Antonia the Younger, first century, Vatican Museum, Musée
du Louvre and Ny Carlsberg Glyptotek.
There were a number of different designs for the shoulder
attachments; the simplest were no more than narrow bands of cloth, often
attached in a position so that there was a loose fold of cloth in between
them that sagged down to form a V-shaped neckline (29). This low
neckline is very characteristic of the stola, and on some examples the
formation of folds carved in stone sculpture suggests that the neckline was
woven (or cut) to form the deep V-shape. A detailed study of the garment
has suggested that most bands are two cords across the shoulder, formed
by having a loop of cord with the cloth pulled through and sewn over
opposite sides of the loop; the junctions were then covered by rectangles
of cloth, forming a horizontal band (colour plate 11; Scholz 1992, Beil. 3).
Unlike the Greek prototype of this type of tunic, the stola was always worn
over an undertunic with elbow-length sleeves. Instead of being belted high
under the breast, as is usual with most tunics, the stola often seems to be
belted at the waist or hips (ibid., Abb. 21, 23, 31, 33).

THE TUNIC

First and second centuries

The gap-sleeved tunic
Freedwomen on Republican tombstones are sometimes shown wearing a
tunic with sleeves fastened at intervals down the arm. Originally a Greek
form of tunic, it was worn by matrons under the stola, but soon became the
outer tunic for upper class Roman women, and is depicted on a large
number of portrait statues and busts. The stola could be made from a
relatively narrow piece of cloth due to the use of bands over the shoulder,
while the tunic had to be made from a wide piece of cloth to create the
elbow-length sleeves. It was made from two rectangles of cloth sewn
together (3.1), but instead of the shoulder seam being sewn up, as in the
male version, it was fastened only at a number of points, the sagging of the
cloth between the fastenings producing the characteristic appearance of
the sleeve (30, colour plate 12).
The fastenings are varied in appearance, but are only c. 1-2 cm across
and are usually either upstanding and roughly spherical, or circular and
flat. They are generally not perfectly round, and often have incisions or
indentations on their surface (31-2). In the past, they were assumed to
have been metal studs or brooches, but their very small size means that
they cannot be brooches, and they are not buttons in the modern sense,
since the two edges of cloth touch but do not overlap. They could possibly
be discs, with a bar or loop on the lower side, sewn onto both edges of the
cloth, and a bust portrait on the Tomb of the Haterii in Rome showing a
‘button’ in the shape of a small flower could support this view (Sinn and
Freyberger 1996, Taf. 5).
30 Tombstone of couple, showing woman wearing gap-sleeved tunic. © Soprintendenza
Archeologica di Ostia.
31 The gap-sleeved tunic, showing details of the fastenings as shown on portrait busts. 1
Faustina the Younger, AD 162-70; 2 Unknown woman, late second century. Both in the British
Museum.
An indication of the colour of these fastenings could help in their
interpretation; however, although this form of tunic appears on many
examples of portrait sculpture, there are very few depictions of it from
coloured wall paintings, and those that do exist are not very helpful. The
clearest examples show the fastenings in a pale colour, whatever the
colour of the tunic, with darker shadows (colour plate 13). This could
either represent light reflecting off metal, or could just be a way to make
the fastenings more visible. It seems strange that, if these fastenings were
indeed made of gold or copper alloy, they are not more often decorated and
treated as an opportunity for display along with other forms of jewellery,
and even more strange that they are not shown on statues as smooth
circular discs or some other regular shape rather than the uneven shapes
depicted. A possible interpretation is that they are not metal, but cloth.
The two edges of the cloth may have been gathered together into a point
(therefore causing the folds seen radiating from each fastening), which
was then sewn round and through, about a centimetre or so below the tip to
fix it in place. The resulting tuft could be left upstanding, or could be
flattened, but either way it usually had folds within it, as modern
experimentation has shown (colour plates 11-2). The sagging folds
created by this form of fastening also help to produce the gap between the
fastenings, which does not happen naturally when the two edges are
simply held together without any form of gathering.

32 The gap-sleeved tunic, showing details of the fastenings as shown on portrait busts.
Unknown women, early third century, British Museum.
There were generally from three to five fastenings to each sleeve, but
up to seven is not unknown. The last fastening was often not at the very
end of the sleeve, leaving short flaps. This type of tunic is shown as being
ground-length, and fastened under the bust by a simple cord.
Short-sleeved tunic
The short-sleeved tunic was made in the same fashion as the male sleeved
tunic, but was simply longer in the body. Tombstones show that it was
worn by some Republican freedwomen, and the fashion seems to have
continued into the first and second centuries for the merchant classes, the
poor and, therefore, also slaves. On some images, it could be argued that
the short-sleeved tunic was worn belted, but on a number it is clearly
unbelted. A wall-painting from Pompeii shows a serving girl in a tavern in
an unbelted tunic, as does a relief of the midwife Scribonia Attice from a
mid-second-century tomb in Ostia (33.1). The musician and some of the
mourners on a relief from the late first- or early second-century Tomb of
the Haterii from Rome also wear unbelted, short-sleeved tunics (33.2). The
tunics of the serving girl from Pompeii and Scribonia’s assistant are calf-
length, which is probably a distinction of class, as shorter tunics required
less cloth (much stress, for example, was put on the fact that the stola of
the upper-class matron covered her feet).
Colour
Both the Greeks and Etruscans, who influenced Roman costume, loved
patterns, borders and bright colours, and in the second century BC, Plautus
mentions the bordered gown, the ‘royal’ or ‘exotic’ tunic and the feather-
patterned in a list of different tunic types (Epidicus, 2.2.228-34). In the
late first century BC and the early first century AD, there are references to
segmentata, bands of patterned decoration, which might be the broad
bands round the hem of tunics, as shown in colour plate 13.

33 Unbelted short-sleeved tunics of the first and second centuries. 1 Midwife Scribonia Attice
and assistant, second-century bas-relief, Ostia, Italy; 2 Musician at funeral, bas relief from the
Tomb of the Haterii, Vatican Museum.
Female costume during the first few centuries of the Empire appears
to make much less use of borders and pattern, but this may result purely
from the bias of the surviving evidence. The most common forms of
decoration were simple stripes of colour, usually two narrow stripes down
the front and back of the tunic. Ovid recommends purple stripes for a pale
woman (Art of Love, 3.270), and a portrait of a woman from Pompeii has a
dark brown stripe on a red-brown tunic. Portraits from Egypt show tunics
with two dark stripes down the front, such as red on green and black on
purple (colour plate 14).
While men seemed to have preferred white tunics, women were more
likely to choose a wider range of colours (Mannering 2000, 285; Olson
2008, 13-4). Linen and silk could only produce pastel colours, but wool
could provide vivid hues: Ovid mentions woollen clothes (presumably
tunics) with borders, in purple, sky-blue, water-coloured, saffron, green,
amethyst, white, chestnut, almond and wax-yellow (Art of Love, 3.169-88).
Some colours, such as violet and yellow, seem to have been closely
associated with women (Olson 2008, 12).

Third and fourth centuries

The short-sleeved tunic
The gap-sleeved tunic continued to be depicted in art until the late third
century. Sometimes this can be explained as examples of Imperial women
wearing ‘traditional’ clothing, or of a new head being put on a pre-existing
statue wearing the old style of tunic, but there are a few examples where
this is clearly not the case, such as the Annona sarcophagus portrait, dated
to the late third century (9.1). However, this betrothal or marriage scene is
unusual, in that not only is the woman not wearing a bridal mantle, as was
usual in such scenes, but that her tunic is slipping down her shoulder in a
way not seen elsewhere with the gap-sleeved tunic. Usually the bare
shoulder is the reserve of goddesses, or women dressed as goddesses, in
tube-dresses fastened by a single brooch at the shoulder, so the exact
interpretation of this scene is therefore not fully understood.
Certainly by the early third century, the women of the Imperial family
had started to wear a new style of tunic which would have made the gap-
sleeved tunic quickly appear old-fashioned. The reliefs on the Arch of
Severus in Lepcis Magna, dated AD 203-4, show the Empress Julia Domna
wearing a gap-sleeved tunic in one scene, and a tunic with fully sewn
sleeves in another (34.1). A bust dated to the early third century has a
similar short-sleeved tunic, probably with striped decoration (34.2). Like
the earlier gap-sleeved tunics, it was fastened high under the bust with a
cloth belt (34.1).
34 Short-sleeved tunics of the third century. 1 Julia Domna, bas relief from the Arch of
Septimius Severus, Lepcis Magna, AD 203; 2 Unknown woman, early third-century bust, Museo
Archeologico Nazionale Napoli, Italy.
Dalmatic
Women also adopted the wide-sleeved, unbelted dalmatic tunic that was
fashionable from the late third century (35.1). The late third- or early
fourth-century catacombs in Rome show a number of Christian women
wearing this style of tunic. Most show women in pale coloured dalmatics
with dark stripes, although the reverse is occasionally known. Usually the
stripes are plain, but a picture from the Cemetery of Priscilla, and one
from the Cemetery of Traso, both show stripes with a decorative pattern
(both women have hairstyles of the type fashionable c. AD 270-310; 35).
Figure 36 shows two styles of decorated tunic stripes with dark and pale
backgrounds. These are both from male tunics, but they give an idea of
some of the simpler patterns available at this time (and can be compared
to the patterns on female tunics shown in 35). Although women
occasionally have decorated stripes, they are never shown during this
period with the roundels used on the male tunics of the same period.
35 Christian women at prayer, wearing unbelted tunics. 1 Late third-century wall painting from
the Catacombs of Priscilla, Rome; 2 Early fourth-century wall painting from the Catacomb of
Traso, Rome; 3 Early fourth-century wall painting from the Catacomb of the Giordani, Rome.

36 Tunic decorations from Egypt in purple and white, California Academy of Sciences. 1 Fifth-
century tunic fragment; 2 Fourth-century tunic fragment.
 More common than the full-length version (at least in the paintings
from the catacombs) is the short tunic, with a curved lower edge reaching
to about the mid-calf. The length may again be an indication of social
status, as some are worn by slaves (colour plate 8, left-hand figure).
Variant versions can have comparatively narrow or tight sleeves (35.1,
35.3).
A mosaic at Piazza Armerina of the early fourth century shows the
mistress of the house on the way to the baths, in the company of her sons
and two female slaves (colour plate 8). The mistress wears an ankle-
length striped tunic with very wide sleeves, while the two slaves wear
mid-calf length versions. The tunic of the mistress may possibly be belted
under the bust, but as she seems to be wearing a mantle draped round her
neck with both ends hanging down her back (a fashion seen more clearly
on the mistress on the Proiecta Casket, 37.1), any possible belt is
obscured. One of the slaves, however, quite clearly does have her tunic
belted under her bust, pulling in her full sleeves at the same time. Some of
the slaves on the Proiecta Casket also wear their tunic belted, as does a
slave in the tomb-paintings of a grave in Silistra (dated to the late fourth
century). Both sets of slaves have tied in their long sleeves, as with the
Piazza Armerina slave, producing long loops, like wings (38). It is likely
that during this period, slaves, or those who had to work, wore their tunics
belted at times, to keep their full skirts and sleeves out of the way, while
the rich had no such need.
If the mistress depicted in the Piazza Armerina mosaic is using a belt,
it is not used to tie her sleeves back at the same time, as they hang down
loose, suggesting that the belt would have to be fed through holes into the
tunic to by-pass the sleeves, as appears to happen on two women making
rose garlands in another mosaic in the villa (Capizzi and Francesco 1989,
83). However, it is possible the sleeves may be narrow near the shoulder
and wider at the wrist, as shown on an image of the biblical Suzanna at a
cemetery at Salonika (Donati 1998, fig. 71). This depiction is one of the
latest examples of the dalmatics worn by a woman, dating to the late
fourth or early fifth century. This form of sleeve deals with one drawback
to the wide sleeves of the typical dalmatic; they can easily expose a large
part of the body. It seems likely that an undertunic usually had to be worn
underneath.
37 Late forms of tunics, mid-fourth-century reliefs in silver from the Proiecta Casket, British
Museum. 1 Proiecta being escorted to the baths; 2 Proiecta putting on make-up.

38 Slaves with belted tunics. 1 Mid-fourth-century relief in silver, Proiecta Casket, British
Museum; 2 Fourth-century tomb wall painting, Silistra, Bulgaria.
Colour
The same wide range of colours current in the first and second centuries
was also used in the third and fourth centuries. The dalmatics could be
either pale or dark coloured, although the decorated stripes are usually
shown as purple and white. The mistress from Piazza Armerina is shown
in a tunic of pink, with broad dark brown and white stripes down the front
of the gown and round the lower part of the sleeves. The slaves
accompanying her also have stripes on their tunics, although theirs are
much narrower; one has a yellow tunic with red and black stripes, and the
other a dark pink tunic with brown stripes (colour plate 8).

Late fourth century

The Proiecta casket, dated roughly to the period AD 330-370, shows the
mistress – Proiecta herself – in three different costumes. The portrait bust
shows her wearing a tunic decorated with an all-over pattern of dots,
belted high under the bust, with long, tight-fitting sleeves (15.1). In the
procession to the bath-house, she wears a full-length dalmatic with a
mantle draped round her neck (37.1). In the final scene, Proiecta is sitting
in a chair, probably applying make-up, rather than doing her hair as is
often suggested (Croom 205, 124), wearing an undertunic with long tight-
fitting sleeves, and a sleeveless overtunic (37.2). It is possible that she has
not yet got round to putting her belt on (as in the portrait scene), but
Proiecta is otherwise fully dressed, including her jewellery, so perhaps this
type of overtunic was worn unbelted. However, another woman also shown
informally at her toilette (in this case putting on her jewellery) is also
shown in an unbelted sleeveless tunic (Blanchard-Lemée 1996, fig. 116).
The fashion for two tunics can be seen more clearly on the portrait of
Serena (c. AD 400). She has an inner tunic with long, tight-fitting sleeves
and a wider tunic over the top fastened by a jewelled belt (39). This
fashion of tight-sleeved inner tunic and looser overtunic mimics that of
fourth-century men (10.6). The fashion is once more for a belted and
ground-length tunic, which fits in with the general late antique trend for
women to be almost completely covered.
39 Serena and her son Eucharius, late fourth-century ivory diptych. © Monza Cathedral
Treasury.

Fifth to sixth centuries

The details of the tunics of the women of the court of Theodora are not
fully clear because of the short mantles they are wearing, but the cuffs
occasionally visible at the wrists suggest they have long, tight-fitting
sleeves (colour plate 2). They may be wearing tunics with long, tight-
fitting sleeves without overtunics, as shown in the fifth-century painting of
Hilaritas (Veyne 1987, between pp 274-5), but a couple of sixth-century
scenes of Christian stories in the Church of San Apollinare Nuovo suggest
that the fashion for the long-sleeved inner tunic and short-sleeved
overtunic still held sway (Bustacchini 1988, 109, no. 5; 115, no. 4). These
women could, of course, be shown in ‘old-fashioned’ clothes as they are
characters in Biblical stories, but their head coverings are very similar to
those worn by Theodora’s court, suggesting that their dress is also
contemporary.
The biblical characters have simple, striped tunics, but the court
women have more elaborate dresses with decorated stripes, a border round
the lower edge, or even roundels, previously seen only on male tunics
(colour plate 2). All of them wear tunics woven with an all-over pattern,
reflecting the taste in the late Empire, imported from the East, for the
ornate, the colourful, and the bejewelled. Theodora herself wears a very
different costume, based on male military dress, which stresses her status
as someone separate from the women of even the highest class who make
up her entourage (perhaps dressing in a ‘mannish’ fashion to indicate that
she had the power usually reserved for men).
Colour
Most women wore tunics in one colour, with stripes down front and back.
The Ashburnham Pentateuch (sixth or seventh century in date, but
featuring women in dresses fashionable from the late fourth century)
shows Rebecca with a red, long-sleeved undertunic, and a dark blue
overtunic decorated with two white stripes with an orange vine scroll
pattern, fastened with a red belt (Weitzmann 1979, pl. XIII). The rich used
cloth with all-over patterns; for example, the ladies of Theodora’s court
wore tunics with a geometric pattern, blue birds against a pale pink
ground, red birds against green and a foliage pattern.

UNDERTUNICS

Various ancient authors use a number of terms for clothing which

apparently refer to a type of undertunic (Olson 2003, 202-3; Olson 2008,
52). In the Republic, some women are shown clearly wearing an
undertunic, visible at the neck, but it is clear the undertunics were not
usually worn with the gap-sleeved tunic, as they would otherwise be
visible between the fastenings. Undertunics are sometimes seen with the
long-sleeved tunics (35.3), but even then they were not universal. In colour
images, they are shown to be white (almost certainly linen), often with
coloured decoration round the neck. A number of the Egyptian mummy
portraits show the decorated neckline (e.g. Walker and Bierbrier 1997, figs
91, 93-4, 100), while a full-length painted shroud shows a white
undertunic with decorated neckline and a short fringe just visible under the
upper tunic (colour plate 15). A surviving example, with both contrasting
decoration at the neckline and a fringe at the hem, was found in a cemetery
in Jordan (Granger-Taylor 2000, fig. 10).
Gradually the undertunic becomes more visible with the introduction
of the long-sleeved undertunic worn with the elbow-length upper tunic
sometime in the fourth century. The undertunic can be white, suggesting it
was linen, but in the very late period it could be coloured and have
decoration round the wrist (Busticchini 1988, 115, no. 4, green undertunic,
with red stripes). A scene in the sixth-century Vienna Genesis shows
Rebecca wearing not only a pink tunic with a matching pink inner tunic,
but also a glimpse of a white undertunic decorated with a black pattern
along its hem (Weitzmann 1977, pl. 24).

BELTS

Roman women always tied their tunics high under the bust, rather than at
the waist. Depictions of women wearing their tunic belted at the waist or
hips are likely to be of goddesses, their priestesses, mythological figures,
or figures copied from Hellenistic models. In the first to third centuries,
belts were no more than twisted cord or cloth, tied in a reef knot with only
short ends hanging down (34.1; 55.2; cf 63.1), in general probably in a
contrasting colour to the tunic. The rich freedwoman Fortunata wore a
‘yellow-green small belt’ over a cherry-coloured tunic (Petronius,
Satyricon, 67.4), while the slave Fotis wore a reddish band over a linen
tunic (Apuleius, The Golden Ass, 11.7).
In the late third and fourth century, the belt, when worn, was still
often no more than a cord (colour plate 8). There was, however, a growing
fashion for elaboration, as can be seen on a painted shroud from Egypt
showing a woman in a purple tunic wearing a belt of red cloth fastened at
the front by a large circular central gem set in gold. This could perhaps be
a brooch, or the gem and its setting could have had a bar across the back
that the cloth was tied onto; either way, the two narrow ends of the belt
hang below the fastening (40). A fourth- or fifth-century fragment of
cloth, now in Berlin, shows a woman in a purple tunic with yellow/orange
stripes (possibly meant to represent gold) wearing a thin, pale red belt
fastened with a large, oval, dark blue or black stone set in gold (von Kalck
1996, no. 365). A letter written by Jerome in AD 385 about a recent
convert records: ‘no gold and jewels adorn her belt; it is made of wool,
plain and scrupulously clean. It is intended to bind her clothing rather than
cut her in half ’ (Letters, To Marcella, 38. 4).
When two tunics became fashionable by the end of the fourth century,
the tie became wider and stiffer, probably made out of leather. Serena
wears a belt decorated with jewels and edged in pearls, fastened by a
central rectangular plate set with an oval stone (39), and a fifth-century
wall painting from the Catacombs of St Januarius in Naples shows a young
girl in a cream-spotted red tunic, with a decorated belt with one very large
central blue or black gem, with further gold set gems and pearls to either
side (Veyne 1987, between pp 274-5). Claudian mentions green jasper as a
stone used to decorate belts (On the Sixth Consulship of Honorius, 523-8).
The waist or the hips are a more natural place to fasten a dress than
under the bust, as the hips or buttocks stop the belt slipping down any
further. Under the bust, however, unless a woman is pregnant, there is
nothing to stop the belt sliding down the ribs. There is no evidence of the
belt being threaded through loops on the tunic, so it is possible it was sewn
onto the dress at intervals to keep it in place. Otherwise, the belt had to be
tied extremely tightly, or the woman had to keep re-adjusting it.
40 Tunic with ornamented belt, painted linen shroud from Egypt. Musée du Louvre.

DINING-CLOTHES

Women changed into a different set of clothes when eating in company in

the same way as men. The dining-clothes consisted of tunic and mantle,
which may possibly have been of the same colour; Martial refers to a man
giving his mistress a gift of ‘leek green dining-clothes’ instead of buying a
toga for Martial himself (Epigrams, 10.29). The Digest, a series of
questions and answers on legal queries, also mentions female dining-
clothes. In this case, a woman called Sempronia Pia had been left Tabian
mantles and three tunics with small mantles (palliola) which, by the terms
of the will, she was allowed to pick out herself. The question seems to be
whether she could only pick the clothes from the individual tunics and
mantles of the deceased woman, or if she could choose them from sets of
dining-clothes, with the implication that there was some distinction in
value between the two (Digest, 34.38.1).
MANTLE, CLOAK AND CAPE
MANTLE

Male and female outer garments had originally been very similar,
consisting of long lengths of cloth draped round the body, but while the
male version was transformed into the toga, the female version remained
very much the same throughout the Roman period. The palla was a large
rectangle of cloth without the curved lower edge of the toga, which
covered the body from shoulder to knee or lower calf. It had presumably
once been intended as protection against bad weather or the cold, but it
soon became a necessary covering for a modest woman, and no
respectable woman would leave her house without her body concealed by
it. In the first century AD, there were complaints about how voluminous
mantles shrouded women (Cleland et al 2007, 136). Women are frequently
shown with the mantle pulled up over the head (41), and not simply when
they are making a sacrifice, which is when men are shown with covered
head (although in their case by their toga or cloak, but not the mantle),
suggesting women covered their heads out of modesty.
The mantle (palla), like the toga, needed at least one hand to keep it
in place. A cloak fastened on the shoulder by a brooch allows both hands
to remain free, but only in very rare circumstances did women wear a
cloak (see below). The palla was no more suitable for working in than the
developed forms of the toga, and while that was not a problem for the rich
who did not work, and who had slaves to carry objects for them, it would
have been cumbersome for the majority of the population. In the first
century, there are a number of depictions of women with their mantles tied
in a great knot on their hips to keep their hands free, but these are usually
religious scenes, with the women playing musical instruments or making
offerings, which may be based on Hellenistic prototypes, and it may not
have been an everyday method of wearing the mantle (eg Nappo 1998, 28,
154; Bandinelli 1970, fig. 141; Strong 1976, pls 52-3).
In the Republican period, the mantle was not so very different to the
toga (cf 41.1 and 10.1), and while there were a number of different ways of
draping it, the mantle stayed the same basic shape over the following
centuries, and never developed the complexities of the toga. One end was
draped over the left shoulder, and then brought round the back, where it
could either be placed over the right shoulder, concealing most of the
figure, or brought under the right arm and then flung over the left shoulder
or draped over the arm (41).
The large mantle seems to have continued throughout the Roman
period, but generally there was a tendency for a smaller, less all-
enveloping form. Egyptian mummy portraits sometimes show a mantle
being worn either like a shawl over both shoulders (42), or with one end
hanging down over the left shoulder and the rest brought round the back,
under the right arm and then across the body to the lower left arm (rather
than back over the shoulder). The mantle across the front of the body is
often shown draped in a curve (colour plate 15; Walker and Bierbrier
1997, no. 114). It could be decorated with the L- and H-shapes also found
on male mantles or, from the third century, roundels (42). A second or
third-century red woman’s mantle from the cemetery at Khirbet Qazone,
Jordan, had four L-shaped decorative elements, with additional paired
arrowheads (Granger-Taylor 2000, figs 8-9, colour plate 3, no. 9). The
mantle was at least 200cm wide; a complete example, from the Cave of
Letters, was 270 cm by 140 cm, and a man’s mantle from At-Tar 250 cm
by 162 cm (ibid., 152, fig. 12).
41 Women wearing the mantle. 1 Statue of unknown woman and daughter, c.50 BC, Palazzo dei
Conservatori, Rome; 2 Second-century statue of Vestal Virgin, Museo Nazionale, Rome; 3
unknown woman, late second-century statue, Museo Capitolino, Rome.
From the later third century, there are a number of depictions of a
new form of mantle, with a wide, decorative border, as shown in a gold-
glass portrait from Brescia (colour plate 16). A painted marble bust has
similar decorated diamonds and ovals (Haufmann 1967, fig. 321), while a
fourth-century bust has a single vine pattern, and a gold-glass portrait has
two vine patterns on the mantle (Weitzmann 1979, 407, no. 363; Harden
1987, no. 157). The mother in the Brescia portrait has a different type of
mantle, tied in a large knot on her breast (colour plate 16). Other gold-
glass examples of this method of fastening the mantle show the short ends
hanging down below the knot (Pillinger 1984, Taf. 7, Abb. 70, 236).
42 Decorated mantle, fourth-century painted linen and plaster mask from Egypt, Royal Museum
of Scotland.
The mantle shown with the fourth-century full-length tunic is much
narrower than the earlier form. It was worn so that there was no visible
cloth hanging down the front; either one end was very short and hidden
under the drapery above it, or both ends were thrown back over the
shoulders; either way, the shoulders and upper arms were covered by an
arc of cloth. Proiecta is shown wearing her mantle in this way, and so
perhaps is the mistress in the Piazza Armerina bath mosaic (colour plate
8; 37.1).
The small mantle continued into the fifth century and beyond.
Serena’s is rolled heavily at the top and is no more than knee length (39),
while the women of the court of Theodora wear ones that are no more than
shawls (colour plate 2). Here, they are worn with one end hanging down
the left side, with the other end taken round the back, over the right
shoulder and thrown back over the left shoulder again. Longer mantles
were also still in use, as shown by some of the women in the biblical
scenes of the Basilica of San Apollinare Nuovo (Bustacchini 1988, 115,
no. 4; 116, no. 1).

Colour
Frequently, the mantle was the same colour as the tunic underneath,
suggesting the concept of a set of clothing, presumably bought or made at
the same time (cf colour plate 14). First and second century portraits from
Egypt show purple mantles and tunics, and sixth-century mosaics show
brown tunics and mantles worn together (Walker and Bierbrier 1997, nos
34, 37; Bustacchini 1988, 116, no. 1). If the mantle was in a contrasting
colour, it could be either darker or lighter than the tunic; portraits from
Egypt show purple with a white tunic, white with a dark blue tunic and
orange with a green tunic (Walker and Bierbrier 1997, nos 51, 86, 111).
Third- and fourth-century mantles sometimes had purple and white
roundels, while fifth-century examples could be patterned. The mantles of
the ladies of Theodora’s court are in a mixture of colours; white with an
overall brown geometric pattern and gold roundels; plain white with gold
roundels; orange with blue dots in sets of three; gold with an all-over
pattern of green crosses imposed on red circles and gold with blue and
black crosses or stars (colour plate 2).

Other mantles
Small versions of the palla were also used (palliola). Other forms of
mantle were called amictus, amiculum and amictorium, meaning garments
thrown on or around a person; these may well simply be generic terms for
a wrap rather than being distinct forms of mantle. Women could also use
the endromis, the thick mantle worn by athletes, as shown by one of the
exercising ladies from the Piazza Armerina mosaic (colour plate 18).

CAPES

Although women are very rarely shown wearing capes, it is clear from the
literary evidence that they could do so when travelling, and presumably
during bad weather. Claudius’ wife Messalina also wore a shoulder cape
(cucullus) when creeping out at night, ‘wearing a night cape and attended
by no more than a single maid’ (Juvenal, Satires, 6.118), while
Diocletian’s Edict refers to capes (caracallae) of coarse linen for the use
of women of the lower classes. The Edict also refers to linen ‘hoods’
(mafortia) worn by women, but as they are listed as being dyed with
purple, they are likely to have been more fashion items than purely
functional, and it is perhaps more likely that they were simply shawls, a
narrow form of mantle that was worn round the shoulders and over the
head. The early fourth-century writer Nonius Marcellus equates the
mafortium with the earlier ricinium, or head veil, and calls it a woman’s
short small mantle (Nonius, 14). One of these ‘hoods’ in the Edict is
described as having vertical stripes made using one pound of purple, and
costing 55,000 denarii, one of the most expensive garments listed in it.

CLOAKS

Women did not, on the whole, wear cloaks in the Mediterranean region.
Some priestesses fastened their mantles with a brooch, but they did not
wear them in the manner of men’s cloaks, with the cloak covering both
shoulders and fastened on the right side. Instead, the cloth was lopped
under one arm, and fastened on the opposite shoulder (cf 55.4). Poppaea,
Nero’s mother, had a general’s cloak made for her, and Theodora is shown
wearing a man’s cloak (colour plate 2), but both are Imperial women and
are ‘honorary men’ because of their position.

TOGAS

Literary references suggest that young girls as well as boys wore the toga
praetexta, although this is not reflected in Roman art, with only fourteen
examples known (Olson 2008, 17). In the first century, there are a number
of literary references to prostitutes and adulteresses wearing the toga
(ibid., 47-51). Martial, for example, suggests that a toga would make a
more suitable gift for a notorious adulteress than expensive scarlet and
violet clothes (Epigrams, 2.39), and Horace contrasts a ‘toga’d maid’ with
an honourable matron (Satires, 1.1.62-3). There seems to be no evidence
that prostitutes had to wear the toga, only that they were the only women
who could. It is not clear why they wore it, or why they would want to; the
first-century toga was even more cumbersome and concealing than the
palla worn by respectable women. Perhaps the prostitute’s version of a
toga was smaller and lighter, or more akin to the Republican toga, but it
seems unlikely that a profession more used to wearing transparent Coan
silk would take kindly to the awkward toga, nor the joys of constantly
taking it off (folding it carefully to avoid creases), and putting it back on
equally neatly, an operation that usually required the help of a second
person.

UNDERWEAR
The extent to which underwear was regularly worn is not clear, for while
bras and briefs are known from images of athletes, acrobats, and
prostitutes, they are functional outerwear rather than underwear.
Nevertheless, the bra seems to have been an accepted part of female
underwear, although there is less evidence for briefs. Cicero’s description
of P. Clodius dressed as a woman in the first century BC lists the clothes
he wore, and on two different occasions mentions both breast-band and leg
coverings, but not briefs (Response to the Soothsayers, 43-4; Against
Clodius and Curio, 15.21-4). Ovid mentions both the breast-band and leg
coverings as hiding places for love letters (Art of Love, 3.621-4), while
Martial’s description of Lesbia’s tunic catching between her buttocks
would also suggest that no briefs were being worn (Epigrams, 11.99.1-5).
When Lucius’ lover in The Golden Ass undressed, he says ‘without delay
she stripped off her clothes, even down to the band that bound up her
breasts’ (10.21) without mentioning the removal of any briefs, and a
woman shown in the process of taking off her tunic on a silver bucket in
Naples Museum has no briefs on underneath (Borriello et al 1986, 98, cat.
63). It is possible women more often wore some form of briefs for the
duration of their periods, but even this may not have been a necessity.

BREAST-BAND

The breast-band was a long length of cloth or leather, wrapped round the
breasts with the end tucked in; the statue of Venus and a mosaic of a
nymph from France show them wrapping or unwrapping their breast-bands
(43, 44.1). In the late Republic the word for breast-band was strophium,
but by the first century AD the term fascia (or fascea) had become more
common. The word simply means strip of cloth or bandage, and it could be
used of bands of cloth used for any purpose, including lower leg coverings.
 The breast-band is often found on depictions of mythological women
connected with love or sex, such as Venus and nymphs, as well as human
lovers and prostitutes at work (eg Capizzi and Francesco 1989, 88; Nappo
1998, fig. 121). While it is not surprising that the prostitutes are not
wearing briefs, the mythological women are also otherwise naked, or
simply have a mantle draped round their hips, suggesting that the
combination of bra and briefs was not the usual ‘set’ of underwear as it is
nowadays.
Experiments by members of the re-enactment group Quinta have
shown that to stay up, the cloth needs to be as much as 5 m long, so that it
can be firmly wrapped round the breasts six or seven times (depending, of
course, on the size of the woman involved). Breast-bands of a shorter
length have a tendency to slip down, even when walking, let alone taking
part in energetic activities such as athletic exercises. The length of the
breast-band may possibly be indicated by a reference in the Scriptores
Historiae Augustae to the birth of Clodius Albinus, where a bandage was
long enough to wrap a baby in:
It was customary in his family that the bandages in which the
children are wrapped should be of a reddish colour. In his case,
however, it chanced that the bandages which had been prepared
by his mother during her pregnancy had been washed and were
not yet dry, and he was therefore wrapped in a breast-band
(fascea) of his mother’s, and this, as it happened, was purple
(SHA Clodius Albinus, 5.9).
43 Statuette of Venus unwrapping her breastband. © Museum Burg Linn, Germany
The word fascea could of course refer to other forms of bandage, as noted
above.
Martial’s gift tag for a breast-band suggests that it could be used to
produce a bit of cleavage: ‘band, compress my lady’s swelling breasts, so
that my hand may find something to clasp and cover’ (Epigrams, 14.134),
but generally the breast-band squashes and flattens the breasts. Ovid, in
fact, suggests that the sight of a large-breasted woman unrestrained by a
breast-band would cure a man of love (Remedies of Love, 338). Of course,
for some women, flattening the breast was the last thing they wanted to do.
There are a number of depictions (generally of mythological women)
where it is clear that the breast-band is wrapped round under the breasts;
assuming this is not simply a Hellenistic convention, it may be that the
band was used in this way to give a little uplift to the bust (44.2).
The breast-band could also be worn as outer wear by athletes, usually
in conjunction with briefs (colour plate 18).
BRIEFS

Briefs were worn by women in the bath-house. Before having a bath, the
Romans liked to work up a good sweat, and depictions of female ‘athletes’
all come from bath-houses, where women could exercise in relative
privacy. Martial mentions a woman playing ball games wearing briefs, and
another woman, being modest, wearing briefs while in the bath-house
(Epigrams, 7.67; 3.87), although it is not clear if the briefs were also worn
in the water. The briefs could be worn alone, or in conjunction with a
breast-band. The mosaic from the family bath-house at Piazza Armerina
shows ten women exercising, eight of them in bikinis holding, among
other things, the palm leaf and wreath of a winner, suggesting the games
were in earnest (colour plate 18). Although some of the bikinis are all red,
others have green or blue breast-bands with red briefs. The briefs are not
tied at the sides, and it is not clear how they stay up without ties or any
form of elastic, unless they are extremely tight.
Other images show the briefs being worn by themselves. Stucco
reliefs of women from a third-century private bath-suite in Carthage show
four women wearing small, black, openwork briefs tied at the sides, with
the ends dangling down to mid-thigh or below (44.3). A very similar
garment has survived in waterlogged conditions at Shadwell, in London. It
is made out of leather in an hour-glass shape, with integral ties on both
sides and an openwork design on the front that may possibly have also
incorporated a central decorative metal fitting (van Driel-Murray 2002,
57-60). Fragments of at least five other pairs have been found in London;
the most complete example of these is made in a slightly different way to
the Shadwell pair, with hemmed edges and the ties sewn on separately, but
without any decoration (Wilmott 1982, 52-5; colour plate 17).
44 Breastbands and briefs. 1 Female entertainer, fourth-century mosaic, Hippo Regius, Algeria;
2 Woman in bed, first-century wall painting from Pompeii; 3 Stucco reliefs of three women from
a private bath-house in Dougga, Tunisia, possibly third-century.

LEGGINGS
Statues of the first and second centuries usually show upper-class women
in sandals without any form of sock, but it is clear that women would wear
some form of puttee on occasion. The word used is fascia or fasciola
(little strip/bandage), suggesting some form of puttee or piece of cloth,
folded round the foot in the modern Russian manner. Cicero refers to a
‘woman’s house-shoes and purple puttees’ (fasceolis; Response to the
Soothsayers, 21.44), and Ovid might be referring to puttees when he
recommended that ‘lean ankles should never be released from their
bonds’, when referring to ways a woman could make herself attractive (Art
of Love, 3.271-2).
Women also wore proper ankle-length socks. The word for sock,
soccus, is also taken to mean ‘slipper’ (as well as being the traditional
footwear of a comic actor), and it is sometimes difficult to tell which is
meant. Catullus mentions flame-coloured socci worn by a bride (Poems,
61.8-10), while Diocletian’s Edict refers to purple, white and gilded socci
in the section on footwear, although without identifying them as either
male or female items of dress (18-23). Pliny may well be referring socks
when he describes women fixing pearls ‘not only to the straps of their
sandals but all over their little socci’ (Natural Histories, 9.56.114). A
third-century painted shroud from Egypt shows a woman wearing socks
with her sandals; dressed in a calf-length cream tunic and mantle and
white undertunic, she has bright red socks (colour plate 15). The second-
century graves from Les Martres-de-Veyre in France also produced a pair
of ankle socks, as well as a pair of knee-high stockings (45). The long
socks, like the ankle socks, were made of pieces of shaped woollen cloth
sewn together, and would have had a narrow cord laced through the top of
them, below the fringe, to act as garters. The ankle socks have a strap to
cross the front of the foot, but there is no surviving method of fastening
them. It is unfortunately not clear how often such long socks were worn, or
if short and long socks were worn according to the season, or if the fashion
for long socks was restricted to the north-west provinces, as the evidence
in art is concealed under long tunics. Full length trousers or hose,
however, were quite firmly the preserve of the man. From the late third
century, enclosed shoes or ankle boots came into fashion, and there is a
lack of evidence for leg coverings. It is unlikely, however, that either the
need for socks or their design changed much.
HAIRSTYLES AND HATS
Hats were rarely worn by women in the Mediterranean region, but instead
women wore their hair in elaborate hairstyles. These styles changed so
frequently during the Empire they can be used to date statues or other
artwork. The hairstyles are depicted on the portraits of Imperial women on
coins, which can be dated from historical sources, and thus a date range
for any particular hairstyle can usually be produced, at least as worn by the
Imperial family. There are drawbacks to this; the coins of the women
cannot usually be dated as closely as those of the Emperors, and often they
were issued posthumously, so that it is not clear whether the hairstyle is
the one used in life, or simply the current style. There are also some reigns
when no coins with female portraits were issued, or the women are shown
with a mantle over their heads, and there are some hairstyles known to
have been worn by women that are not shown on coins at all.

45 Woollen knee and ankle socks from a second-century grave at Les Martres-de-Veyre, France.
Although there are a number of changes in the styles, these changes
occurred over five centuries or more. Some styles could last for twenty or
thirty years, so that it was quite possible for some women to wear the
same hairstyle most of their adult life. Only a few of the Imperial women
are shown with more than one hairstyle, and it is likely that new fashions
were introduced with a new Empress to distinguish herself from her
predecessor. Women wishing to be fashionable, wherever they lived in the
Empire, could use the official statues and coin portraits to follow the latest
hairstyles from the Imperial court if they so wished. Some women
presumably continued to wear the fashions of previous empresses, just
because they were used to them and did not like change.
Literary evidence also suggests that hairstyles could be more varied
than the art evidence suggests. Writing about AD 1, Ovid commented:
‘Nor is there but one form of adornment; let each choose what becomes
her … An oval face prefers a parting upon the head left unadorned ...
Round faces would fain have a small knot on top of the head, so that the
ears show. Let one girl’s locks hang down on either shoulder … Let
another braid her hair like girt-up Diana … This one it beseems to let her
waving locks lie loose; let that one have her tight-drawn tresses close
confined; this one is pleased by the adornment of the Cyllenian tortoise-
shell; let that one bear folds that resemble waves … Even neglected hair is
becoming to many’ (Art of Love, 3.135-54).
Women lower down the social scale probably did not bother with the
more ornate hairstyles of the rich. Slaves are shown as having simple
hairstyles, probably pulled back into a plain bun, in contrast to the
fashionable styles of their mistresses. In the Piazza Armerina bath
mosaics, the mistress is shown with her wide plait sitting on the top of her
head, while her two accompanying slaves have much simpler styles
(colour plate 8). In other mosaics within the villa, spectators at a fight and
wreath-makers all have plain hairstyles, while only some mythological
Naiads have the plait on the top of the head, complete with decorative pins
(Capizzi and Francesco 1989, 75, 79, 81, 83). The rich lady putting on a
necklace in the fifth-century Sidi Ghrib mosaic has a striped cloth cap
covering her hair, while her two slaves have uncovered hair (Blanchard-
Lemée 1996, fig. 116).
Coin portraits are almost exclusively side views. More details of the
hairstyles, including front and back, can be obtained from portrait busts
and statues. These often show the hair in great detail, including faithfully
reproduced small plaits and curls, and yet they very rarely show any
hairpins. One of the few statues to show a pin is a Trajanic bust with a
single pin across the back of a plaited bun; another three examples also
show a single pin each (Breve 2003, fig. 86; Stephens 2008, 117). It is
possible to push pins in far enough to hide them, but pins with pearls,
beads, or ornately carved heads were clearly meant to be seen. The stone
portraits also rarely depict necklaces or ear-rings, so it could simply be a
convention of portrait art not to show these details, but a recent study has
shown that the ornate hairstyles could be produced very easily by sewing
the hair into position without the need for pins; some of the hairstyles
could probably not be created any other way (Stephens 2008).
All adult women would have had long hair, the only exceptions being
women who habitually wore wigs, slaves who had had their own hair cut
off to provide the hair for such wigs and possibly some priestesses, as well
as women who had been ill. Wigs or hair-pieces are mentioned in the
literary record, presumably worn by those women who did not have
enough of their own hair for the more elaborate hairstyles (Olson 2008,
74). Women would not have appeared in public with short hair from
choice. The hair was always worn pinned up, unless the woman wanted to
show distress by not taking care of her appearance, such as at funerals.

c. 40 BC – c. AD 14
The hairstyle of Livia (wife of Augustus) consisted of a strip of hair on the
top of the head brushed forward over the forehead and then folded back,
forming a knot (nodus) at the front, held in place by a tie or pin. The strip
of hair then crossed the top of the head in a band, while the hair round the
face was waved and lightly twisted towards a low bun, usually made up of
a thick plait or plaits with a plait wrapped round its base (46.1).
46 Hairstyles of the first and second centuries. 1 40 BC – AD 14; 2 AD 14 – 37; 3 AD 38 – 64;
4 AD 64 – 8; 5 AD 80 – 98; 6-8 AD 98 – 138.

c. AD 14 – c. AD 37
The nodus went out of fashion, but the waved band of hair round the face
and the low bun of plaits bound round with another plait continued (46.2).

c. AD 38 – c. AD 64
Agrippina I (mother of Caligula) modified this style; the waved band
round the face was now flat on the top, before breaking out in a bank of
short curls on either side, before being rolled in a loose band to join the
bun at the nape of the neck. This had now become elongated so that the
loops of plaits hung down the back of the neck, although it was still bound
round at the top. There was often a single ringlet on either side behind the
ears (46.3). Agrippina II (third wife of Claudius) was portrayed with a
very similar style, although she was usually shown on coins with two
ringlets.

c. AD 64 – c. AD 68
Poppaea Sabina (second wife of Nero) had almost the same style, but now
the bank of curls was much wider, extending back over half the head
(46.4).

c. AD 69 – c. AD 79
Vespasian’s wife and daughter had died before he became Emperor, so no
official statues or coin issues of them were produced. It is likely that the
hairstyles in this period were closer to the following rather than the
proceeding styles, as people would not have wanted to associate
themselves with the disgraced Imperial families.

c. AD 80 – c. AD 98
Coins of Julia Titi (daughter of Titus) and Domitia (wife of Domitian:
coins of AD 82-3) show them with two different hairstyles. Both styles
had a wedge-shaped raised bank of tight curls in a crescent from ear to ear.
One style then had the hair pulled back in numerous small, separate plaits
that again hung in long loops at the nape of the neck, loosely tied with
other small plaits (46.5). The second style had the plaits drawn up into a
large, plump bun or coil of plaits near the top of the back of the head.

c. AD 98 – c. AD 138
Although Plotina (wife of Trajan) became Empress in AD 98, no coins
depicting her were produced until AD 112, when she is shown with a
hairstyle developed from the previous style. Round her face was a narrow
band of false hair, downturned in front of her ears, marked out with C-
shaped waves, while above it her hair was brushed over a raised pad to
form a peak. The hair was then divided into numerous small plaits drawn
down to the nape of her neck, where they hung in long loops, as in the
previous century (46.6). At the same time, the deceased Matidia (Trajan’s
sister) and her daughter Marciana are shown on coins with a different
style: there was the same narrow band of hair and the peak (or often two or
three peaks of hair), but behind them the small plaits were twisted up into
an open circle. From the side, this open circle was often slightly twisted,
forming almost a Z-shape lying on its side (46.7). The second hairstyle
worn by Matidia and Sabina probably developed into one of the best
known of Roman hairstyles (but one that is not depicted on coins), where
the bank of curls grew extremely tall. From the front, the bank of curls
started out being wider than it was tall, but gradually developed until it
was taller than it was wide (46.8). Sabina (Hadrian’s wife) wore the same
style as Matidia and Marciana, and another, simpler style, closer to
Plotina’s style: her loose, plain hair was drawn up over a pad and pulled
back in a loose pony-tail at the back of her neck, with the ends folded
under and tied rather low down, towards the end of the tail. A ribbon ran
from in front of the pad to tie at the back of the head (47.9).

c. AD 138 – c. AD 146
Faustina I (wife of Antoninus Pius) died in AD 141, so most of her coins
are posthumous, but they again show that the Empress chose a new
hairstyle that was very distinct from that which had gone before. Her hair
was parted in the centre, and set in soft waves round the head to the back,
where the hair was divided into a number of plaits which were then loosely
twisted together up the back of her head and then coiled into a neat, flat
bun right on the top of the head (47.10-1; 30; colour plate 19). On some
statues the bun grows much taller and becomes almost conical in shape.

c. AD 146 – c. AD 160
Faustina II (Faustina I’s daughter and wife of Marcus Aurelius) had a
hairstyle different from her mother’s. The hair round her face was divided
up to form long overlapping curls pulled back away from her head,
sometimes to a braid worn like a headband. The rest of her hair was
divided into sections that were lightly rolled before being pulled to the
back of her head, where they were coiled into a flat open bun made up of
plaits. A distinctive element of the bun was the ends of a number of plaits
that came from under it, their ends tucked back into its centre (47.12,
colour plate 19).
c. AD 160 – c. AD 180
When she grew older, Faustina adopted a new hairstyle. It grew more
bulky, with more body, and was set in waves, before being pulled back to a
large, round bun, low at the back of her neck, sometimes depicted as
coiled plaits but sometimes shown on statues as a looser bun with an
interlaced design. Occasionally, the hair round the face was shown twisted
or plaited before it fed into the bun. Faustina’s daughter Lucilla is shown
with the same hairstyles (47.13; colour plate 19).

47 Hairstyles of the second and third centuries. 9 AD 98 – 138; 10-1 AD 138 – 46; 12 AD 146 –
60; 13 AD 160 – 180; 14 AD 180 – 200 15-6 AD 200 – 217.

c. AD 180 – c. AD 200
Crispina (wife of Commodus) is sometimes shown wearing the same style
as the elder Faustina II and Lucilla, but in some of her coin portraits the
style has developed further. The hair was still heavily waved, but now fully
covered the ears before being drawn back to the bun, which was now much
larger, reaching almost to the top of the head. The wife and daughter of
Didius Julianus (reigned for one year, AD 193) had the same large bun.
Marble statues show quite clearly that these buns were false pieces, with
elaborate interlacing star patterns (47.14). The early coins of Julia Domna
(wife of Septimius Severus) show her with the same huge bun, although
the hair round her face drooped down onto her neck before being brought
up to the bun. Julia Domna’s hairstyles were always very large: other
women copied the style, but usually without the same bulk.

c. AD 200 – c. AD 217
Julia Domna’s second hairstyle was equally large. The hair was still
heavily waved, and the hair round the face was often twisted into a cable
pattern before being drawn back to the bun (47.15). The bun could be
large, covering the whole of the back of the head in a single spiralled plait,
or it could be low and oval, at the nape of the neck. Plautilla (wife of
Caracalla, AD 202-5) is shown with this second style: her hair was not as
bulky, but it still covered her ears, was heavily waved and had a low spiral
plaited bun (47.16). She is also shown with a very different style, with her
hair divided into sections and twisted in rows before being pulled back to a
flat bun at the back of her head.

c. AD 218 – c. AD 240
Julia Maesa (grandmother of Elagabalus) is shown with her straight hair
covering her ears and pulled back to a spiralled bun covering the back of
her head, but her hair did not hang down so low on her shoulders, nor did
her bun cover as much of her head as Julia Domna’s bulky styles. Coins
also show her with waved hair and a small, oval bun made of a spiralled
plait at the nape of her neck. This low, very small oval bun style, now with
the hair tucked behind the ears, was worn by all of Elagabalus’ wives,
Severus Alexander’s wife and his mother, Julia Mamaea (although
Elagabalus’ first wife Julia Paula is also shown with a style very similar to
that of Plautilla, see above) (48.17, colour plate 19). Marble portraits,
however, more frequently show the style with the slightly higher, larger
bun as worn by Julia Maesa and the wife of Balbinus on his sarcophagus
(AD 238; Kleiner 1992, fig. 356), although with hair tucked behind the
ears. The hair could be either straight or waved, and the bun made up of
one or two spiralled plaits; but either way, the top of the plaits formed two
distinctive ‘wings’ under the ears when seen from the front (cf 47.16,
48.17).

c. AD 240 – c. AD 268
Sabinia Tranquillina (wife of Gordian III: coins of AD 241-4) is shown
with two hairstyles; the low bun, and a new style where the heavily waved
hair was brought to the back of the neck, plaited or braided and then
pinned up the back of the head, sometimes so high that the top was just
visible on the top (48.18-9). Marble statues show that this was a wide, flat
bun, the details of which are often only sketchily carved. Some statues
seem to depict a wide, multi-strand braid of hair, or a number of smaller
braids sewn together (see Stephens 2008, fig. 19), while others seem to
show a hairnet (Nowicka 1998, 122; for late third-century date see
Michaeli 1998, 64). The same style was worn by a number of the wives of
the succeeding short-lived Emperors. A different hairstyle, not shown on
coins, which might belong to the same period had the hair forming wings
behind the ears, then being lightly twisted up the back of the head and onto
the top, with the ends tucked back in. It is difficult to see how such a loose
twist could ever be kept up (48.20).

c. AD 270 – c. AD 307
Ulpia Severina (wife of Aurelian) developed the style worn by preceding
Imperial women by bringing the bun right down to the forehead, where the
ends were tucked in (48.21; Hartley et al 2006, no. 97). The section of bun
over the forehead could be decorated with spherical-headed hairpins, or
pearls sewn onto the net. A wall painting from a catacomb shows two pins
or pearls (35.2), while a portrait from Egypt clearly shows three gold-
headed pins (Doxiadis 1995, figs 11-2).

c. AD 307 – c. AD 326
Helena (mother of Constantine) is shown wearing either a hairstyle similar
to that of Ulpia Severina, or another where an extremely thick plait was
wrapped round her head in the manner of a diadem (48.22). On some
statues, the plait is decorated with jewels. Fausta (wife of Constantine I),
however, is shown with two hairstyles that are closer to styles of two or
three hundred years earlier: waved hair, brought back to a flat-topped bun
wrapped round with a plait low on her neck, as worn c. AD 14 – c. AD 37,
or sections of hair twisted and brought back to a coiled plaited bun high on
the head, similar to some styles worn by Faustina II and others c. AD 145
– 160. These simple styles matched her husband’s change from the
Tetrarchic look to a neo-first century appearance, harping back to an
earlier age. It was, however, a short-lived fashion.

c. AD 326 onwards
After this date, it is almost impossible to distinguish the hairstyles of
Imperial women on coins, as they are invariably depicted with elaborate
diadems and head-dresses (colour plate 2). As the Imperial family used
their headgear increasingly to distinguish themselves from ordinary
people, fashionable women could no longer copy the Empresses’
hairstyles in the same way as in previous centuries. The fact that
respectable Christian women now had to keep their head covered inside as
well as outside the house also meant that there was less opportunity for
display of ornate coiffures. St Paul wrote that, as a woman was made from
man, and for man, she should wear a veil on her head as a symbol of man’s
authority over her (1 Corinthians 11, 10). The hair was now fully covered
so that often not even a curl was visible, and the fashion for elaborate
display was past. In the middle of the fourth century, the wife shown in a
tomb at Silistra wears a white cap with a bulky light blue wrap over the
top (Nowicka 1998, 121). In the late fourth century, Serena, the wife of
Stilicho, has a bulky hairstyle looking rather like a cottage loaf covered in
a striped cloth (48.23).
48 Hairstyles of the third to sixth centuries. 17 AD 218 – 240; 18-20 AD 240 – 268; 21 AD 307
– 326; 23 late fourth century 24-5 fifth century.
Fifth to sixth century
In the sixth century, the ladies of the court of Theodora (Empress AD 527-
48) also still wore their hair concealed beneath hoods, so it can be
assumed a similar fashion was followed through-out the fifth century.
There are two main styles in this period, the first of which had a cap that
framed the face closely, with a narrow horizontal band that could be either
plain or decorated (48.24). On some women, this band gave the cap a
cottage-loaf profile, similar to Serena’s hairstyle. The second style
consisted of a tight-fitting cap covered by a larger hood in a different
colour (48.25; colour plate 2).

HAIR PINS

The more elaborate hairstyles were probably sewn together without the
need for hairpins, creating very stable hairstyles that could be slept in if
necessary (Stephens 2008, 124). Sewn hairstyles required hairdressers to
help produce them, and, just as importantly, to undo them, and would have
been reserved for the rich (ibid.). Some hairpins seem to have been more
ornamental than functional, like those shown in a number of Egyptian
mummy portraits where long pins stick out from hair piled up on top of
the head, sometimes placed diagonally and occasionally horizontally
(colour plate 14; Doxiadis 1995, figs 40, 72). However, those women with
less ornate hairstyles, lower down the social scale, seem to have used
hairpins to keep their hair in place, and bone pins can be found in some
numbers on archaeological sites. Numerous inhumation burials have pins
positioned close to the skull (Philpott 1991, 150), and there are a couple of
examples of pins still found in situ in hair. A mummified head of a fifteen-
year-old girl dating to the early second century has a plait drawn up to the
top of the head, and held in place by four pins, two of bone, one of
tortoiseshell and one of silver (Walker and Bierbrier 1997, no.302).
Another head, from York, shows a bun of hair held in place by two
remaining jet pins (Hartley et al 2006, no. 108). Hair that has survived in a
number of other burials includes the second-century burials at Les
Martres-de-Veyre, France where there are the remains of two simple plaits
and a fourth-century grave at Poundbury, which has the remains of five-
and six-strand braids (Audollent 1922, 39; Farwell and Molleson 1993,
206; pl. 49).

HAIR OIL

Women, like men, used hair oil. In the first century, Martial wrote a gift
tag for a gold hair pin which suggests the use of hair oil: ‘so that your
moistened hair may not damage your bright silks, insert a pin to hold up
your twisted hair’ (Epigrams, 14.24), while in the second century, the
narrator of The Golden Ass, in describing his obsession with women’s hair,
talks of hair ‘anointed with the lotions of Arabia’, and mentions
worshippers of Isis with oil-moistened hair (Apuleius, 2.9; 11.10). The
plait of hair from Poundbury, and hair from another woman in the same
cemetery, seem to have been oiled (Farwell and Molleson 1993, 206). The
waves in the hair frequently seen on the portraits may have been made by
oil and finger-styling, or heated curling irons. Statues can show the same
hairstyle with both waved and smooth hair, so it would seem this was a
personal choice.

HAIRNETS AND COVERINGS

Hairnets made out of gold were sometimes worn by the rich during the
first century AD, although as hairstyles grew more complicated, hairnets
went out of fashion (49.3). According to a biographer of Severus
Alexander, hairnets were also known in the early third century, when the
Emperor includes one in a list: ‘the women of the imperial household
should be content with one hairnet, a pair of ear-rings, a necklace of
pearls, a diadem to wear while sacrificing, a single mantle ornamented
with gold and one robe with a decorated border, not to contain more than
six ounces of gold’ (SHA, Severus Alexander, 41.1). However, as the
Scriptores Historia Augustae was written in the late third or fourth
century, when hairnets worn over the wide plait may have come back into
use, this could be an anachronism.
In the late Roman period, the coverings became less net-like and
more cloth-like, concealing more of the hair. The late fourth-century grave
portrait of Aelia Arisuth shows her dressed in a blue tunic, with a short
white mantle wrapped round her shoulders and her hair covered by a white
cap. In the fifth century, the whole of the head was covered by a snood,
sometimes gathered at the front (49.4). Some seem to have been made out
of cloth, but surviving examples from Egypt are made out of sprang. There
were two basic shapes among the Egyptian examples: rectangular, bag-
shaped ones and tapering triangular ones, and they could be brightly
coloured and patterned (Jenkins and Williams 1985, ill. 2-3, figs 9-11; von
Falck 1996, nos 333a-c).

HATS

Modest women were supposed to keep their heads covered when out of
doors, but this generally took the form of draping their mantle over their
head rather than wearing a hat, since the elaborate hairstyles frequently
made hats impossible (41). In the first centuries BC and AD, there are
some references to the turban (mitra) being worn by rich women, and it is
possible this is the cloth wrapped round the head like a headscarf seen on
women in Hellenistic paintings (49.1-2). Catacomb paintings of Christian
women praying sometimes show them with a veil covering their head and
hanging down their back or over their shoulders, but rarely crossed over
like a shawl round the neck (35). One is clearly opaque cloth, depicted as
white with two wide, dark bands and a fringe (35.1), but others appear to
be thinner and semi-transparent (Nicolai et al 2009, fig. 148).
The freedman’s cap could also be worn by women on occasion;
women who had probably been freed in the will of the deceased are shown
wearing it on the reliefs of the Tomb of the Haterii in Rome (50). Straw
hats, like those worn by men, would have been used by labourers or
travellers, but they were never fashionable for the rich.

49 Hats and headcoverings. 1 Mythological figure from the first century BC wall painting of the
Punishment of Cupid, Pompeii; 2 Mythological figure from first-century BC Villa of Mysteries,
Pompeii; 3 Portrait of a woman, wall painting from Pompeii; 4 Sixth-century portrait bust of
woman, Metropolitan Museum of Art, New York.
50 Freedwoman with loose hair and wearing a cap at the funeral of her ex-mistress, Tomb of the
Haterii, Vatican Museum.

PARASOLS

Parasols (umbellae and umbracula) were used to protect rich women from
the harmful rays of the sun (as tans were definitely unfashionable).
Martial wrote a gift tag for one: ‘accept a parasol to defeat the fierce sun’
(Epigrams, 14.28), while in The Golden Ass a Triton ‘protects Venus from
sunburn with a silk parasol’ (4.31). Although women could carry one
themselves, they could equally get a slave or, better yet, an admirer to
carry it for them, as Ovid suggests: ‘do you yourself hold her parasol
outstretched upon its rods’ (Art of Love, 2.209). A relief from a tombstone
from Rome shows that the parasols could be folded in the manner of the
modern umbrella (51.1). A wall painting of a female figure, perhaps the
personification of summer, has one decorated with tassels (51.2), while
Juvenal refers to a parasol of green (Satires, 9.50).

FANS
Women used a fan (flabella) to keep themselves cool. Some were folding,
with handles made of ivory, such as those found at York (51.3), perhaps
with the skin made from strips of vellum sewn together, as on early
medieval examples. A tombstone from Carlisle and a mosaic from North
Africa both show this form of fan (51.4; see also 71). Others were
rectangles of woven palm set on a long handle (51.5), while there is
literary evidence for fans made from peacock feathers (Claudian,
Eutropius, 1.109; Propertius, Elegies, 24.11). Martial refers to leek-green
and purple fans (Epigrams, 3.82.11; 10.30.15). Men did not usually carry
fans, but they could still be fanned by slaves.
51 Parasols and fans. 1 Folded parasol, bas relief from Avezzano, Italy; 2 Allegorical figure
holding parasol, wall painting from Pompeii, Musée du Louvre; 3 Ivory fan handle (one of a pair)
from a grave in York; 4 Allegorical figure of ‘summer’, mosaic from Djebel Oust, Tunisia; 5
Woven palm fan from Egypt, Staatliche Museen zu Berlin; 6 The mistress from the Lord Julius
mosaic, Carthage.

FOOTWEAR
Study of surviving examples has shown that only rarely was there any
difference in the style of footwear as worn by men and women (van Driel-
Murray 2001, 194). Men could wear flimsy sandals, and women boots very
similar in style to those being worn by soldiers, only distinguishable by
their size. During those periods when it was fashionable to wear tunics that
reached to the floor it is not always possible to see what women wore on
their feet, and there is the usual problem of identifying the terms used in
literature with the surviving artefacts. As with male versions, the footwear
could be made from leather, palm, rush and similar plants, cloth and wood.
Pliny also mentions cork being used for the soles of women’s winter shoes
(Natural Histories, 16.13.34).

SANDALS

Sandals were a long-lived fashion, being worn from the first century
through to the fourth century, and probably beyond. In the first and second
centuries, women are shown wearing sandals with delicate thongs from
between the big toe and its neighbour, to half-way down the sole
(Goldman 1994, figs 6.3a, 6.5b). On portrait statues the thongs are
sometimes only painted onto the stone, but when depicted the thongs were
often cut into arrow-shapes (52). This central section could also be
decorated with pearls (colour plate 15; Pliny, Natural Histories, 9.6.114).
Very often there was another thong round the back of the foot (53; see also
the seated woman – probably a mythological figure – in colour plate 13).
In mainstream art, women are shown wearing these sandals with bare
feet, but slotted socks could also be worn with them (20; colour plate 15).
Although these sandals had delicate uppers, they usually had thick soles,
often several centimetres thick on sculptures. Thick soles show that such
sandals were intended to be worn outside, and in the north-west provinces
existing examples usually have hobnails (van Driel-Murray 2001, 192). In
this part of the Empire, if not elsewhere, the shape of the sole changed
from a naturalistic shape with indents to mark the toes to a narrow and
sinuous design, in contrast to the triangular shape of male sandals (ibid.,
fig. 3). A slightly different form of sandal has a band over the foot without
a strap between the toes. An example from Saalburg in Germany has a
decorative stone set on the strap, while one from London has a highly
complex tooled and gilded decoration (Goldman 1994, 111; fig. 6.14).
52 Sandal, early second-century statute, British Museum.

BOOTS AND SHOES

Women also wore shoes and ankle-high boots. During the fourth century,
when the short tunics made footwear more visible, women are almost
invariably shown wearing enclosed boots. The female figures shown in the
Silistra tomb wear light brown and white ones, with a seam or fastening up
the front (38.2; Barbet 1998, 115, 117). Other examples have a row of dots
up the front (see the female figure to the left, colour plate 8), which may
possibly be decoration, such as pearls. Diocletian’s Edict refers to
women’s boots for sixty denarii, shoe-socks and double-soled ox-hide
shoes for fifty, and single-soled sandals for thirty denarii. The use of gold
decoration on the shoe-socks or the shoes added about twenty to thirty
denarii to the price (Grazer 1975, section 9.17).
53 Sandals, bas relief from Avezzano, Italy.

SHOE-SOCK

Whether soft shoes or true socks, socci were also worn by women. Both
Caligula and Nero were criticised by later biographers for wearing
womanish socci, but it is not clear whether the shoe-socks themselves or
their decoration were considered effeminate.

SHOES WITH LOW VAMPS

In the fifth century, it is likely women wore the same type of low shoes
that the men wore, but long tunics tend to hide the details in art. Surviving
examples from Egypt are slightly more enclosed, such as a fifth-sixth
century example from Egypt, made of red or purple leather with gilt
decoration, including a circular motif with embroidered stars (Hartley et
al 2006, no. 117; see also the equally highly decorated red and purple
shoes dating to the fifth to eighth centuries: von Falck 1996, no. 429).

CLOGS AND PATTENS

Women could also wear wooden-soled clogs. The narrator in the Satyricon
was attacked by a ‘half-blind old woman wrapped round in the dirtiest of
linen and set up on unequal wooden-soled sandals’ (Petronius, Satyricon,
95). The pair from a woman’s grave at Les Martres-de-Veyre, France
seems to have been lined with sheepskin, although none of the straps has
survived (54).
54 Wooden clogs from second-century tomb at Les Martres-de-Veyre, France.

COLOUR

As well as being decorated with pearls, shoes could be decorated with

punched designs, openwork, gold thread and gold-leaf. Ovid writes of
‘snow-white shoes’ (Art of Love, 3.271), while a bandit disguised as a
woman also wore ‘thin white shoes’ (Apuleius, The Golden Ass, 7.8) and
the wife of a rich freedman wore Greek-style white shoes decorated with
gold (Petronius, Satyricon, 67.4). Some of these may have been made out
of alum-tawed skins, which naturally produces white leather. Leather
could be dyed black with a ferrous sulphate called shoe-maker’s black,
dyed red with madder, yellow by the bark of the nettle tree and be
bleached by the sumach plant (Pliny, Natural Histories, 34.26.112 and
34.32; 14.17.47; 16.53.124; 13.12.55). Tertullian writes indignantly of a
brothel-overseer wearing mullet-red shoes (On the Mantle, 4.10), which
might suggest that this colour was as symbolic of rank for women as it
was for men. A third- or fourth-century shoe from London has an inner
layer of leather in a contrasting pale colour to the outer purple layer,
which is cut into a delicate openwork pattern and sewn with gold thread
(Allason-Jones 2005, fig. 43). Theodora wears white and gold shoes
decorated with blue to match her tunic, while the ladies of her court wear
orange/red shoes (colour plate 2). This orange-red seemed to be a popular
colour in the sixth century, for it is used for the footwear of the Virgins in
procession, and almost all the women in the biblical scenes in the mosaics
of the Basilica of St Apollinare Nouvo on Ravenna, as well as many of the
women in the Vienna Genesis and the Rossano Gospels. (Bustacchini 1988,
123; 115, no. 4; 116; Weitzmann 1977, pls 24-9). The few women not
wearing orange-red shoes have black boots or shoes instead (ibid., pl. 26;
Bustacchini 1988, 110, no. 4).

SPECIAL COSTUMES
WEDDING DRESS

Roman marriage did not require a set ceremony as nowadays, as it was

more a matter of mutual agreement, public acknowledgement, and the fact
that the couple lived together in the husband’s house. As documentary
evidence of a marriage was often needed for legal reasons when dealing
with property, marriage contracts might be drawn up, but they were not
compulsory, and their presence or absence had no bearing on the validity
of the marriage. A celebration in front of witnesses helped to provide the
public acknowledgement and proof of mutual agreement, so weddings
could be celebrated in style, including torch-lit processions, feasts and a
specially decorated bridal chamber by those with money.
Except for the wealthiest Romans, the idea of special, elaborate
clothing to be worn only on the wedding day is modern. In the past, it was
more usual for a dress bought or made for the occasion to be worn again
afterwards. When the autobiographer Laura Ingalls Wilders got married in
1885, she simply wore a dress that she had already started making, as it
would be brand new for the ceremony; it happened to be black, but its
newness was more important than the colour (Ingalls Wilder 1970, 222).
There was, however, a recognised costume for Roman women rich
enough to be able to afford it. The most important feature was the flame-
coloured mantle (flammeum). Pliny uses the word luteum to describe the
colour, the same word he uses to describe an egg yolk, so it may have been
a dark yellow to orange colour, but it was also described as being the
colour of saffron, of the sunset, of flame and of Jupiter’s lightning
(Cleland et al 2007, 14). The mantle was worn in the same way as the
palla, but instead of the edge resting on the top of the woman’s head, it
was pulled forward to hide half of her face (55.1). This flame-coloured
mantle was originally worn on an everyday basis by the Flaminica Dialis,
the wife of the priest of Jupiter, and as she was unable to divorce, the
mantle may have come to represent constancy and was therefore used in
wedding celebrations.
A wreath of flowers and herbs was worn under the mantle, and the
hair was dressed in a special manner using woollen ribbons (vittae). A
single reference suggests that the hair was divided into six separate parts
and there has been much discussion on what form this took, but the
description relies solely on the interpretation of a single word, which has
also been taken to mean ‘twisted and bound’, or ‘cut’. As the hairstyle is
usually concealed under the mantle and wreath, the most that can be said
about it is that it was bulky (55.1).
55 Bridal and religious costumes. 1 Bride, bas relief from sarcophagus, Mantua, Italy; 2 Vestal
Virgin, third-century statue, Museo Nazionale, Rome; 3 Priestess of Isis, relief from tombstone,
Athens; 4 Priestess of Isis, relief from tombstone, Athens.
The other, less important elements of the costume included a straight
tunic (tunica recta) that was traditionally specially woven for the
occasion. One author says it was worn the night before the wedding, along
with a flame-coloured hair net (Olson 2008, 21). For most women this
‘straight’ tunic was probably no different from a normal tunic, and it was
certainly fastened under the bust in the usual manner, with a woollen cord
tied in a Herculean (reef) knot that only the bride’s new husband could
untie. The poet Catullus also mentions flame-coloured socci (Poems, 61.8-
10). This outfit was worn by brides in the first century AD, and Claudian is
still referring to a flame-coloured mantle worn by Serena when she
married Stilicho in the late fourth century (Stilicho, 1.80-4).

NIGHTWEAR

Wearing special clothes for bed (as nowadays) does not appear to have
been universal. The poor are unlikely to have invested in special clothes
just for sleeping in, and may have worn their daywear or simply nothing at
all. It is also difficult to say how much personal choice and night
temperatures come into it. Some modest women apparently did wear
clothes, such as the wife Martial describes, who wore ‘breast-band, tunic
and obscuring mantle’ (Epigrams, 11.104), while Propertius talks of a
woman sleeping wrapped in a mantle, and another wearing a night-cap of
silk (Elegies, 3.21.8; 2.29.15). The word camisia might refer to a form of
undertunic used as nightwear by both men and women (Olson 2003, 210).

RELIGIOUS COSTUMES

Roman priestesses sometimes wore a distinctive form of costume to

distinguish them from other women. Tertullian refers to women choosing
their religion purely for the sake of the clothes:
some are initiated into [the mysteries] of Ceres for the sake of an
all-white dress or the honour of ribbons (vittae) and the privilege
of a [religious] cap, while those with a contrary disposition
towards dark clothes and a dull woollen covering for the head
run free on Bellona’s mount; others commend Saturn for the
ostentatious purple stripes and the red Galatian wrap (On the
Mantle, 4.10).
The clothes of the religious cults were often old-fashioned, or used
unusual methods of fastening, and were usually closer in style to that worn
by goddesses than their fellow humans. It is noticeable how much use was
made of brooches to fasten shawls or mantles at a time when brooches
were usually reserved for male costume in mainstream fashion (55.2,
55.4). This may either be the influence of provincial clothing being
brought in with the often foreign cults, or of the outfits goddesses were
usually shown wearing, with tunics fastened with brooches in the old-
fashioned manner.

Vestal Virgins
The Vestal Virgins are the most well-known of Roman priestesses, holding
a very special role in Roman public life. Images of Vestal Virgins show a
variety of tunics and hairstyles, and there is some debate about whether
they actually represent what the women themselves wore (La Follette
1994, 59). Generally, their dresses seem to be based on Hellenistic tunics,
with long overfolds. The most important element, as with the bride, was
the covering for the head (Cleland et al 2007, 206). Their hair was
concealed with infulae, red and white woollen ribbons coiled round the
head and tied at the back, with the ends hanging down over the shoulders
in long loops (Servius, Commentary on the Aeneid, 10.538). Sometimes
these ribbons are shown to be arranged almost as a turban, or more in the
form of a simple headband (La Follette 1994, figs 3.1-3.6). When
sacrificing, they wore a short white mantle with a purple border
(suffibulum) over their head, fastened by a brooch (55.2).

Priestesses of Isis
The costume that seems to have been worn by these priestesses was also
based on ancient clothing; in this case, on that of the goddess herself. The
costume consisted of a fringed mantle fastened with a knot, worn over a
ground-length tunic with long, tight-fitting sleeves. The mantle was a long
rectangle of cloth, fringed on the short sides; the mid-point on the long
side was held in a bunch between the breasts, with the two ends taken
under the arm, and over the opposing shoulders (Goldman 1996, fig. 15).
The two ends were then tied in a knot, with the cloth held between the
breasts, with the fringed ends hanging down below. Statues of the goddess
from the third century BC show her wearing a very similar outfit, although
in this case the mantle is the main item of dress, and it is shown as thin
and figure-hugging, unlike the rather more robust Roman interpretations
of it (ibid., 246; fig. 4).
 This is shown on some tombstones of priestesses, but it is not clear
whether the outfit was worn only for religious activities, or for everyday
use (55.3). There is even the possibility that the priestesses were depicted
as Isis, and never actually wore the costume in real life.

Other priestesses
A tombstone of a priestess of Isis shows the woman not wearing her
mantle tied in an Isis knot, but folded over and fastened on her right
shoulder (55.4). Another tombstone of a priestess of Jupiter from Mainz
also shows a woman wearing her mantle fastened in a similar way. A
statue, now in the Vatican Museum, of a young girl holding a sacrificial
dish in one hand, has her mantle folded to form a long narrow band over
her shoulder and across her body, which is held in place by her belt. All
these unusual ways of wearing the mantle would distinguish a priestess
from normal women.

MOURNING

Women in mourning, like men, wore dull-coloured clothes. The essence of

mourning was to demonstrate grief by a lack of interest in personal
appearance, so women are often also shown with their long hair loose (49),
and would have abandoned their jewellery and make-up (Olson 2008, 42).
It is not known how long such dark clothing was worn after the funeral,
but women seem to have been expected to wear it for longer than men
(Cleland et al 2007, 127). A dull-coloured mantle with a purple border
called a ricinium was associated with women in mourning (Sebesta 1994b,
50; Olson 2008, 42).

JEWELLERY
Women completed their costume with jewellery, the most common items
of which were necklaces, earrings, finger-rings, and bracelets. These could
be made from a number of materials, such as gold, silver, copper alloy,
iron, glass, bone, jet, amber and precious or semi-precious stones,
depending on the wealth of the woman. Gold was the preferred metal for
those who could afford it, probably because of its warm colour; although
some jewellery in silver is known (particularly finger-rings), it seems to
have been preferred for domestic plate. Roman jewellery on the whole
depended on colour rather than glitter for effect, so the most sought after
and expensive jewellery was made from pearls rather than diamonds. The
Romans knew of diamonds, but they had no way to cut them and generally
they were used as drills or engravers rather than as jewellery. Modern gem
stones have faceted faces, and are set in foil to let the light catch them and
make them glitter, but although the Romans used precious stones, they
usually left them in their natural state. Emeralds were particularly popular,
as they naturally form hexagonal crystals, and only need to be drilled to be
used as beads. The most popular shapes for stones, according to Pliny,
were those with an elongated or lentular shape, while those with sharp
angles were least regarded (Natural Histories, 37.75.196).
Jewellery was priced according to the weight of precious metal and
the number of stones used, while craftsmanship, however exquisite, rarely
added to the value. A wedding contract of AD 260 from Egypt, for
example, listing the dowry of the bride, describes the jewellery by weight:
a necklace in common gold … of the kind called maniaces,
having a stone and weighing apart from the stone thirteen
quarters, a brooch [?] with five stones set in gold, weighing apart
from the stones four quarters, a pair of ear-rings with ten pearls
weighing apart from the pearls three quarters, a small ring
weighing a half quarter … making the total of the whole dowry
one mina and four and a half quarters of common gold and for
the valuation of the clothing 620 drachmae’ (Oxyrhynchus
Papyri, 1273).
The fictional Trimalchio tried to impress guests by pointing out that his
wife had golden jewellery (including armlets, anklets, and a hairnet)
weighing 6.5 lbs (2.9 kg), while he had an armlet of 10 lbs (4.5 kg). To
prove his point, he then proceeded to weigh the jewellery in front of his
guests (Petronius, Satyricon, 67). Then, as now, the amount of jewellery
worn would have depended on the occasion, and on the taste of the woman
involved. Pliny complained about Lollia Paulina, a consort of Nero, being
overdressed for ‘an ordinary betrothal banquet’, being ‘covered with
emeralds and pearls interlaced alternately and shining all over her head,
hair, ears, neck and fingers, the sum total amounting to the value of
40,000,000 sesterces’ (Natural Histories, 9.58.117).
 The ancient authors often disapproved of the use of jewellery, or any
other unnecessary adornment of the body, as being frivolous, expensive for
the men who had to do the buying and un-Roman, ignoring the ideal of
simplicity (Olson 2008, 85, 88). However, a woman used the quality or
quantity of jewellery she wore to demonstrate her wealth and thus her
status, both to society at large and to her own social circle (ibid., 98). Livy
had the tribune Lucius Valerius, arguing for the repeal of a war-time law
forbidding women certain ornaments, point out that ‘No offices, no
priesthoods, no triumphs, no insignia, no gifts or spoils of war can come to
[women]. Elegance and ornamentation and care of self, these are the
insignia of women: in these they rejoice and take delight’ (Histories,
34.7.8-9).
Designs in precious metals were often copied in cheaper materials for
the lower classes. The designs also changed over time, with a general trend
towards more flamboyant and ornate designs (Stout 1994, 78, 80). Gold
coins set in pendants, bracelets and finger-rings, and intricate openwork
designs became ever more popular from the second century onwards.

PEARLS

According to Pliny, ‘the whole value [of pearls] lies in their brilliance,
size, roundness, smoothness and weight, qualities of such rarity that no
two pearls are found that are exactly alike’ (Natural Histories, 9.56.112).
During the first century AD, authors frequently used pearls as a symbol of
expensive jewellery, and wrote disapprovingly of their use. Pliny also
noted that ‘women spend more money on their ears in pearl earrings, than
on any other part of their person’ (ibid., 11.50.136). Worst of all, in his
opinion, were those people, male or female, who were so decadent that
they sewed pearls not only onto their shoes, but onto their socks (ibid.,
37.6.17, 9.56.114).
In the late second century, Fronto talks about the ‘celebrated string of
pearls, which every-one talks of ’ belonging to a woman called Matidia,
and noted the interest in what would become of them after her death. It is
clear from art that pearls continue to be the favourite jewel through-out
the Roman period. The third- or fourth-century gold and glass portrait
from Brescia shows the daughter with pearl necklace and ear-rings (colour
plate 16), while the sixth-century mosaic of the Empress Theodora shows
her wearing a head-dress set with pearls, with long strings of pearls
hanging down either side of her face and more pearls on her necklace and
brooch (colour plate 2).

FINGER-RINGS

Women also wore intaglios set in finger-rings, and, like men, would wear
rings on the upper joints (56). In the late first century, Pliny mentions
plain iron rings sent to the woman as a betrothal ring, but this was not
universal, and there was no real concept of our modern engagement ring,
nor were wedding rings obligatory (Natural Histories, 33.4.12).

56 Finger-rings and bracelets. A woman holding wreath and phial, painted linen shroud from
Egypt, Staatliche Kunstsammlungen, Dresden, Germany.

BRACELETS

Simple bracelets made from copper alloy wire are extremely common
finds in the archaeological record, and were made in a variety of styles.
More elaborate, larger forms of bracelets were made in gold, and could be
decorated with jewelled settings (56; Johns 1996, pl. 2). Bracelets made up
of glass or jet beads are also known.

EARRINGS
Earrings were worn in pierced ears. The simplest forms were loops,
sometimes hung with a bead, that were often meant to be worn semi-
permanently (Allason-Jones 1989, fig. 1). Those who could afford a choice
of earrings wore pendant earrings, which were made in a number of
designs. A loop threaded with beads or pearls, a gem set in gold with a
pendant pearl or bead below, or a horizontal bar with three pendant pearls
were all very common second- and third-century designs (colour plates
14, 16, 20). Pearls were naturally a popular stone, with drop-shaped ones
particularly favoured (Pliny, Natural Histories, 9.56.114). The future
Emperor Vitellius, when sent from Rome to Lower Germany, ‘took a
valuable pearl from his mother’s ear and pawned it, to defray the expenses
of his journey’ (Suetonius, Vitellius, 7).

NECKLACES

Few necklaces are shown in mainstream art other than a short string of
large beads, which were presumably frequently pearls. The Egyptian
portraits often show three or four necklaces worn at the same time, from
tight chokers to long loops over the breast, but it is unclear how common
this was elsewhere in the Empire (colour plate 14). Due to the relatively
high necklines of tunics, the short necklaces were often half-hidden (see
the mother in colour plate 16); and there was definitely no plunging
neckline to show off an elaborate necklace, as was the fashion in later
centuries.
The most common form of necklace was a string of small glass
beads, or a delicate necklace of beads threaded onto twists of gold wire
(Johns 1996, figs 5.11-13; Guido and Mills 1993, fig. 72; Weitzmann
1979, no. 286). The beads on the wire necklaces could be all glass or
precious stones in their natural state, in either case often alternating with
pearls. More elaborate forms had the beads joined by decorative elements,
such as quatrefoils or Hercules knots (colour plate 20). More expensive
necklaces had large precious or semi-precious stones (or even paste) in
gold settings, such as the necklace with amethyst and emeralds now in the
British Museum (colour plate 20).
Sometime in the late third or early fourth century, a fashion for large
collar necklaces worn over the top of the tunic started, and continued until
at least the sixth century (42, colour plates 2, 8). Some appear to be made
up of two or three rows of pearls and beads threaded together to create the
wide collar, while others have gems spaced further apart set on metal
plates, and others have the gems set in gold frames (cf Capizzi and
Francesco 1989, 33, 35, 73-5, 89). A surviving example from Egypt is
made up of separate metal plaques with a mixture of intricate openwork
designs, and was originally set with 112 gems. The central settings are
missing, but it still has pearls, emeralds and aquamarines set in it (57).

OTHER FORMS

Other less common forms of jewellery included hair ornaments that hung
down the central parting, armlets worn on the upper arm, anklets and body
chains (Johns 1996, figs 6.6-7; colour plate 18). Body-chains consisted of
long chains that crossed the body diagonally from shoulder to hip, with a
large jewelled setting where the chains crossed (ibid., pl. 6). The woman
preparing for a bath depicted on a silver bucket in Naples Museum wore
one under her tunic, and continued to wear it while she bathed (Borriello et
al 1986, 99, cat. 63). This woman is sometimes identified as Venus, and as
most women shown wearing a body chain seem to be mythological,
surviving examples may have been used on statues rather than by humans.
57 Necklace of gold with pearls and semi-precious stones. Early seventh century, Staatliche
Museen Preussischer Kulturbesitz, Berlin.
5
 CHILDREN’S CLOTHING

BABIES

Babies were routinely wrapped in swaddling clothes to keep them

immobile (Cleland et al 2007, 184). The swaddling clothes could take the
form of either a blanket or cloth tied round the baby with bands or strips of
cloth, or of a long strip of cloth wrapped round and round (58). The birth
of the future Emperor Clodius Albinus, as described in the Scriptores
Historiae Augustae, provides some insight into the swaddling clothes of a
newborn baby. His father, Ceionius Postumus, wrote a letter announcing
the birth: ‘A son was born to me on the seventh day before the Kalends of
December, and so white was his body at birth that it was whiter than the
linen cloth in which we wrapped him’, a circumstance which led to the
choice of name for the child, Albinus, meaning ‘white). Later, he was
wrapped in bandages of a reddish colour, as was the tradition of his family.
The physician Soranus gave detailed instructions on the bandages to
be used:
take soft woollen bandages which are clean and not too worn out,
some of them three fingers in breadth, others four fingers.
Woollen, because of the smoothness of the material and because
linen ones shrink from the sweat; soft, so as not to cause bruises
when covering the body which is still delicate; clean, so that
they may be light and not heavy, nor of evil smell, nor irritate
the surface by containing natron; and not worn out, for whereas
new ones are heavy, worn-out ones are too cold, and sometimes
rough as well and very easily torn. They must have neither hems
nor selvedges, otherwise they cut or compress’ (Gynecology
2.14[83]).
He also gave detailed instructions on how to wrap the baby up, and
recommended that the baby stay swaddled for between forty and sixty
days.
58 Babies in swaddling clothes. 1 Votive object, Musée Archéologique de Dijon, France; 2
Tombstone of Aeliola, Musée de Metz, France.
59 Children depicted on the Ara Pacis, Rome.

CHILDREN

There was not a great distinction between child and adult in the Roman
world, as offspring, even as adults, were under the authority of their father
until his death. In the poorer classes, children could be less than ten years
old when they joined the adult world of work and helped their families
earn a living. The minimum legal age for marriage was twelve for a girl
and fourteen for a boy, although in practise it was generally a few years
later. There was usually a ten-year age gap between husband and wife, so
in aristocratic families women married for the first time in their early
teens and men in their early twenties, while for the lower classes women
tended to marry in their late teens and men in their late twenties.
In general, children wore the same clothes as their parents (59; see
also father in 8 and his son in 39), wearing tunics, mantles and capes, in all
their variations, as well as the toga for special occasions. There is,
however, some evidence that young women sometimes wore a different
form of tunic to that of their elders. Although some girls are shown in
tunics belted under the bust, an Augustan tombstone with a girl beside her
parents shows the girl wearing a tunic belted at the waist, and a statue of a
young girl standing beside a seated woman (presumably her mother) now
at Chatsworth shows her in a tube-dress with a thigh length overfold,
belted at the waist and worn over a sleeveless tunic (Olson 2008, fig. 1.5;
Strong 1969, pl. 115). A sarcophagus now in the Louvre depicting girls
playing also shows the same type of dress (60). The advantage of this style
is that by reducing the length of the overfold, the tunic can be made longer
for the growing girl.

60 Young girls playing a ball game, bas relief from sarcophagus, Musée du Louvre.
On formal occasions, children were made to wear the same costume
as their parents, including the toga. A tombstone from Ostia shows a five-
year-old boy in the toga, while the procession of the Imperial family on
the Ara Pacis shows boys of about two and six also in togas, although the
chances of getting a child of two to wear a toga and keep it in order do
seem rather remote. On special occasions, boys wore the toga praetexta,
which was edged with a purple border, a form of toga also worn by
magistrates. Some literary references refer to girls also wearing the toga
praetexta, but it is not clear how common this actually was (Olson 2008,
15, 17; Cleland et al 2007, 32).
For girls, there was no rite of passage between childhood and
adulthood beyond marriage itself. Boys’ coming-of-age as an adult citizen
was marked by a ceremony when the toga praetexta was laid aside in
honour of the all-white toga virilis. There was no fixed age for this
ceremony, which was decided by the family, but it was generally around
the age of fifteen or sixteen. This ceremony would have been held only for
the children of rich families, and it is not clear how long it continued after
the first century. For men, there was a period of ‘adolescence’ in their late
teens and early twenties, between boyhood and marriage, which was often,
in tradition at least, a time for profligate behaviour. There was no
equivalent period for women, who moved straight from daughter to wife.

61 Hairstyle of young girl, statue of c.50 BC, Palazzo dei Conservatori, Rome.

Hairstyles
It seems that as soon as a girl had hair long enough to pin up, it was worn
up. Even relatively young girls, with hair only just below shoulder length,
are shown with three plaits that are fastened together on the top of the
head (61). The girl on the Ara Pacis is about ten or eleven years old and
already has her hair tied in a bun like the older women (59). Young girls
are also described as wearing fillets (vittae), although this is hard to find
in the artistic record. They are not shown with their head covered by their
mantles (Cleland et al 2007, 32).
6
 BEAUTY

WOMEN

At different periods over the centuries, different parts of the female body
have been seen as attractive, such as a neat ankle, a snow-white bosom or a
rounded shoulder, while the desired shape for the body itself has changed
over time, with various features of the body being either exaggerated or
concealed. For example, in the late fifteenth century, women’s dresses had
a high waist-line, and women are usually shown standing with an arched
back and folds of their skirts held in front of their stomach to give an
impression of pregnancy. In the nineteenth century, however, huge bell-
shaped skirts emphasised a very narrow waist, while in the 1920s
fashionable women were expected to be flat-chested in their straight, low-
waisted dresses. In the Roman period, the perfect body shape seems to
have been small-breasted and wide-hipped (Olson 2003, 204).
The rich Roman woman almost always had a high waistline under the
bust, skirts long enough to cover the ankles, elbow-length sleeves and a
relatively high neckline. The costume of the respectable woman could be
very enveloping, as Horace pointed out when he compared the dress of a
matron and a prostitute:
In a matron, one can only see her face, for … her long tunic
conceals all else. But if you seek forbidden charms that are
invested with a rampart … many obstacles will then be in your
way – attendants, the sedan, hairdressers, parasites, the stola
dropping to the ankles, the mantle wrapped round – a thousand
things which hinder you from a clear view. In the other – no
obstacle. In her Coan silk you may see her, almost as if naked, so
that she may not have a poor leg, an unsightly foot; you may
measure her whole form with your eye (Satires, 1.2.95-104).
The way for a respectable woman to show off her charms was to adopt the
clothing of the brazen. Women could wear tunics of extremely thin cloth
so that, although fully clothed, the shape of their bodies was visible, and
there are numerous complaints about this fashion in literature. The best
material to use was silk which, since it had to be imported from China,
was seen by some as yet another example of eastern fashions corrupting
the weak. Pliny complained about the huge distances involved and the
amount of work in preparing the thread ‘to enable the Roman matron to
flaunt transparent clothing in public’, and pointed out that whoever had
invented silken clothing ‘has the undeniable distinction of having devised
a plan to reduce women’s clothing to nakedness’, and Seneca spoke of the
‘clothes that hide nothing of the matron’s body’ (Pliny, Natural Histories,
6.20.54; 11.26.76; Seneca, Debates, 2.5.7).
Ovid’s books Painting the Face and The Art of Love are full of advice
on how women could make themselves more attractive to men. In essence,
he suggests that women should choose clothes and hairstyles to suit their
colouring and body shape, and to emphasise their best features, whatever
they were. In his companion book, The Remedies of Love, on how to fall
out of love, he rather cynically comments that men ‘are won by dress; all
is concealed by gems and gold; a woman is the least part of herself ’ (343-
4).
In the late Empire, women’s mantles generally became smaller and
less concealing, but in contrast, their tunics covered more of the body.
Necklines were hidden under collar necklaces, sleeves covered the arms
right down to the wrists and even the hair was concealed under snoods or
caps. More than ever, women had to use their clothes and their jewellery,
rather than the appeal of naked flesh, to make themselves attractive.
Women also tried to improve their beauty through the use of make-up
and hair dyes. Cosmetics were used to produce a white skin, to
demonstrate their pampered indoor life out of the sun, with some rouge to
add a touch of colour, and a black liner to emphasize the eyes (Croom
2005). Hair could be dyed black with (amongst other things) elderberries,
gall-nut, ashes of wormwood, ivy, myrtle fruit, sumach, oak-galls, sage
and an ore called ink-stone (Pliny, Natural Histories, 15.71.180; 24.5.10;
27.28.52; Dioscorides, On Medical Matters, 2.210; 1.155; 1.147; 1.146;
3.40; 5.119); it could be dyed red with an application of lees of vinegar
and mastic oil, or young walnut nuts (Pliny, Natural Histories, 23.32.67;
15.24.87), or yellow with the nettle-tree, the lees of wine with oil, yellow-
berry buckhorn, henna or saffron (Dioscorides, On Medical Matters,
1.171; 5.132; 1.132; 1.124; Tertullian, On the Apparel of Women, 2.6).
MEN

The desired shape for a man’s body has also changed over time, from the
unbelted calf-length tunics of the thirteenth century, to the skin-tight hose
and codpiece of the sixteenth century and the tucked-in waists and padded
shoulders of the early nineteenth century. The Roman tunic covered some
or all of the arms, and was generally belted at the waist or hips, leaving
most of the leg bare; even when trousers were adopted, they were skin
tight hose so that the shape of the legs was still visible. The exact length of
the tunic varied depending on who wore it, with soldiers wearing short
tunics and senators long ones, although there was obviously some room
for personal choice. Horace writes of ‘Maltinus who walks with his
garments trailing low; another, a man of fashion, wears them tucked up
indecently high as far as his waist’ (Satires, 1.2.25-30). Men who took too
much care over their appearance were regarded as either popinjays or
effeminate, but there was also the usual tension between conservative
older men and fashionable youths over what was acceptable. As there was
little scope for variety in the cut of the tunic or mantle, stylish young men
used bright colours to make themselves attractive, and Quintilian pointed
out that ‘purple and scarlet clothes go ill with grey hairs’ (11.1.31). Men
were supposed to be tall and athletic, and are usually shown with a darker
skin tone than women, consistent with a healthy outdoor life. Some men,
like the women, needed help with their looks, and wore shoes with raised
soles to make them taller, or resorted to wearing wigs, curling their hair
with tongs, plucking out all their body hair or wearing patches to cover
blemishes.
7
 PROVINCIAL CLOTHING

INTRODUCTION

The provinces of the Roman Empire stretched from Spain to Syria, from
Scotland to Egypt, and covered a great many different nations and tribes.
The Romans imposed their rule on these people, set up an administrative
system and demanded taxes, but they also accepted that different people
had different cultures, and did not attempt to convert them all to an Italian
Roman way of life.
Outside the Mediterranean region, the people in the provinces could
be divided into three broad groups. The smallest group were the
administrators and civilian and military leaders who were outsiders,
usually from the Mediterranean region, such as the Provincial Governor
and his staff, and the Commanding Officers of legions and auxiliary units.
The second group were people native to the area, but who adopted a
Roman way of life, such as the rich ruling classes, and those who lived in
towns and followed a money-based economy. Lastly, there was the third
group, mainly the rural dwellers who were very little influenced by the
Roman way of life, and continued to live their lives as they always had.
Mainstream Roman costume would have been seen in the provinces,
on the statues and busts of the Imperial family that would be found in
every major city or town. The outsiders, the Roman administrators, would
also have worn their own costume, and even the toga would have made an
appearance in the palaces of the governors on official occasions. It is also
likely that some of the members of the native ruling class adopted Italian
fashions. Tacitus, for instance, mentions Agricola introducing Roman
clothing to the Britons, amongst other examples of the trappings of the
Roman way of life: ‘the wearing of our dress became a distinction, and the
toga was frequently worn’ (Agricola, 21.2), and this may have trickled
down to other layers of society.
The second group of people also adopted a Roman way of life,
reflected in the buildings they lived in, the way they cooked and the
objects that they bought. Most noticeable, perhaps, is the fact that they
adapted the Roman method of burial, with a grave marker made out of
stone with a Latin or Greek inscription, and frequently also a portrait of
the deceased. These people could afford a tombstone, sometimes quite
elaborate, and had it made to a Roman pattern – and yet they are shown
wearing native costume. Living a Roman way of life did not mean to them
having to wear Mediterranean styles of clothing. The third group of people
(such as those in rural areas), those least touched by a Roman way of life,
would also have worn native costume, but as there are few images of them,
it is not clear if what they wore was exactly the same as those influenced
by the Romans.
It was recognised that different nationalities wore different style of
clothes. In the sixth century, Isidore wrote:
Each nationality has its own costume belonging to it, such as the
Parthians and their wide, flowing trousers, the Gauls and their
soft square mantles, the Germans and their shaggy coverings, the
Spaniards and their stringes, the Sardinians and their garments
of wild animal hides … Nationalities are distinguished by their
costume just as they differ in their languages. The Persians cover
their arms and legs with drawings and their heads with a turban.
The Alani are distinguished by their pointed hats. The Scotti
raise the hackles with their ugly dress as well as their barking
tongues. The Alamanni are clothed in their cloaks [the sagum],
the Indians in linen. The Persians wear jewels, the Chinese wear
silk, and the Armenians wear quivers (Etymologies, 19.23.1-3,
6).
Provincial costume would have changed over the centuries, just as
mainstream fashion did. The following sections, however, simply give an
introduction to the most typical costumes in the different areas of the
Empire without following their history in any detail. Looking at provincial
costume as a whole, there are two striking points to consider. First, it is
often the women who retain their native fashions while the men adopt
Greek or Roman styles, so that a gravestone can show a woman in native
costume and the man in a toga or in a Greek tunic and mantle (62).
Secondly, women’s dress also tends to be more distinctive, and it is
particularly noticeable that both the brooch and the hat play a much larger
role than in mainstream fashion.
 The best evidence for provincial clothing comes from tombstone
portraits showing the deceased in the manner they wished to be
commemorated. Another source of evidence is the images of prisoners-of-
wars used by Romans to decorate triumphal monuments. The danger of
these depictions, of course, is that they may be more stock images of
barbarians than fully accurate portrayals, but many of them show quite
detailed non-Roman clothing, and at the very least give a general idea of
what these people would have worn. They certainly show what people
thought barbarians wore, and must have been recognisable images for the
spectators, since the artwork came without the explanatory labels found in
modern museums.

62 Tombstone showing both native and mainstream fashions, second century, Landesmuseum
Joanneum, Graz.
Mosaics of the late Empire show people of all ranks wearing the late
form of tunic decorated with roundels. It is not clear if native fashions
died out to be replaced by a more universal fashion, or whether these
images owe more to mosaic pattern-books than to local custom. It is clear,
however, that some forms of native costume are not depicted beyond the
second century.

The following sections are not meant to be a comprehensive province by

province examination of costume, but instead are intended to give a
flavour of the varied styles of dress worn in different areas of the Empire.

BARBARIANS
Romans considered the wearing of skins and furs to be an attribute of
‘barbarians’, even if worn by their own auxiliary troops. Tacitus described
how a disguised Germanicus wore a ‘wild-beast’s skin over his shoulders’
when he moved incognito amongst his own troops, and the army under
Emperor Vitellius presented ‘a savage sight, dressed as they were in
shaggy skins of wild beasts’ (Annals, 2.13; Histories, 2.88). Claudian
dismissed Rufinus as being ‘thorough in his barbarity [because he] draped
tawny skins of beasts about his breast’ (Rufinus, 79). Fur had no
significance as an indicator of wealth as in the medieval and later periods
up to the twentieth century.

63 Greek costume. 1 Tombstone of couple, Athens; 2 Tombstone of Eutychia, Athens.

GREECE

Men
The male Greek costume was adopted in a number of countries where
there had been Greek rule or influence before the Romans, so that under
the Empire it was almost as widespread as Italian fashion. As has been
seen, it was even worn at Rome. The costume consisted of a calf-length
tunic, often unbelted, with a large mantle worn over the top (called a
himation, the equivalent of the pallium) and openwork shoes (63.1).
Soldiers and hunters wore belted knee-length tunics with cloaks.

Women
Female Greek dress had also influenced Roman costume, as can be seen by
the adoption of the gap-sleeved tunic, the belt high under the bust and the
large, all-enveloping mantle. However, gravestones from Greece show that
a different form of tunic was fashionable during the Roman period. This
was full length, with elbow length sleeves; sometimes it was belted under
the bust, but more commonly at the waist (63). Daughters or female slaves
are shown wearing a tube-dress with long overfold, belted at the waist,
possibly over a short-sleeved undertunic (63.2).

SYRIA

Like many provinces, the coastal cities of Syria were more Romanised
than the cities in the interior. The large trade city of Palmyra, some way
inland, has a large collection of tombstones and statues that show both
Roman and native styles of dress.

Men
Tombstones or painted burial chambers show the men dressed in two
different styles of costume: they either wore Greek clothes consisting of
tunic and mantle (B. Goldman 1994, fig. 10.16), or tunic, mantle or cloak,
and trousers based on Iranian and Parthian designs (64.1). The Iranian and
Parthian costumes are based much more on tailored clothing, fitted and
shaped to the body, in contrast to the looser Graeco-Roman fashions (B.
Goldman 1994, 164). The Parthian tunic was mid-thigh in length,
sometimes with side splits, with long, tight-fitting sleeves. It was
decorated round the neck, wrists and hem, and frequently had an extra
central stripe down the front (sometimes replaced by two stripes in the
third century). The tunic is always shown with plentiful, stylised folds, and
although the clothes may well have been made of thin cloth, the excessive
use of folds must be an artistic convention. The tunic had a thin belt, often
decorated, fastened with a reef-knot, and with the two long ends tucked
under the belt, leaving very characteristic loops on either side (64.2). The
man in Figure 64.1 has a slightly more unusual belt, with a narrow cord for
the actual tie. Serving boys, probably slaves, are shown wearing a baggy
tunic with elbow length sleeves. Underneath, they wear an undertunic with
a fringed edge, visible on one side only under the hem of the overtunic.
Over the tunic, the men either wore a large mantle in the Greek
manner that was mid-calf in length and hid most of the tunic underneath,
or they wore a much smaller cloak fastened on the right shoulder with a
circular brooch. The brooch was pinned in such a way that the hole left for
the neck was very large, and it would slip down the left arm (64.1). The
cloak, unlike the mantle, often had a decorated border.
The trousers were baggy, Iranian in style, and are shown with as many
fine folds as the tunic. The trousers narrowed at the ankle, and were
always worn tucked into ankle boots (if, indeed, they did not have integral
feet like the hose), and usually had a central band of decoration down the
front. Until about AD 150, much of the trousers was hidden by loose over-
trousers, similar to the idea of a cowboy’s chaps; they were attached only
at the sides of the leg, so that their top edge hung down at the front to
show the trousers underneath (64.2; B. Goldman 1994, 165). The top edge
had another band of decoration.
A rarer item of dress, most common in the third century, and shown
worn by rich, important men, was the ‘coat’. This was another item of
Parthian dress, in some ways like a modern cardigan, open down the front,
and worn over another item of dress (in this case, the tunic), collarless and
with tight-fitting sleeves (64.3). It could also be worn crossed over at the
front and tied with a belt, in the manner of a modern dressing-gown.
Footwear consisted of highly decorated, soft ankle boots, either with
a thin sole for indoor use, or a plainer version with a sole for either indoor
or outdoor use. Hats are not shown on the tombstones, apart from the tall,
cylindrical hat worn by the priests, sometimes decorated with a wreath
half way up the hat. That late ‘Roman’-style decorated tunics were also
worn can be seen by the fragment of tunic found in the city with the purple
and white interlace roundel so frequently seen in late Roman art (7.1).
Paint surviving on the sculpture or tomb wall paintings gives some
idea of the colour of the clothes; green trousers with a red stripe, pale red
tunic with white central stripes worn under cloaks of white, pale green and
light and dark brown (Colledge 1976, 101-2). A man called Maqqai is
shown wearing a red coat over a blue tunic with a red central band and red
trousers with a yellow stripe, attended by boys in red tunics and blue
trousers, and is shown in a blue coat in another image (ibid., 87, 100). A
man called Hairan, wearing a Greek style tunic and mantle, is shown in
white, presumably representing white linen, while his wife wears more
brightly coloured clothes (ibid., 84, pl. 114). As all the fragments of linen
recovered from the city are undyed, the colourful clothes shown in art
must have been made from wool or silk. Fragments of silk that have been
recovered are multi-coloured and highly patterned (ibid., fig. 55).

Women
Women wore an undertunic which could either have relatively narrow
sleeves (sometimes with a decorated band half way between shoulder and
elbow) or wide sleeves (occasionally fringed). Both versions had a band of
decoration round the wrist and two stripes down the front. Over the top,
they wore a tube-dress fastened on one shoulder only, the left, with a
massive brooch. To begin with, the brooches were trapezoidal in shape, but
later huge circular ones – the size of a fist – came into fashion. The brooch
could have one or two keys hanging from it, or three pendant chains
(64.4). Such large brooches would be very heavy, and would pull the
overtunic down, suggesting that they must have also been pinned to the
undertunic to keep them in position. The overtunic was worn belted at the
hips, but as the tunic was bloused out over it, the form of the belt is
uncertain. The women usually, although not always, kept their heads
covered. They wore a decorated band with a turban over the top, both
covered by a veil that reached to the hips or below.
64 Syrian costume. 1-2 Third-century tombstones of men from Palmyra; 3 Syrian god, bas relief
from Rome; 4 Tombstone of unknown woman, Palmyra.
Exquisitely carved busts of Palmyrene ladies covered in jewellery are
justly famous, but such ostentation was not universal. Most women are
shown only with ear-rings and bracelets, although they could have head-
and central-parting ornaments, ear-rings, up to seven necklaces, a large
brooch with pendant decoration, and a number of bracelets and finger-
rings. Sometimes wide anklets were also worn (64.4). The wife of Hairan,
in a tomb dated AD 149/150, has an undertunic of green with a red stripe,
and an overtunic shown as white with folds picked out in brown, and black
shoes (ibid., 84, pl. 114). Another woman has a tunic of green with an off-
white band edged in red.

JUDAEA

There are no contemporary portraits of Jews in the province during the

Roman period, so an understanding of their costume has to come from a
study of a few surviving garments and literary evidence.

Men
Men’s costume was very similar to the basic Greek costume used
extensively in the eastern Empire, consisting of a tunic and mantle. The
usual set of a man’s everyday clothing can be seen in a list giving the order
in which clothes were removed at a bath: ‘first he removes his shoes, then
the hat, then the mantle, then he unfastens the girdle, then his tunic, and
after he unties the undergarment’ (Tosefta Derech Evetz, Perek Ha Niknas
1). Surviving examples of tunics from a collection of c. 135 in the Cave of
Letters show that they were made of two rectangles of cloth sewn together,
decorated with two stripes (Roussin 1994, fig. 11.1). The mantle was
rectangular, and was distinguished from those worn by other races by the
fact that it had a tassel on each of the four corners (Deuteronomy 22:12),
while examples from the Cave show that by the mid-second century,
mantles had the L-shaped corner decorations used throughout the Empire
(Roussin 1994, 184; fig. 11.3). Interestingly, Jews were not supposed to
use hobnails in their shoes. This was a long-lasting tradition that had
started during times of persecution, when the sound of hobnailed boots or
the imprint of hobnails in the ground could instantly be recognised as
those of Roman soldiers without danger of confusion (ibid., 188). Men
were supposed to wear only white, but although this remained common, it
is clear that it was not always strictly enforced.

Women
The best evidence for women’s costume comes from a wall-painting
showing the finding of Moses from the synagogue of Dura-Europos in
Syria, although this needs to be treated with the usual caution. While the
women with the Pharaoh’s daughter wear Greek costume, the women of
Moses’ family and their attendants apparently wear native Jewish
costume, similar in many ways to that worn in Syria. They have an ankle-
length undertunic with wide, elbow-length sleeves and a shorter tube-dress
over the top, fastened on the left breast. A hip-length veil hangs over their
heads, and their mantles (decorated with L-shapes) are wrapped round
their hips with the top part folded over, although it is not clear how this is
fastened (65).

65 Judaean costume, as illustrated by a wall painting of the discovery of the baby Moses, Dura-
Europos, Syria.
The veil was an important part of the costume, as women by custom
(but not by law) had to keep their head covered when outside. Italian
Roman women were also supposed to keep their head covered when
outdoors, but it seems that Jewish women were more strict on this point:
‘among the Jews, it is so usual for their women to have the head veiled
that this is the means by which they may be recognised’ (Tertullian, On the
Military Crown, 4). Women were allowed to wear coloured clothes,
excluding the colour red, which was regarded as the colour of non-Jewish
women (Roussin 1994, 186).

EGYPT

Information on clothing in Egypt comes from two main sources: mummy

portraits buried with the dead, and garments that have survived in Egypt’s
dry conditions. Mummy portraits executed on wooden panels and inserted
into the wrapping of the mummy are most often only head and shoulder
portraits, so the amount of information they can provide on clothing is
therefore limited (colour plate 14). In the later period, half length or full
length portraits were painted onto large pieces of linen that were then used
as shrouds for the mummies (colour plate 15). Mummy portraits, of
course, are only evidence for the costumes of people rich enough to afford
the mummification process.
A huge number of cloth fragments have been recovered from Egypt,
although only a small number of these are complete garments. An almost
complete child’s tunic, now in the Laing Art Gallery, Newcastle, has a tuck
that shortens the tunic by some eight centimetres (three inches). It is
visibly sewn up on the inside of the tunic (colour plate 21, left hand side
near the accession number label), and unpicked on the complete side (to
the right), with the once-hidden cloth a brighter colour than the outer face
of the tunic. A similar tuck is found on a surprising number of complete
tunics from Egypt (see also 70, a tunic from France), so it seems that
instead of turning up a hem to shorten a tunic, the Romans preferred to put
a tuck in at the waist.

Men
Men are almost always shown wearing white tunics decorated with stripes
of purple, red or dark brown, with a white mantle worn over one shoulder.
Some men, identified as soldiers, wear a dark-coloured cloak, either
fastened on the right shoulder with a brooch, or simply draped over the left
shoulder in the manner of the mantle, but recognisable as a cloak because
of the brooch left pinned to a fold (Walker and Bierbrier 1997, 87-8). Later
portraits show an undertunic visible at the neckline, decorated in a simpler
way than the female version, usually with just two sets of lines front and
back (ibid., figs 95, 113; Doxiadis 1995, figs 9-10). The later full-length
portraits show men in mid-calf length tunics, wrapped up in a mantle
almost as long, with simple sandals (66.1). Other shrouds show men
wearing tunics with purple and white roundels (ibid., fig. 7).
There are a number of portraits, both male and female, with a
horizontal band of decoration at the waist, or across the chest. Some
clearly pass over both arms, and must be related to burial rites rather than
to costume, but there are some where the band is clearly on the body of the
tunic, as shown on a praying boy from a painted pot (66.2). A similar
horizontal band can be seen on the tunics of some historical figures on
decorated silver plate, and on the Christian figures from Lullingstone villa
(Weitzmann 1979, nos 427; Wild 1987, 41, fig. 6, pl. XII). In both these
examples, the band comes above a belt holding the tunic in, so it certainly
seems to be a decorative element rather than a wide belt or sash, and it
may be a genuine fashion of the fourth century and later.

66 Egyptian costume. 1 Painted linen shroud, Pushkin Museum, Moscow; 2 Sixth- or seventh-
century painted pot, Staatlichen Museen zu Berlin.
Women
There are no known depictions of the gap-sleeved tunic, although their
existence in Egypt cannot be totally ruled out as the upper arms are not
always shown or are covered by the mantle. Where the full shoulder can be
seen, the tunics have sewn sleeves with stripes that run over the shoulder.
In the first and second centuries, tunics were generally decorated with two
stripes, often edged in gold in the paintings (Walker and Bierbrier 1997,
figs 33, 38, 40). There are no full-length portraits of this period, so the
length of the tunic is unknown, and nor is it known if they were worn
belted or unbelted. The mantle is shown draped round one or both
shoulders (colour plate 14).
From the late second century, the use of a white undertunic seems to
have become more common (colour plate 13). The body of a young girl
found in a grave at Mostagedda wore three tunics, the innermost of which
had braid of red wool woven in a lozenge pattern in white linen round
neckline and sleeve, and a fringed hem (Lochhead 1990, 44; fig. 8), while
the undertunic of the woman in colour plate 13 had red decoration of
infilled triangles at the neck. Such full-length shrouds show third-century
and later tunics to be mid-calf in length, and worn unbelted. In the larger
portraits, the mantle is shown carried over both arms, so that it is draped
down between them in a big loop (Walker and Bierbrier 1997, fig. 114).
The mantle worn by the woman in colour plate 13 is the same colour as
the tunic and therefore not very clear, but it seems to have been carried in
the same way. These mantles could be decorated with the typical L- and H-
shapes.
A couple of late third- or early fourth-century full-length portraits
show women in calf-length tunics decorated with wide stripes, sometimes
patterned. The shrouds are fragmentary, and have a large zone of
decoration over the lower body, but in the area of the waist they have an
element of decoration that matches that of the tunic. These may be more
examples of mantles matching the colour of the tunic (see colour plate
13; Walker and Bierbrier 1997, fig. 114), but if not they could be
horizontal bands of decoration, as seen on some men’s tunics (Doxiadis
1995, figs 90-1).
Two portraits of a similar date show women in white tunics with
purple stripes. They have wrist-length sleeves to the undertunics, which
are tied at the wrist and at one or two other places up the arm, forming
puffed sleeves (42). The neckline is hidden by the heavy collar-type
necklaces of the fourth century. The mantles are decorated with roundels.

NORTH AFRICA

The tombstones showing native influence are very basic in style, and
therefore only very limited information can be gleaned from them. Pre-
Roman reliefs from Carthage show men and women in long unbelted
tunics, the women also with a long rectangular mantle worn pulled over
the head. Roman period tombstones show folds in the clothes, depicted as
parallel lines. Men seem to wear a tunic and mantle (67.1), while the
women wear a ground- or ankle-length tunic with a long overfold,
sometimes belted at the waist (67.2). On other figures, the mantle, worn
draped over the left shoulder and under the right arm, obscures the details
of the tunic. At Ghirza, hunters and farmers are shown wearing very short
tunics, sometimes belted with sashes (King 1982, 123). All the late period
mosaics, such as that of Lord Julius, show people wearing the mainstream
fashions of the Mediterranean region.
67 North African costume. 1 Tombstone of man, Carthage; 2 Tombstone of woman, Carthage.

SPAIN

Spanish art outside mainstream art also relied on simple images of human
figures for portraits (68.1-2). It is likely that details were added on in
paint, so it is not clear if women wore their tunics unbelted or not. A
gravestone of an innkeeper shows a servant probably wearing a short tunic,
belted either under the bust, or at the waist like other slaves and servants
from elsewhere (68.3). However, there is an image on another tombstone
of a person weaving, who is shown with no folds above the area of the
waist, and plentiful vertical folds below (Tarradell 1969, pl. 111). The fact
that the person could be wearing a belted tunic with folds shown only
below the belt, or a form of skirt or sarong, highlights the limitations of
these forms of simple images. Strabo described the costume of the
Lusitania area: ‘all the men dress in black, for the most part in coarse
cloaks (sagum) in which they sleep … but the women always go clad in
long mantles and gay-coloured gowns’ (Geography, 3.3.7).

68 Spanish costume. 1 Man in short tunic, bas relief from tombstone, Museo Arqueológico,
Burgos; 2 Tombstone of AAia, Museo Arqueológico, Burgos; 3 Serving woman, bas relief from
tombstone, Museo Arqueológico de Mérida.

GAUL

J. P. Wild has carried out extensive research on the provincial clothing of

the north-west provinces, including Gaul, Britain and Germany. Both men
and women wore a form of tunic Wild calls a ‘Gallic coat’, using the word
in the sense of a medieval rather than a modern coat; in other words, it was
a tunic to be worn inside, and was not bad weather outerwear.

Men
The native dress of the southern Gauls originally included trousers; Pliny
explains that the province of Narbonne was originally called Bracata
because the inhabitants wore trousers (bracae), while Propertius describes
a Belgic chief as wearing ‘striped trousers’ (Elegies, 4.10.43).
In the north, more Romanised Gauls wore the Gallic coat. The male
version of it came to just below the knee, with wrist- or elbow-length
loose sleeves, and a wide-slit neck-hole (Young 2000, fig. 13). It could
have a relatively narrow body, with distinct sleeves (69.2), or it could be
more caftan-like, with sleeves tapering to the wrist (69.1). The lower edge
was frequently curved, and a few examples are fringed both at the hem and
sleeves (Wild 1985, 369-70; Wild 1968, fig. 3; Roche-Bernard 1993, 143).
It was always worn unbelted, even outside, and thus was very different to
the mainstream tunic. It was usually worn over an undertunic, generally
visible at the neck. A narrow scarf, usually tucked in at the neck, was also
worn (Wild 1985, 376).
The form of outer-wear used with the gown was almost invariably the
cape, either the long hooded version, or the short shoulder cape. The long
cape is often shown with the characteristic V- or W- shape at the front,
where the lower part is not fastened, and is also shown with one or both
sides rolled up onto the shoulder to leave the arms free for work (Roche-
Bernard 1993, 26, 29, 143). In the provinces, the long cape was not seen
purely as poor weather or travelling clothes, but as a respectable covering,
and men were quite happy to be shown wearing it on their tombstones
(72). This outfit was worn by men such as doctors and wine-merchants, as
well as tenant farmers and slaves (Wild 1985, 374). The belted tunic was
also worn, and is shown on men such as hunters (ibid., fig. 18; Selzer
1988, Abb. 208, 252). In the later period, mainstream fashions became
more common through-out the provinces, and mosaics show the later,
roundel-decorated tunics.

Women
Women wore the same form of gown, in the same variety of styles,
although their version was naturally longer, and the undertunic often more
visible (Wild 1985, 392). The undertunic was sometimes ground-length,
with a rounded neck and tight-fitting sleeves slightly longer than elbow-
length (69.4, colour plate 18). The gown worn over the tunic was
approximately mid-calf in length, with wider, elbow-length sleeves, and
could have a slit, rounded, or V-shaped neckline (69.3-4). As with the male
version, the coat was not worn belted. Women never wore the cape, but
instead they wore a mantle, as in Mediterranean fashions, although
generally smaller than the palla. It was often long and narrow, and was
worn like a modern long scarf, wrapped round the shoulders with a long
end hanging down front and back (69.3), while other examples are even
smaller and were simply wrapped round the shoulders (71). In some areas,
women also appear to wear a hat, a tight-fitting cap that covered all their
hair (Wild 1985, 392). Female slaves wore a short-sleeved tunic, usually
ankle-length as befitted their lowly rank, tied with a belt closer in
appearance to a sash than a cord, either at the waist or under the breast
(ibid., fig. 41).
A remarkable group of waterlogged graves dating to the second
century at Les Martres-de-Veyre in France contained preserved clothes, as
well as other belongings and grave offerings (Audollent 1922). The six
graves were discovered in 1851 and 1893, and were therefore not
excavated in a modern manner, so much information has been lost, but a
number of clothes and shoes have survived for study. In the grave of a
woman, estimated to be about thirty years old, there were yellow, black
and green beads and leather sandals with cork soles, while in the grave of
another, older woman, there was a surviving plait of hair and wooden
clogs (54). A pair of cloth ankle socks apparently also came from this
group of graves, but from which is unknown (45.2; but see Desrosiers and
Lorquin 1998, 55).
69 The Gallic coat. 1 Relief from tombstone, Trier, Germany; 2 School-boy, bas relief from
Neumagen, Germany; 3 Tombstone from Regensburg, Germany; 4 Tombstone from Til-Châtel,
France.
The most complete group came from the grave of a young woman,
and consists of a large tunic, a long length of cloth said by the excavator to
be a belt, hobnailed shoes, knee-length socks and a plait of hair (70; ibid.,
fig. 2 shows the tunic and sash displayed on a mannequin in the late
nineteenth century). The belt is extremely long, being 4.30 m (fourteen
feet) including fringes, and may be an example of the sash-belt apparently
worn by slaves in the north-west provinces, although why this woman was
wearing one is not clear, as it is apparent from the grave goods she was not
a slave. The sash-belt may have been worn more commonly than the
tombstone evidence would suggest, as part of an everyday working
costume not usually commemorated by the deceased, who preferred to be
remembered shown in their ‘Sunday best’ clothes. The tunic has a large
tuck round the waist, a feature seen on a number of tunics from Egypt (cf
colour plate 21).

70 Romano-Gallic costume. The clothing and hair from Grave D at Les Martres-de-Veyre,
France.

BRITAIN

Men
Trousers, tunic and a cloak were probably the native dress of the Britons,
as of the Celtic and Germanic tribes. A fragment of a monumental statue
from North Africa possibly shows a captive Briton, wearing a short cloak,
no tunic and trousers with different forms of checks on either leg (Piggott
1967), and Martial refers to the baggy ‘old trousers of a poor Briton’
(Epigrams, 11.21.9). At some stage, it appears that the Gallic coat was
introduced into Britain from the Continent, for although not many
tombstones of male civilians survive, they do include images of coat and
cape, such as on the Carlisle tombstone of a mother and son where the boy
wears an unbelted coat and undertunic, under a cape with the characteristic
W-shaped lower edge (71).

71 Romano-British costume. Tombstone of mother and son, Tullie House Museum, Carlisle.

Women
The pre-Roman costume for women seems to have been differed between
women in the south and north of the country. In the south, at least some
women seem to have worn a tube-dress with paired brooches, but the
brooches are not found in the north. However, a Romanised version of this
outfit seems to have worn through-out the whole country during the first
and second centuries, since pairs of Roman-style brooches connected by a
short chain are found in the archaeological record (e.g. Johns 1996, fig.
7.7). From the late second century, if not earlier, women wore the Gallic
coat (71, colour plate 18). A tombstone from Chester shows a female
slave wearing a short-sleeved tunic belted under the breast (Wright 1955,
pl. XXXI, no. 120).

RHINE VALLEY

In this area, women wore a very distinctive costume, while their husbands
wore the usual Gallic outfit (72). The women’s costume, however, had
disappeared from gravestones by the late second century.

Women
The women wore a long undertunic with long, tight-fitting sleeves with a
high neck, sometimes decorated with a frill. Usually, a torc with a pendant
was worn below the frill. Over the top of the undertunic was worn a tube-
dress, fastened by brooches at both shoulders, although one side is
sometimes shown as having slipped down the arm (72). Other examples
have it fastened only on one shoulder (Selzer 1988, Abb. 5). Another
central brooch pinned the overtunic to the undertunic, which was open
down the front and fastened with a number of brooches set vertically
(Wild 1985, 393-4).
The mantle was worn over the top of this, draped under the left
shoulder, and fastened by a brooch on the right shoulder. This is an elegant
but not very warm way of wearing a mantle, and it is not clear if another
mantle was worn over the top for warmth when needed, or if this mantle
was unfastened and worn in a more conventional way; there are certainly
depictions of the mantle being worn draped over one shoulder in the more
usual manner. Surviving paint on one tombstone shows a pale green
undertunic, a red overtunic and a dark grey cloak (Wild 1968, 220).
72 Costume of the Rhine Valley. Tombstone of Blussus and Menimane, from Mainz, Germany.
Some women wore a tight-fitting cap or hair net over their hair that
was worn with a thick plait or pad of hair wrapped round the head above
the brow. The tight-fitting cap shows both this ring of hair, and sometimes
also a small bun on the nap of the neck (Wild 1985, fig. 35, pl. 12, no. 43).

LOWER GERMANY

In the area of the Ubian tribe, women may have dressed in another
distinctive form of clothing. The best examples of it are worn by mother
goddesses, but sarcophagus portraits and other reliefs of mortal women
show them in the same form of costume, so it may be that the goddesses’
costume does reflect the clothing of the area (ibid., pls VII, XII, XIII).
Alternatively, it may be that the costume of the goddesses was worn by
their priestesses, as with those of Isis.
The costume consisted of a ground-length undertunic and slightly
shorter overtunic. The details of the upper part of the tunic are unknown,
as over the top of this was worn a large semi-circular mantle, draped
shawl-like over both shoulders and fastened on the breast with a brooch.
The most spectacular part of the outfit was the hat, a very large circular
creation that looks almost halo-like from the front (73). Very few hats
show any form of texture, so it is not very clear how they are made or
fastened, but one example has a button with loops hanging down from it
against one cheek, which is possibly a drawstring, while others have a
decorated rod against their cheek that may be the end of a long decorative
hair pin (ibid., pl. XII, 43).

73 Costume of Lower Germany. Statue of a Mother Goddess, Rheinisches Landesmuseum,

Bonn, Germany.

GERMANY BEYOND THE ROMAN FRONTIER

Men
Trousers were seen as the defining costume of both the Celt and the
German. Tacitus describes the Germans as wearing:
a thick cloak (sagum), fastened with a brooch … They spend
whole days on the hearth round the fire with no other covering.
The richest men are distinguished by the wearing of
underclothes; not loose, like those of the Parthians and
Sarmatians, but drawn tight, throwing each limb into relief ’
(Germania, 17).
A Roman relief of captured Germans shows them wearing a short cloak
and nothing else, while a man from the tribe of the Suebi (with his hair
worn in the characteristic knot) wears trousers and a short tunic over a
bare chest (Selzer 1988, Abb. 47). They were unlikely, however, to go
bare-chested through a German winter, and other images show them in
long-sleeved tunics.
The Germanic trousers had integral feet like hose, as a surviving pair
from a pool or bog deposit in Denmark shows (74.2). A tunic was found in
the same deposits, with side slits and tight sleeves. The sleeves were made
of woollen cloth woven to a different pattern to that of the body, and the
faint patterning created by the weaving would have been visible (74.1).

74 Woollen clothing from a bog or pool deposit, Thorsborg, Denmark. 1 Tunic; 2 Trousers.

Women
Tacitus describes German women as wearing: ‘the same dress as the men,
except that for the women a trailing linen garment, striped with purple,
was frequently in use: the upper part of this costume does not widen into
sleeves, and their arms and shoulders are therefore bare, as is the adjoining
portion of the breast’ (Germania, 17).

DACIA AND MOESIA

Trajan’s Column and the Adamklissi Trophy commemorating Trajan’s

victories in the Dacian Wars show both native male prisoners and their
womenfolk. The fashions are very Germanic in style. The Dacian men are
shown in a tunic with wrist-length sleeves and high side slits fastened by a
leather belt with a long hanging strap-end, worn over ankle-length trousers
(75.2). Dacian women wear a short-sleeved long tunic that, from the
direction of the folds, suggests a draw-string at the neck (75.3). A mantle
was worn tied round the waist or hips.

75 Dacian costume from the Tropeaum Traiani, Bulgaria, c.AD 109. 1 Suebian prisoner with
hair tied in knot; 2 Dacian prisoner; 3 Dacian woman.
76 Costume of Pannonia and Noricum. Tombstone of a girl, Klagenfurt, Austria.

NORICUM AND PANNONIA

Men
Most men shown on tombstones are wearing togas or mantles (62), so they
seem to have adopted mainstream fashions very early. One distinctive item
of dress from Pannonia was the pill-box hat that was adopted by the
Roman army in the late third century (Vegetius, On Military Matters,
1.20).

Women
There are a number of surviving tombstones showing the typical first-
century Norican costume for young women (shown bare-headed and
therefore possibly unmarried). It consisted of an ankle-length undertunic
with a rounded neckline and skin-tight, wrist-length sleeves with wide
cuffs (perhaps formed by decorative bands of tablet-weaving), worn under
a tube-dress fastened on either shoulder by a large brooch with a tall
upstanding tail (surviving examples can be incredibly large: Garbsch
1985, Abb. 4-5). The upper tunic was shorter than the undertunic, and on a
tombstone from Klagenfurt it is shown tucked under, either pinned or
gathered by a drawstring (76), although on other tombstones no such
gathering is shown. The upper tunic was fastened at the waist with a wide
leather belt decorated with metal fittings and fastened by a buckle. From
the belt hung two thin straps with decorated strap-ends and a central, wider
strap with a rectangular terminal (ibid., Abb. 12). Some women also wore
an extra brooch in the centre of the breast, which seemed to have no
functional purpose. The mantle, when used, was worn over both shoulders,
in the manner of a shawl. Most women also wore a hat with a curled or
folded brim that sat on hair combed away from the face in two large rolls
(62, 77). Other forms of hat included a tall cylindrical hat with a veil or
mantle worn over the top, and a large boat-shaped hat, possibly made out
of fur (77.2-3).

77 Hats worn in Pannonia and Noricum. 1 Tombstone from Neumarkt, Austria; 2 Tombstone
from Lendorf, Austria; 3 Tombstone of Umma, Landesmuseum, Vienna.
 CONCLUSIONS

Modern everyday clothing is generally chosen for convenience and

comfort and, apart from uniforms, is rarely symbolic in any way. One of
the most important functions of clothing any time before the second half
of the twentieth century was to show social status. Perhaps the last
example of instantly recognizable high status clothing in modern times is
the fur coat, for although designer clothes might be hugely expensive, they
are not always recognised by the ordinary person in the street. In the
Roman period, there had to be a visible division between those who used
clothes as a status symbol, since they could afford large quantities of
expensive cloth and because they were not required to work, and those
who had to have functional and practical clothing because they had to earn
their living, either as slave or free man. Therefore, both the amount of
cloth used in the clothes and their colours would immediately say
something about the wearer to the casual observer. Further elements of
costume, their use controlled by law, such as the colour of the shoes or the
width of stripes, would provide additional information. A wearer was
therefore identified by a number of different elements: the quality of the
cloth, its age, condition and colour, as well as the way the clothes
themselves were worn, such as the length of the tunic and whether if it was
worn unbelted. Finally, the type of clothes might identify their wearers:
the freedman’s cap for ex-slaves, the toga for citizens, or the shoulder
capes particularly favoured by the lower classes.
Clothes also helped to identify race or nationality. Native costume
was worn, particularly by women, during the first and second century, even
by those who had a Romanised lifestyle. Rome itself must have seen many
visitors wearing their traditional clothing, and in some circumstances
mainstream fashion adopted or adapted items of clothing from the
provinces. The clothes themselves could be imported over long distances,
or made more locally in the style of the foreign clothing. Diocletian’s
Edict, therefore, mentions clothes from Britain, Gaul and Africa, as well
as Egypt and Asia Minor.
The Roman Empire is frequently considered to be a ‘western’
civilisation because it had provinces in what is now western Europe, but it
had equal interests in eastern Europe, and in fact more often looked to the
east for inspiration (sometimes to the disgust of the conservatives), and
eventually even its capital moved east to Constantinople. This interest in
the east can also be seen in costume. At first, Rome looked to Greek
culture, and such items as the mantle worn by men and the gap-sleeved
tunic worn by women were adopted from Greek costume, while later
Rome looked to the Persians and adopted the use of highly patterned and
decorated cloth. From beyond the Empire, they imported gems from India
and silk from China.
The most important trend in the development of Roman costume over
the centuries was the gradual tendency for the body, both male and female,
to be covered up. Men went from wearing short-sleeved, knee-length
tunics with bare legs to long-sleeved, calf-length tunics with skin-tight
hose. Women, already well covered in the first century when wrapped in a
mantle, ended up with only face and hands visible, even in the privacy of
their own home. In the east, the use of patterned materials and more
tailored clothes continued to evolve as the Byzantine Empire developed,
and thus became part of the clothing of the medieval world.
1 The Emperor Justinian and his court, mosaic of AD 546-8, San Vitale, Ravenna. © Scala,
Florence.
2 The Empress Theodora and her court, mosaic of AD 546-8, San Vitale, Ravenna. © Scala,
Florence.
3 Men at a bakery, wall painting from Pompeii. © Museo Archeologico Nazionale, Napoli.
4 Hunters, mosaic from Piazza Armerina, Sicily. © Sonia Halliday Photographs.
5 The method of putting on a third-century toga. For reference, the end of the toga that hangs
down the front of the body has been coloured blue, and the end that hangs down the back of the
body has been coloured red. Drawn by R. Oram.
6 A reconstruction of a cape. © Cohors V Gallorum.
7 This reconstruction of the cape has an extra piece of cloth at the neck that can be unfastened,
producing a higher neckline to give better protection in bad weather. © Cohors V Gallorum.
8 Family and slaves in procession to baths, mosaic from Piazza Armerina, Sicily. © Sonia Halliday
photographs.

9 A dead or wounded hunter, mosaic, Museo Borghese, Rome.

10 Coin portraits. From left: top line – Trajan AD 101-2; Hadrian AD 134-8; Septimius Severus
AD 205. Bottom line – Severus Alexander AD 231-5; Constantius I c. AD 300; Constantine I c.
AD 321. © Arbeia Roman Fort and Museum.
11 Reconstruction of a stola in wool, worn over a gap-sleeved tunic in linen. © Cohors V
Gallorum.
12 Reconstruction of a gap-sleeved tunic in silk. © Cohors V Gallorum.
13 Woman having her hair dressed, watched over by mythological figures. © Museo Archeologico
Nazionale, Napoli.
14 Portrait of a woman from Hawara, Egypt, c. AD 100-20. British Museum EA74712. © British
Museum.
15 Painted shroud of a woman, Egypt. The Metropolitan Museum of Art, Rogers Fund, 1909.
(09.181.8). The Metropolitan Museum of Art 2010 © Photo Scala, Florence.
16 Gold and glass portrait roundel of unknown family. Third- or fourth-century, Museo Civico
dell’età Cristiana, Brescia. © Scala, Florence.
17 Reconstructed leather briefs based on examples found at Queen’s Street and Shadwell, London
© Cohors V Gallorum.
18 Women exercising before a bath, wearing briefs and breastbands, mosaic from Piazza
Armerina, Sicily. © Sonia Halliday Photographs.
19 Coin portraits. From left: top line - Faustina I AD 141+; Faustina II AD 145-6. Bottom line -
Lucilla AD 161-80; Julia Mamaea AD 222-35. © Arbeia Roman Fort and Museum.
20 Jewellery of first- to third-century date, including three necklaces, a pair of ear-rings, and a
fragment of bracelet. They are made from gold, pearls, emeralds, citrines, amethysts, garnets, and
quartz. © Copyright The British Museum.
21 Fragment of child’s decorated tunic from Egypt, showing one sleeve and a tuck at waist height.
© Laing Art Gallery, Newcastle upon Tyne.
22 Tombstone portrait of Regina, shown wearing a Gallic coat, Arbeia Roman Fort and Museum.
© Arbeia Roman Fort and Museum.
23 Replica of the tombstone of Regina, showing how Roman tombstones would originally have
been brightly painted. Carved by K. McQueen and J. Jackson.
24 A selection of men’s provincial costumes.

25 A selection of women’s provincial costumes.

26 A selection of undyed wools.
27 Samples of hand-dyed wool showing some of the colours available using natural dyes.
28 A man’s tunic in linen, the dimensions of which are based on a tunic found at Khirbet Qazone,
Jordan. It has fold lines from being stored on a shelf, and finer crease lines round the waist where
it has been belted.
29 The tunic being worn, with elbow-length sleeves created from its great width.
30 A long-sleeved tunic in linen, of the type worn in the third-century.
31 A long-sleeved tunic in linen, of the type worn in the third and fourth centuries. The purple and
white decoration is based on patterns from Palmyra, Syria, and Egypt.
32 A long-sleeved tunic in wool, with the multi-coloured decoration popular from the fourth
century onwards. The pattern of the shoulder decoration is based on a sixth-century example from
Egypt.
33 A fourth-century cloak with roundel decoration. © Cohors Prima Gallica.
34 Leather shoulder cape.
35 Hose cut from woollen cloth. It has to remain baggy round the ankle, to provide enough room
to let the foot through, and round the knee, to allow the leg to bend without the cloth splitting.
36 A puttee based on the example from Søgaards Mose.
37 A puttee consisting of a long strip wound round the leg.
38 A soldier wearing a Pannonian hat. (Cohors Prima Gallica).
39 A gap-sleeved tunic in linen, of the type worn during the first and second centuries.
40 A short-sleeved tunic in wool, of the type introduced during the late second century.
41 A linen dalmatic of the type fashionable from the late third century, worn with a colour
necklace.
42 An overtunic of wool worn with an undertunic of linen, of the type fashionable from the fourth
century to the sixth century.
43 The all-enveloping palla, 1.5 m wide, and over 2 m long.
44 A rich woman holding a fan, accompanied by her slave holding a parasol © Cohors Prima
Gallica.
45 A Romano-British tube dress of wool worn over an undertunic of linen, fastened by brooches.
46 The open end of the tunic fastened by second-century trumpet brooches.
47 The tube-dress laid out to show its very simple design.
48 Women wearing the Gallic gown. © Cohors Quinta Gallorum.
 GLOSSARY

abolla form of mantle

balteus literally ‘belt’; also used to describe the diagonal folds of
cloth from shoulder to armpit on the toga
bardocullus shoulder cape
bracae a word that seems to have been used for both knee-length
breeches and ankle-length trousers
byrrus form of cape
calceus shoe; particularly the form of shoeboot worn with the toga
caligae openwork shoes worn by soldiers
cenatoria a ‘set’ of tunic and mantle worn during evening meals
chlamys Greek form of cloak adopted by the military
consul highest rank of Roman magistrate. Chosen annually, there
were only ever two at any particular time
cucullus shoulder cape
dalmaticus late form of wide-sleeved tunic
endromis thick mantle used by athletes
equestrian a social rank (also translated as ‘knight’) based on a
monetary qualification
fascea/fascia literally ‘bandages’; used of puttees worn by men and
women, and of the breast-band worn by women
flabellum fan
flammeum flame-coloured mantle worn by brides
institia literally a strap or bandage; a distinguishing element of the
stola
knight a social rank (also translated as ‘equestrian’) based on a
monetary consideration
lacerna/laena form of cloak worn by civilians
mafortium shawl or hood worn by women
mantle rectangle of cloth used as an outer covering
matron respectable married woman of some social status
mitra turban
paenula form of cape
paludamentumform of scarlet cloak worn by military generals
pallium/palla a mantle
pero ankle boot
ricinium head veil
senator a social rank based on a monetary consideration
sagum thick cloak that originated in the western provinces,
particularly worn by soldiers and country folk
sinus literally a fold; also used to describe the deep loop of folds
on the toga across the front of the body on the right-hand
side
snood cloth or net cap worn by women to keep hair in place
soccus cloth sock; possibly also refers to cloth shoe
soleas shoes worn inside the house
sprang-work a form of fabric made on a frame by knotting rather than
weaving (cf crochet and knitting)
stola female upper tunic that became symbolic of a Roman
matron
suffibulum short mantle worn over the head by Vestal Virgins
synthesis literally ‘set’; taken to mean a set of tunic and mantle worn
during evening meals
tablion inset of coloured cloth on cloaks worn by court officials in
the late Empire
toga an elaborate form of mantle that became the national dress
of the Roman citizen
tube-dress form of female tunic made from a tube of cloth, fastened
by brooches on the shoulders
tunicus/tunica the main garment of both man and woman, with or without
sleeves
umbellam parasol
umbo literally ‘boss’; used to describe the small loop of cloth
produced just above the waist on some of the early forms of
draping the toga
vittae ribbons worn round the head for religious reasons
umbracula parasol
 REFERENCES

In most cases the translations have been taken from the Loeb editions,
with some modifications, particularly in the matter of the terminology
used for clothing. The translation of Soranus is taken from that of O.
Temkin, and that of Vegetius from the work by N. P. Milner. The
translation of Tertullian is that published in the Ante-Nicene Christian
Library (volume 28) and that of Sulpicius Severus from The Nicene and
Post-Nicene Fathers (2nd series, volume 2).

d. = died; fl. = flourished

AUTHORS AND ABBREVIATIONS USED IN THE TEXT

name title date

Ammianus Roman History d. 400
Marcellinus
Apuleius The Golden Ass fl. 160
Aulus Attic Nights d. 175
Gellius
Cato On Farming d. 149 BC
Catullus Poems d. 54 BC
Cicero Letters to Friends d. 43 BC
Second Speech against Catiline
Response to the Soothsayers
Against Clodius and Curio
Claudian On the Fourth Consulship of Honorius fl. 400
Against Eutropius
Epithalamium
Second Book against Rufinus
On the Consulship of Probinus
On Stilicho’s Consulship
Columella On Farming fl. 50
Digest Extracts of Roman law
Festus fl. 150
Fronto Letters d. 168
Horace Satires d. 8 BC
Letters
Epodes
Iuvenal Satires d. 130
Josephus Jewish Wars /
Martial Epigrams d. 102
Ovid Art of Love d. 17
Ibis
Metamorphoses
Remedies of Love
Persius Satires d. 62
Petronius Satyricon fl. 60
Pliny Natural Histories d. 79
Pliny the Letters d. 113
Younger
Plautus Epidicus d. 184 BC
Comedy of Asses
Propertius Elegies d. 16 BC
Prudentius Daily Round fl. 400
Mansoul
Quintilian The Education of the Orator d. 95
Seneca Debates d. 65
Servius Commentary on the Aeneid fl. 390
SHA Scriptores Historiae Augustae; a collection late third – early
of biographies by various authors fourth century
Sidonius Letters d. 488
Statius Silvae d. 96
Strabo Geography /
Suetonius Lives of the Caesars d. 160
Sulpicius Dialogues d. 425
Severus
Tertullian On the Mantle d. 220
 On the Military Crown
Tibullus Elegies d. 18 BC
Varro On Farming d. 27 BC
Vegetius On Military Matters fl. 386
 WEAVING TERMINOLOGY

A loom, of whatever design, is basically a wooden frame to support and to

keep taut the threads used during weaving. The threads that hang vertically
are called the warp. The thread that is passed backwards and forwards
horizontally is called the weft. Cloth is woven by passing the weft over
one thread and under the next, so that on a loom there must be one set of
threads hanging towards the front and another set hanging towards the
back. The gap between these two sets of threads is called the shed. The
threads at the back are attached by loops to a rod at the front, so that by
pulling this rod forward the back threads can be pulled in front of the other
threads. The rod that changes which threads are to the front is called the
heddle rod. Two or more heddle rods can be used to bring forward the
threads in different orders, which will produce patterns in the cloth itself.
The warp can be prepared for the loom by the production of a
starting border. This is a narrow braid, woven using small cards instead
of a loom, but on one side of the braid the weft is left hanging in very long
loops. The finished braid is attached to the bar along the top of the loom,
and the loops are cut at the bottom so that the back half of the loops drop
back and the front half are brought forward, thus producing the shed. The
last warp threads on either side which will form the edge of the woven
cloth are often bundled together to create a strong edge. This woven edge
is called the selvage or selvedge.
The different groupings of warp and weft threads produce different
weaves, such as tabby where the weft passes under and over single warp
threads, and half-basket where the weft passes under and over two warp
threads. Other weaves include basket weave, two-over-two twill,
herringbone twill and diamond twill.
When woven, the cloth (often already in the shape of a garment) was
sent to the fuller for finishing. The natural grease in sheep’s wool
(lanolin) was removed, if not required to make the cloth waterproof, and
the cloth was shrunk. To make woollen cloth softer, the nap was raised;
this involved running something with a rough surface over the cloth to pull
up the fibres slightly without catching the threads themselves, producing a
fluffy surface. The nap was then trimmed to an even height, given a finish
similar to that found on blankets.

SPRANG

Sprang-work is a method of producing cloth by knotting rather than

weaving (cf crochet and knitting). It was carried out on a small frame,
with adjacent threads twisted round each other.
 BIBLIOGRAPHY

Adams, N. K. and Crowfoot, E. 2001 ‘Varia Romana: textiles from a

Roman army dump’, in Walton Rogers et al 2001, 30-7
Adkin, N. 2000 ‘Did the Romans keep their underwear on in bed?’,
Classical World, 93,
619-20
Allason-Jones, L. 1989 Ear-rings in Roman Britain, British
Archaeological Report, British Series 201
Allason-Jones, L. 2005 Women in Roman Britain (2nd ed.), York
Audollent, M. A. 1922 Les Tombes Gallo-Romaines à Inhumation des
Martres-de-Veyre (Puy-de-Dòme), Paris
Bandinelli, R. B. 1970 Rome, the Centre of Power. Roman Art to AD 200,
London
Barbet, A. 1998 ‘Le tombeau de Silistra’, in Blanc 1998, 114-7
Bender Jørgensen, L. 2000 ‘The Mons Claudianus Textile project’, in
Cardon and Feugère 2000, 253-63
Bieber, M. 1977 Ancient Copies: Contributions to the History of Greek
and Roman Art, New York
Bishop, M. C. and Coulston, J. C. N. 2006 Roman Military Equipment
from the Punic Wars to the Fall of Rome (2nd ed), Oxford
Blanc, N. (ed.) 1998 Au Royaume des Ombres. La Peinture Funéraire
Antique, Saint-Romain-en-Gal
Blanchard-Lemée, M. 1996 Mosaics of Roman Africa: Floor Mosaics
from Tunisia, London
Borriello, M. R., Lista, M., Pappalardo, U., Sampaolo, V., and Ziviello,
C. 1986 Le Collezioni del Museo Nazionale di Napoli, Rome
Bowman, A. K. and Thomas, J. D. 1983 Vindolanda: the Latin Writing
Tablets, Britannia Monograph Series 4
Breve, G. 2003 La Casa Romana di Spoleto, Spoleto
Brodribb, G. 1987 Roman Brick and Tile, Gloucester
Bustacchini, G. 1988 Ravenna, Capital of Mosaic, Ravenna
Capizzi, C. and Francesco, G. 1989 Piazza Armerina: the Mosaics and
Morgantina, Piazza Armerina
Cardon, D. and Feugère, M. 2000 (eds) Archéologie des Textiles des
Origines au Ve Siècle, Montagnac
Carroll, D. L. 1988 Looms and Textiles of the Copts, San Francisco
Cleland, L., Davies, G., and Llewellyn-Jones, L. 2007 Greek and Roman
Dress from A to Z, Abingdon
Coarelli, F. (ed.) 2002 Pompeii, New York
Colledge, M. A. R. 1976 The Art of Palmyra, London
Cottica, D. 2007 ‘Spinning in the Roman world: from everyday craft to
metaphor of destiny’, in Gillis, C. and Nosch, M. B. (eds), Ancient
Textiles: Production, Craft and Society, 220-8
Croom, A. T. 2005 ‘Roman cosmetics’, Arbeia Journal 8, 119-27
Croom, A. T. 2007 Roman Furniture, Stroud
Donati, A. 1998 Romana Pictura, Milan
Doxiadis, E. 1995 The Mysterious Fayum Portraits Faces from Ancient
Egypt, London
Desrosiers, S. and Lorquin, A. 1998 ‘Gallo-Roman period archaeological
textiles found in France’, in Bender Jørgensen, L. and Rinaldo, C., Textiles
in European Archaeology, Göteborg, 53-72
Driel-Murray, C. van 1987 ‘Roman footwear: a mirror of fashion and
society’, in Friendship-Taylor, D. E., Swann, J. M. and Thomas, S., Recent
Research in Archaeological Footwear, Association of Archaeological
Illustrators and Surveyors Technical Paper 8, 32-42
Driel-Murray, C. van 2001 ‘Vindolanda and the dating of Roman
footwear’, Britannia 32, 185-97
Driel-Murray, C. van 2002 ‘Leather ‘bikini’ or briefs’, The Roman Tower
at Shadwell, London: a Reappraisal, Museum of London Archaeology
Studies Series 8, 57-60
Falck, M. von (ed.) 1996 Ägypten Schätze aus dem Wüstensand,
Wiesbaden
Farwell, D. E. and Molleson, T. I. 1993 Poundbury Volume 2: the
Cemeteries, Dorset Natural History and Archaeological Society
Monograph 11
Feder, T. H. 1978 Great Treasures of Pompeii and Herculaneum, New
York
Ferrua, A. 1991 The Unknown Catacomb: a Unique Discovery of Early
Christian Art, Florence
Follette, L. la 1994 ‘The costume of the Roman bride’, in Sebesta and
Bonfante 1994, 54-64
Fremantle, W. H., Lewis, G. and Martley, W. G. 1893 Nicene and Post-
Nicene Fathers, second series, 6
Fuentes, N. 1987 ‘The Roman military tunic’, in Dawson, M. (ed.), Roman
Military Equipment: the Accoutrements of War. Proceedings of the Third
Roman Military Equipment Research Seminar, 41-76
Garbsch, J. 1985 ‘Die norisch-pannonische Tracht’, Aufstieg und
Niedergang der Römischen Welt 12.3, 546-77
Gentili, G. V. 1959 La Villa Erculia di Piazza Armerina: i Mosaici
Figurati, Rome
Gilles, K. 2006 ‘Gold brooch with inscriptions’, in Hartley et al 2006,
136-7
Goldman, B. 1994 ‘Graeco-Roman dress in Syro-Mesopotamia’ in
Sebesta and Bonfante 1994, 163-81
Goldman, N. 1994 ‘Roman footwear’ in Sebesta and Bonfante 1994, 101-
29
Goldman, N. 1996 ‘Isis revealed: cult and costume in Italy’, in Bailey, D.
M. (ed.), Archaeological Research in Roman Egypt, Journal of Roman
Archaeology supplementary series 19, 246-58
Granger-Taylor, H. 2000 ‘The textiles from Khirbet Qazone (Jordan)’ in
Cardon and Feugère 2000, 149-62
Granger-Taylor, H. 2006 ‘Dalmatic’, in Hartley et al 2006, 162-4
Graser, E. R. 1975 ‘The edict of Diocletian’ in Frank, T., An Economic
Survey of Ancient Rome, Volume 5: Rome and Italy of the Empire,
Baltimore, 305-421
Guido, M. and Mills, J. M. 1993 ‘Beads (jet, glass, crystal and coral)’, in
Farwell and Molleson 1993, 100-2
Hald, M. 1980 Ancient Danish Textiles from Bogs and Burials,
Copenhagen
Harden, D. B. 1987 Glass of the Caesars, Milan
Hartley, E., Hawkes, J., Henig, M., and Mee, F. 2006 Constantine the
Great: York’s Roman Emperor, York
Hattatt, R. 2000 A Visual Catalogue of Richard Hattatt’s Ancient
Brooches, Oxford
Hendzsel, I., Istvànovits, E., Kulcsár, V., Ligeti, D., Óvari, A. and
Pásztòkai-Szeöke, J. 2008 ‘“On the borders of east and west”: a
reconstruction of Roman provincial and barbarian dress in the Hungarian
National Museum’, in Gleba, M., Munkholt, C. and Nosch, M. (eds),
Dressing the Past, Ancient Textiles Series 3
Hoffmann, M. 1974 The Warp-weighted Loom, Oslo
Ingalls Wilder, L. 1970 These Happy Golden Years, London
Jenkins, I. 1991 ‘Pendants for eunuchs’, Jewellery Studies 5, 85-93
Jenkins, I. and Williams, D. 1985 ‘Sprang hair nets: their manufacture
and use in Ancient Greece’, American Journal of Archaeology 89, 411-8
Johns, C. 1996 The Jewellery of Roman Britain, London
King, A. 1982 Archaeology of the Roman Empire, Feltham
Kleiner, D. 1992 Roman Sculpture, Yale
Lochhead, V. 1990 ‘The conservation of four Egyptian tunics’,
Archaeological Textiles Occasional Papers 10, 41-8
Lucas, A. T. 1956 ‘Footwear in Ireland’, Journal of the County Louth
Archaeological Society 13, 309-91
Mannering, U. 2000 ‘Roman garments from Mons Claudianus’, in Cardon
and Feugère 2000, 283-90
Michaeli, T. 1998 ‘The pictorial program of the tomb near Kibbutz Or-ha-
Ner in Israel’, Assaph 3, 37-76
Nappo, S. 1998 Pompeii, London
Nicolai, V. F., Bisconti, F. and Mazzoleni, D. 2009 The Christian
Catacombs of Rome: History, Decoration, Inscriptions (3rd edn),
Regensburg
Nowicka, M. 1998 ‘Le portrait funéraire dans la peinture antique’ in Blanc
1998, 120-3
Olson, K. 2003 ‘Roman underwear revisited’, Classical World 96, 201-10
Olson, K. 2008 Dress and the Roman Woman. Self-preservation and
Society, Abingdon
Pfister, R. and Bellinger, L. 1945 The Excavations at Dura-Europos Final
Report IV part II: the Textiles, New Haven
Philpott, R. 1991 Burial Practices in Roman Britain, British
Archaeological Reports British Series 219
Piggott, S. 1967 ‘An ancient Briton in North Africa’, Antiquity 42, 128-30
Pillinger, R. 1984 Studien zu römischen zwischengoldgläsern, Vienna
Pritchard, F. and Verhecken-Lammens, C. 2001 ‘Two wide-sleeved
linen tunics from Roman Egypt’, in Walton Rogers et al 2001, 21-9
Reinhold, M. 1970 ‘History of purple as a status symbol in antiquity’,
Collection Latomus, 116, 35-73
Roussin, L. A. 1994 ‘Costume in Roman Palestine: archaeological
remains and the evidence from the Mishnah’, in Sebesta and Bonfante
1994, 182-90
Scholz, B. I. 1992 Untersuchungen zur Tracht der römischen matrona,
Cologne
Sebesta, J. L. 1994a ‘Tunica ralla, tunica spissa: the colours and textiles
of Roman costume’, in Sebesta and Bonfante 1994, 65-76
Sebesta, J. L. 1994b ‘Symbolism in the costume of the Roman woman’, in
Sebesta and Bonfante 1994, 46-53
Sebesta, J. L. and Bonfante, L. 1994 The World of Roman Costume,
Madison
Selzer, W. 1988 Römische Steindenkmäler Mainz in Römischer Zeit, Mainz
Shelton, J. 1988 As the Romans Did: a Sourcebook in Roman Social
History, Oxford
Sinn, F. and Freyberger, K. 1996 ‘Die Ausstattung des Hateriergrabes’, in
Daltop G. and Oehler, H. (eds), Museo Gregoriano Profano ex-
Lateranense: Katalog der Skulpturen, Mainz, 43-59
Stephens, J. 2008 ‘Ancient Roman hairdressing: on (hair)pins and
needles’, Journal of Roman Archaeology 21, 111-32
Stone, S. 1994 ‘The toga: from national to ceremonial costume’, in
Sebesta and Bonfante 1994, 13-45
Stout, A. M. 1994 ‘Jewelry as a symbol of status in the Roman Empire’,
in Sebesta and Bonfante 1994, 77-100
Strong, D. 1976 Roman Art, Harmondsworth
Strong, E. S. 1969 Roman Sculpture from Augustus to Constantine, New
York
Sumner, G. 2009 Roman Military Dress, Stroud
Taradell, M. 1969 Roman Art in Spain, New York
Tomlin, R. S. O. 1988 ‘The curse tablets’, in Cunliffe, B (ed.), The Temple
of Sulis Minerva at Bath, Volume 2: the Finds from the Sacred Spring,
Oxford University Committee of Archaeology Monograph 16, 59-260
Trilling, J. 1982 The Roman Heritage: Textiles from Egypt and the Eastern
Mediterranean 300 to 600 AD, Washington
Veyne, P. (ed.) 1987 A History of private Life, Volume 1: From Pagan
Rome to Byzantium, Cambridge
Walker, S. and Bierbrier, M. 1997 Ancient Faces: Mummy Portraits from
Roman Egypt, London
Walton Rogers, P. 2007 Clothing in Early Anglo-Saxon England AD 450-
700, CBA Research Report 145
Walton Rogers, P., Bender Jørgensen, L. and Rast-Eicher, A. (eds) 2001
The Roman Textile Industry and its Influence, Oxford
Waterer, J. W. 1976 ‘Leatherwork’, in Strong, D. and Brown, D. (eds),
Roman Crafts, London, 178-93
Weitzmann, K. 1977 Late Antique and Early Christian Book Illumination,
New York
Weitzmann, K. (ed.) 1979 Age of Spirituality, New York
White, K. D. 1977 Country Life in Classical Times, London
Wild, J. P. 1963 ‘The byrrus Britannicus’, Antiquity 37, 193-202
Wild, J. P. 1968 ‘Clothing in the north-west provinces in the Roman
Empire’, Bonner Jahrbücher 68, 166-240
Wild, J. P. 1970 Textile Manufacture in the Northern Roman Provinces,
Cambridge
Wild, J. P. 1985 ‘The clothing of Britannia, Gallia Belgica and Germania
Inferior’, Aufstieg und Niedergang der Römischen Welt 12.3, 362-423
Wild, J. P. 1987 ‘The dress of the Lullingstone figures’, in Meates, G. W.,
The Roman Villa at Lullingstone, Kent, Volume 2: the Wall Paintings and
Finds, Maidstone, 40-1
Wild, J. P. 1993 ‘The textiles’ in van Driel-Murray, C., Wild, J. P.,
Seaward, M., and Hillam, J., Vindolanda Research Reports, New Series,
Volume III: the Early Wooden Forts. Preliminary Reports on the Leather,
Textiles, Environmental Evidence and Dendrochronology, 76-90
Wild, J. P. 2002 ‘The textile industries of Roman Britain’, Britannia 33,
1-42
Wilmot, T. 1982 ‘Excavations at Queen Street and Roman wells in
London’, Transactions of the London and Middlesex Archaeological
Society 33, 1-78
Wilson, L. M. 1924 The Roman Toga, Baltimore
Wright, R. P. 1955 Catalogue of the Roman Inscribed and Sculptured
Stones in the Grosvenor Museum, Chester, Chester
Young, A. 2000 ‘Representation of cloth vendors and the cloth trade on
funerary reliefs in Roman Gaul and Italy’, in Cardon and Feugère 2000,
215-33