University of Virginia Library

V. 7

DEVICES FOR HEATING,
LIGHTING, VENTILATION,
AND COOKING

V.7.1

THE CENTRAL HEARTH AND THE
LOUVER

THE MEANING OF LOCUS FOCI AND TESTU

One of the striking typological characteristics of the guest
and service buildings on the Plan is the squares in the
center of these houses referred to varyingly as locus foci and
testu (fig. 358A-C). The first of these terms clearly refers
to the "fireplace" or open hearth in the middle of the floor
which heats the house. The second, testu, requires some
explanation. It has generally been taken to be an abbreviation
for the word testudo ("turtle" or "roof"),[226] but I do
not think that this is the correct interpretation. Testu, as an
abbreviation for testudo, with no sign given to indicate that
a part of the word is missing, is not consistent with the
author's other abbreviations,[227] and there is no reason to
assume that he departed from his normal procedure because
of the smallness of the space in which the word had
to be inscribed. The testu square in the Hospice for Pilgrims
and Paupers (fig. 358B) is large enough to accommodate
the whole word testudo and several other words if
necessary. What the scribe had in mind, in my opinion, was
the word testu, exactly as it is written—a rare yet perfectly
meaningful term.

Testu (indeclinable) means the "lid of a pot."[228] It is
closely related to testa, which means "shell," either the
shell that covers a testaceous animal, the human skull
(testa hominis), or human artifacts of comparable construction,

[ILLUSTRATION]

359. WIJCHEN, GELDERLAND, THE NETHERLANDS

This Iron Age house of modest design was unearthed on a chain of hills skirting
a moor that once may have formed an outer arm of the Maas. The house was
internally divided into two areas one of which contained its hearth. Four
independent posts surrounding the hearth are most reasonably explained as the
supports of a small roof protecting an opening in the main roof admitting light
and air and also allowing the smoke to escape.

NEAR THE RIVER MAAS (10KM WEST OF NIJMEGEN)

PLAN. IRON AGE HOUSE [after F. Bloemen, 1933, 5, fig. 5]

such as clay bowls or pots. Testudo is a derivative of
both these words. It shares with them the basic meaning
of "protective cover." In everyday language testudo meant a
"tortoise" or "turtle"; in military language it was the name
for the protective covering formed when soldiers held their
shields overhead and locked them together. By analogy, in
architectural terminology—both classical and medieval—
testudo came to be the word used for "roof," usually a roof
of timber, but also by extension, a "vaulted roof."[229]

Even supposing that the author of the Plan had in mind
testudo, rather than testu, it is unlikely that he referred to
the principal roof of the building; rather, he must have
meant a roof equal in size with the square in which
the term was written; and since this square is designated
both as testu and as locus foci, it is most probably to be understood
as "a protective shield above the hearth," the purpose
of which must have been to form a cover for a central smoke
outlet. Such openings in the roof above the hearth are, in
fact, a feature of the protohistoric and early medieval building
tradition just discussed, and they remained an intrinsic
part of vernacular buildings throughout the Middle Ages.

[ILLUSTRATION]

360. KRAGHEDE, VENDSYSSEL, DENMARK

This Danish Iron Age House is of very similar design to the Dutch specimen
shown in the preceding figure. It shows four independent inner posts related to
the hearth in an identical manner. In houses of relatively small dimensions it
made sense to support the hearth-protecting lantern by poles rising from the
ground. In larger houses, as the subsequent figures show, this was accomplished
by timbers forming part of the roof frame.

PLAN. IRON AGE HOUSE

[after Hatt, 1928, 254, fig. 25]

THE LIGHT AND SMOKE HOLES OF THE
NORDIC SAGA HOUSE

This roof device is well attested in the Nordic Sagas where,
according to its function, it is referred to varyingly as
"smoke hole" (reykháfr, reykberi), "light inlet" (ljóri), or
"air inlet" (vindgluggr, vindauga).[230] Little is known about
the size and shape of these devices, but apparently they
were large enough to be used as an escape hatch when all
other passages were blocked. The Sagas abound with tales
of exits made in this manner. A passage from the Vatnsdœla
Saga gives a typical example: "And so was this [house]
arranged that from that pile of goods, one could step up
into a big smoke hole [í einn storan reykbera] which was
over the hall [er á var skálanum] and when the marauder
investigated the pile, þorsteinn was outside" [var þorsteinn
úti, the sense being: þorsteinn had gained his freedom by
escaping through the smoke hole].[231]

The openings of these light and smoke holes could be
closed by means of wooden shutters (spjaeld) or boards
(fjöl) which were placed in position with the help of a pole

[ILLUSTRATION]

361. NEW COLLEGE, OXFORD, ENGLAND

THE GREAT HALL, 1378-1386

[after Loggan, 1675]

The lantern-surmounted ridges of these buildings built approximately one hundred years apart attest the effectiveness of the device on larger
structures of monumental character.

or rope; or they were screened with transparent membranes
made from the stomach lining of a hog mounted on movable
frames (skjágluggr, skjávindauga). We read of the first type
in Haralds saga harđrađa: "The king then let a board
(fjöl) be moved in front of the light hole (ljórann) so that
only a small opening was left . . . Einar entered and said,
`Dark is it in the King's Council Hall (málstofa).' At the
same moment men rushed on him. . . ."[232] The second type is
mentioned in an equally dramatic passage of the GullÞoris
saga, where Þorir, finding himself trapped in the hall
by Þorbjörn's housecarls, with all exits blocked, "grabbed
a pole and raised it under the `skin hole' (skjárinn) and
there went out and pulled up the pole, and then ran up to
the mountains."[233]

PROTOHISTORIC EVIDENCE FOR
LIGHT AND SMOKE HOLES

Evidence for the existence of poles rising from the ground
to form a canopy around and over the hearth has been
found in aisled Iron Age houses at Hodorf, Germany (fig.
307), and Wijchen, Holland (fig. 359), as well as in the

Migration Period houses of Nauen, Germany; also in
single-span Iron Age houses at Kraghede, Vendsyssel,
Denmark (fig. 360); Källberga, Alunda (Uppland), Sweden;
and the Migration Period village Vallhagar on the island
of Gotland, Sweden.[234]

[ILLUSTRATION]

363. NAPA VALLEY, CALIFORNIA

CALIFORNIA DAIRY BARN WITH LOUVERS

The most common and widely diffused survival form of the testu of the Plan are the
lantern-surmounted openings in the roofs of barns found at every community of the
great farm belt of Canada and the United States.

The relatively rare occurrence of these hearth poles
suggests, however, that in general the protective shields
were mounted directly on the roof rather than on special
supports. The latter system would have been entirely inappropriate
in larger halls, since it would have required an
underpinning of timbers entirely out of scale with the
superstructure that it served to support.

MEDIEVAL EXAMPLES

A fairly accurate picture of these smoke and air hole
coverings may be obtained from some of the old engravings
of early English college halls, for instance, those of Oxford's
New College (fig. 361) and Magdalen College (fig. 362), as
shown in David Loggan's illustratious Oxonia Illustrata of
1675.[235] New College Hall, the oldest of the surviving college
halls of Oxford, was built between 1378-1386 by William
of Wykeham. Magdalen College was founded in 1448 by
William of Waynflete, but its buildings were not completed
until 1480. Both these halls were built during a period when
new discoveries in the technique of roof construction made
it possible to dispense with the two rows of roof-supporting
posts which formerly divided the hall into a nave and two

accompanying aisles. Thus it became possible not only to
cover the space in a single span but also to lift the roof upon
walls of considerable height. Yet even in this new and more
fashionable hall, which permitted large windows, the traditional
opening in the roof above the hearth was retained
as the principal exit for smoke. The roof of the hall of
Magdalen College shows what extraordinary dimensions
these openings could obtain.

The medieval term for these smoke holes is fumerium
("smoke hole") or lovarium (identical with Old French
louvert, "opening"). The so-called Liberate Rolls of King
Henry III, issued in 1216 and 1272 (verbal directive for
repair and construction of houses owned by the crown)
make frequent reference to these devices.[236] Loggan's engravings
of the halls of New College and Magdalen College
show how these smoke holes were covered by a simple
saddle roof, which looks like a portion of the main roof cut
out and raised over the hole. In Loggan's Oxonia Illustrata
saddle-like louvers appear only on the roofs of the earlier
college halls. On the roofs of the later halls the saddle-like
louver was replaced by a flèche or lantern, a Gothic development
and one which the author of the Plan of St. Gall
is not likely to have envisaged.[237]

[ILLUSTRATION]

365. JEAN LE PRINCE. LES LAVANDIÈRES, 1770, PARIS

ÉCOLE DES BEAUX-ARTS, COLLECTION ARMAND-VALTON

(AFTER LE PAYSAGE FRANCAIS. 1926, PL. LV)

The saddle-like version is the simpler and, unquestionably,
the older form, and this type of fumerium is in my
opinion what the author of the Plan had in mind when he
used the term testu.

MODERN SURVIVAL FORMS

Superstructures of this type, almost extinct in the Old
World, are a common feature of the timbered barns in the
great farm belt of the United States (fig. 363). This device
was brought over by early settlers, along with the very type
of building for which it had been invented in the Early
Iron Age. Hundreds of thousands, if not millions, of aisled
American hay and dairy barns of timber are ventilated even
today by openings in the roof ridge, which are shielded by
elevated sections of the main roof and which still retain the
shape of what in the Middle Ages was probably the most
common means of controlling light and smoke.

The design of such a device from a barn in the vicinity
of Benicia, California (fig. 364), is probably as good a guide
for the reconstruction of the louvers of the guest and service
buildings on the Plan of St. Gall as any equivalent found in
Europe, where this particular device disappeared rapidly in

[ILLUSTRATION]

367. GRIMANI BREVIARY

VENICE, LIBRARY OF ST. MARK'S

LABORS OF THE MONTH OF FEBRUARY, DETAIL, fol. 2v

[photo: Alinari 39316]

residential architecture once the open fire was replaced by
hooded chimneys, a development that must have been
nearly complete by the beginning of the fifteenth century.

Despite a thorough search among Flemish, Dutch, and
German landscape drawings, etchings, and paintings—
media which have richly and vividly preserved the architectural
panorama of the medieval countryside—I have
been able to trace only a single case of survival in post-medieval
architecture, and a belated one at that. In an ink
drawing of 1770 by Jean le Prince, entitled Les Lavandières
(fig. 365) there is shown in the center of the bridge that
crosses a stream an old rectangular house with an opening
in the ridge which is shielded by a raised portion of the
main roof above the spot where in the period of construction
of this house there must have burned an open fire.
Together with the saddle-shaped superstructures over the
ridge of the halls of New College and Magdalen College at
Oxford, this drawing of Jean le Prince may retain the most
truthful visual record of the device which in the guest and
service buildings of the Plan is referred to as testu.

[ILLUSTRATION]

368. DUTCH BIBLE (UTRECHT, 1465)

RUTH LYING WITH BOAZ

VIENNA NATIONAL LIBRARY, ms. 2177, fol. 153v.

[after Byvanck and Hoogewerff, I, 1922, 116.B]

The details of the roof flaps und the curved levers by means of which they were
opened and shut are clear enough to leave no doubt about their identity with the
18th-century examples shown in figs. 369 and 370. Observe in the background the
haystack with a roof that can be lowered and raised, a Carolingian example of
which is shown in fig. 326.F.

HINGED HATCHES

However, in the Lowlands during the sixteenth century,
there existed a related device which, judging from the frequency
of its appearance in the illuminations of the Grimani
Breviary and other Franco-Flemish manuscripts of the
same period, must also have been a common feature in late
medieval house construction. A considerable number of
houses represented in the landscapes of these manuscripts
have smoke holes covered by wooden hatches hinged to the
ridge, which could be raised or lowered by means of pulleys.
This device appears in three different places in the Grimani
Breviary: the July representation (fig. 366), the March
representation, and the well-known February representation
(fig. 367), in which a wisp of smoke can be distinguished
rising in gentle spirals into the chilly winter air from an
open fire burning directly below the smoke hole on the
simple clay floor.[238] Again, it is depicted in the September
representations of the breviary of the Museum Mayer van
den Bergh at Antwerp[239] and in an illustration, which depicts

[ILLUSTRATION]

ROOF FLAPS ON AN 18TH-CENTURY FARM, THE NETHERLANDS

369. CROSS SECTION

The roof flaps are shown in closed position, with cord arrangement visible.

[after Uilkema, 1916, 21, fig. VII, and 27, fig. IX]

A simple arrangement of cords and eyes or small blocks caused the right-hand cord to open the left-hand flap, and vice-versa. These cords
were cleated off below to maintain the flap open which, when released, closed of their own weight.

[ILLUSTRATION]

370. EXTERIOR PERSPECTIVE

One roof flap is open (the second is omitted from the drawing for clarity).

the lying together of Ruth and Boaz, from a Dutch Bible
illuminated around 1465 (fig. 368).[240] The technical details
of how such roof flaps were operated are well explained in
the sketches of a vent of a Frisian farmhouse of the eighteenth
century, published by K. Uilkema (figs. 369 and
370).[241]

A wooden hatch or lid of this kind would be in complete
accord with the term testu ("lid"), but the dimensions of
many of the testu squares of the Plan of St. Gall, some of
which are as large as 10 feet square (Hospice of the Paupers,
House for Distinguished Guests), speak against this being
the type used. Hinged lids of such dimensions would be
unmanageable. The saddler roof is the simple and the earlier
form, and for this reason in our reconstruction of the guest
and service buildings on the Plan, we have chosen the latter
version.

V.7.2

CORNER FIREPLACES WITH CHIMNEYS

A PREROGATIVE OF HIGHER-RANKING MEMBERS
OF THE MONASTIC COMMUNITY

In addition to the central open fireplace which forms the
primary—and in the majority of houses, the only—source
of heat, some of the guest and service buildings are provided
with another device for heating individual rooms. Its
symbol is an ovoid loop (fig. 371 A and B) always found in
the corner of a room. In the Abbot's House, it is designated

[ILLUSTRATION]

PLAN OF ST. GALL

371.A ABBOT'S HOUSE

371.B HOUSE FOR DISTINGUISHED GUESTS

CORNER FIREPLACES

Corner fireplaces with chimney stacks are installed in double-storied structures and
in houses where the seclusion of separate living or bedrooms deprives occupants of
the heat of the central fireplace (as in the House for Distinguished Guests).

as caminata. Caminata (short, presumably, for camera
caminata) is the medieval word for a room that has its own
fireplace (caminus).[242] In this sense the term is used to indicate
the bedrooms for distinguished guests (caminata cum
lectis; fig. 371B) and the bedroom in the Porter's lodging
(caminata portarii). By contrast, in the Abbot's living room
as well as his bedroom, both of which are heated by corner
fireplaces, the word caminata is inscribed into the interior
of the loop-shaped symbol that indicates the presence of
these heating devices, and therefore must have been used,
in these two instances, as synonyms for caminus, i.e. the fireplace
itself.

Individual fireplaces indicated in this way on the Plan
are found either in buildings having no central hearth, or
in rooms separated from the common hall by wall partitions
for the sake of greater privacy. Such fireplaces were the
prerogative of the higher ranking members of the monastic
polity and of the sick. They are primarily associated with
masonry structures. Besides appearing in the Abbot's
House, they are also found in the living quarters of the
monastic officials next in rank: the Porter, the Master of

the Outer School, the Master of the Novitiate, the Master
of the Infirmary, and the Master of the Hospice for Pilgrims
and Paupers.[243] For reasons of health they occur in
all the sick wards (Novitiate, Infirmary, House for Bloodletting,
House of the Physicians). They are found with less
frequency in the guest and service buildings. Here again
their presence is determined by considerations of rank or
functional necessity. We find them in the bedrooms of the
noblemen in the House for Distinguished Guests, in the
bedrooms of the Physician and the Gardener. Conversely,
they never occur in the buildings that accommodate the
humbler members of the monastic community. They are
completely absent from the houses for the workmen and
craftsmen, the Hospice for Pilgrims and Paupers, and the
houses that shelter serfs or livestock and their keepers.

ETYMOLOGY AND SHAPE SUGGEST DERIVATION
FROM THE OVEN

Caminus comes from the Greek κάμινος, which meant an
"oven, furnace, or kiln for smelting, baking, or burning

[ILLUSTRATION]

373. LE PUY-EN-VELAY (HAUTE-LOIRE), FRANCE

WALL FIREPLACE, 12TH CENTURY

One of the earliest and most elegant of the surviving medieval fireplaces, it is set
against a flat wall, its conical hood constructed with consummate skill supported by
cusped brackets rising from two short columns.

earthenware and bricks."[244] In classical Latin the term
apparently had come to mean "a furnace which supplies
the heat for a room or an apartment."[245] On the Plan of St.
Gall it is used in this sense in connection with the large
firing chambers (caminus ad calefaciendü) of the hypocausts
which heated the Monks' Dormitory and Warming Room
as well as the living and sleeping quarters of the novices and
the sick. In medieval Latin caminus is the standard term for
a wall or corner fireplace with a chimney, as may be inferred
from Old High German glossaries, where it is translated
both by "oven" (ofan) and "chimney" (scorenstein).[246] The
etymology of the term caminus—its original meaning of
"kiln" or "oven" and its eventual change to mean chimney
(French: cheminée; German: Kamin)—suggests that the
medieval wall or corner fireplace is, developmentally, an
offspring of the baker's oven or potter's kiln. This assumption
makes sense for functional as well as etymological
reasons. When the fire was moved against the wall, it had
to be enclosed, and the age-old solution for enclosing a fire
was the ovoid or round oven of the baker or potter. When
the baking oven had, in this manner, been transformed into

[ILLUSTRATION]

374. FRANKFURT-AM-MAIN, GERMANY

RÖMERTURM, CORNER FIREPLACE, 13TH CENTURY

The hood has the shape of a beehive. Other examples of this design are found
in the Berchfried of Castle Schönburg, near Naumburg (by some ascribed to the
11th century) and in the Berchfried of Petersberg near Freisach (see Piper,
1895, figs. 476 and 480).

a heating unit that formed an integral part of the living
room, the proper functioning of the fire and the health of
the people whom it served required a smoke flue. At precisely
what point in history this feature was introduced is
as yet uncertain.

[ILLUSTRATION]

375. PLAN OF ST. GALL. CORNER FIREPLACE

HOUSE FOR DISTINGUISHED GUESTS

Our reconstruction of the corner fireplaces in the bedrooms of the guests is
related to the design of the chimney shown in fig. 373. The hood masonry rests
on a lintel fashioned by contilevered curved stones seated deep in the masonry,
separated on corbels likewise deeply embedded. A voussior as illustrated, or a
joint, may occur at center.

TIME OF INVENTION OR ADOPTION
IS PROBLEMATIC

Moritz Heyne[247] and Joseph Schepers[248] ascribe its invention
to the Romans, but this supposition has recently
been shattered by André Parrot's extraordinary discovery
of two fireplaces from the bathrooms of the Palace of Mari
in Mesopotamia, which date from the beginning of the
second millennium B.C. (fig. 372).[249]

The smoke flues of these fireplaces consist of conical
hoods encasing a vertical stack of tubular flue tiles, with an
opening at the bottom for the fire which burned on a platform
that formed a quarter of a circle. Hence the Romans
cannot claim to be the first inventors of this device. They
may have rediscovered it, but until a chimney-type Roman
corner fireplace has actually been excavated—and so far no
one has had the good fortune to find one—even this assumption
must remain hypothetical. Conversely, it must be
stressed that wall or corner fireplaces with chimneys were
not a feature characteristic of Germanic house construction,
and were not known at all in the northernmost areas
held by Germanic peoples. We know this is so not only
because of hundreds of house sites that have actually been

[ILLUSTRATION]

377. SANNAP, HALLAND, SWEDEN

[after Erixon, 1947, 426, fig. 545]

CORNER FIREPLACE WITH OVEN
SURMOUNTED BY CONICAL
HOOD

excavated, but also because, when chimneys were finally
introduced at the Norwegian court during the reign of Olaf
Kyrre (1067-93), this was an event so unusual that it was
considered worthy of being recorded in Snorri's Lives of the
Kings of Norway: "It was an old custom in Norway that
the King's high seat was in the middle of the long bench.
The ale was borne round the fire. King Olaf was the first
to install corner fireplaces."[250] (Ofnstofur, the Old Norse
term, like the Latin caminus retains etymological consciousness
of the fact that the masonry fireplace is an offspring of
the oven !) Iceland resisted this innovation even longer. The
first masonry-built Icelandic wall fireplace was constructed
in 1316, in the timbered hall of Bishop Laurentius at
Hólar.[251]

EARLIEST VISUAL & LITERARY EVIDENCE
IN MEDIEVAL EUROPE

The Plan of St. Gall, to the best of my knowledge, offers
the earliest visual evidence of the use in Europe of corner
fireplaces with chimneys. Literary evidence, however, of
such heating devices in individual rooms goes back as far
as the sixth century. In 584, in connection with a donation
to the church of St.-Marcellus ad Cabillonum (Châlons-sur-Saône),

[ILLUSTRATION]

378. MUSEÉ CONDÉ, CHANTILLY,
FRANCE

LES TRÈS RICHES HEURES DE JEAN DE
FRANCE DUC DE BERRY.

FEBRUARY: DETAIL.

The miniature depicts rural life during the month of
February, portraying a snow covered landscape with a
farmhouse, a pigeon house and a stack of beehives. Inside
the farmhouse the mistress and two servants warm themselves
before a crackling fire. The hood and mantle of the fire
place are built in masonry. The chimney shaft is braided
in wicker-work, presumably daubed inside to prevent it
from catching fire.

The manuscript, one of the finest of its kind, was illuminated
between 1411 and 1416 by Pol de Limburg, the most
distinguished of a small group of Flemish artists who,
trained in the tradition of French illumination of the
fourteenth century, under the inspiration of contemporary
Italian painting laid the foundations for the realistic style of
the brothers van Eyck.

the Frankish king Gunthram I directed the
construction of a royal guesthouse (hospitole), the description
of which (solarium cum caminata and lobia, galleried
porch) is strikingly reminiscent of both the Abbot's House
on the Plan of St. Gall and the royal residence at Anappes
of the Brevium exempla.[252]

EARLIEST EXTANT MEDIEVAL
CORNER FIREPLACES

The earliest extant Continental chimneys date from the
twelfth century. They form niches in the masonry walls and
are surmounted by conical hoods often braced at the sides
by pillars and brackets. They are usually constructed on a
full circular plan, the heating chamber forming the rearward,
and the hood the forward, segment of the circle. Two
typical examples of this species, dating from the twelfth
and thirteenth centuries, respectively, are found in Le Puyen-Velay
(Haute-Loire), France (fig. 373)[253] and the so-called

[ILLUSTRATION]

379.A SAALBURG, HESSE, GERMANY. FORTIFIED ROMAN FRONTIER CAMP

"PILLARED" HYPOCAUST [after Fusch, 1910, pl. xv]

Although slow in making warmth felt in the rooms served
by the system, the hypocaust had the great advantage,
once the building was warmed by convection and
conduction, of providing through radiation from walls and
floor an even temperature for long periods of time, in
relatively large spaces, using only a small amount of fuel.
Ideally suited for the large warm and hot rooms of
Roman baths, the hypocaust was used in the northern
provinces of Rome almost universally in private homes.

Römerturm in Frankfurt a.M., Germany (fig. 374).[254]
This hooded circular fireplace, in my opinion, is the type
that the draftsman of the Plan had in mind when he used
ovoid symbols for the private heating units of the leading
monastic officials and the monastery's distinguished guests
(fig. 375). In this connection attention must be drawn to
certain oven-shaped corner fireplaces still in use today in
rural buildings of Sweden, two typical examples of which
are given in figure 376 and 377.[255] Both of these are, in fact,
oven and fireplace combined.

I do not doubt that the corner fireplaces of the Plan of
St. Gall were intended to be built in masonry, although
there is evidence for the existence in the Middle Ages of
fireplaces with wooden hoods. A group of wooden chimneys,
mounted on frames of oak and filled with wattle and
daub, was published in Nathaniel Lloyd's History of the
English House.[256] We may assume that fireplaces constructed
entirely of wood or of both wood and stone were equally
common on the Continent, because of their appearance in
late medieval manuscripts and paintings. A typical example
of this mixed technique is the fireplace of the farmer's
house in the charming winter landscape of the February
representation of Jean de France's Très Riches Heures (fig.
378),[257] and an example of a fireplace built entirely of wood
is the one at the rear wall of a wooden cottage in the Livre du
Cuer d'Amours Espris of René, Duke of Anjou.[258] These
medieval wall and corner fireplaces with wooden chimneys
were, in my opinion, not an original conception, but rather
an adaptation to Northern building materials, performed on
a relatively humble social level, of a heating device that
formed no part of the Northern building tradition.

WESTWARD DIFFUSION FROM THE NEAR EAST
BY-PASSING ROME?

Roman custom of heating a house or its individual apartments
by means of hypocausts stands in marked contrast
to the open fire that burned on the floor of the Germanic
house. The Roman heating unit was not only enclosed; it
was concealed. The medieval open chimney combined the
advantages of both; the fire was enclosed, as in the Roman
type, and yet it was visible, as in the Germanic open fireplace.
We do not know exactly when or where this combination

[ILLUSTRATION]

379.B ST.-REMY, BOUCHES-DU-RHÔNE (NEAR ARLES), FRANCE

RUINS OF A "PILLARED" HYPOCAUST IN A HOUSE OF THE ROMAN SETTLEMENT OF GLANUM

These two illustrations show typical use of the Roman
hypocaust system in houses of transalpine Europe. No
Roman villa of any significance in the vast stretches of
land that extended from Provence to the borders of
Scotland lacked such a facility.

first took place. One might be tempted to guess that
it occurred in an area where Roman and Germanic culture
merged. But Parrot's discovery of corner fireplaces with
chimneys in the Mesopotamian Palace of Mari, dating as
early as the beginning of the second millennium B.C., suggests
that we have to contend with a third influence, from
the East. It is possible that the medieval wall or corner
fireplace is a Near Eastern idea, cast into Roman masonry
in Merovingian times, which permitted the installation of
open fires in individual rooms, without endangering the
safety of the building. Perhaps it was the close ties established
between the Near East and the West through the
monastic conquest of Merovingian Europe, in the fifth
century, as well as the ubiquitous presence in the ports and
inland cities of Gaul of Syrian, Egyptian and Jewish tradesmen
that opened up the channels for the westward diffusion
of this heating device which seems to have bypassed
Rome.[259]

[ILLUSTRATION]

380. SILCHESTER, HAMPSHIRE, ENGLAND

Roman hypocaust of the channeled type. The floor of the room is removed to reveal the hypocaust substructure. This system, although not
quite so common, was as widely diffused as the pillared type.

[after Joyce, 1881, pl. vii]

V.7.3

HYPOCAUSTS

TWO TRADITIONAL ROMAN TYPES:
CHANNELED AND PILLARED

Since the history of the Roman hypocaust has been fully
documented elsewhere, we may confine ourselves here to
the most summary review.[260] This heating system, developed
to perfection by the Romans, is found not only in their
baths but also in practically every Roman villa north of the
Alps. The Roman hypocaust was either of the channeled
or the pillared type.[261] In the latter, a good example of which,
from the Roman camp Saalburg, is shown in fig. 379, the
floor of the room to be heated was raised by short columns,
usually two feet high, a shallow chamber thus being formed
below the floor level. The heat, generated in a furnace that
was built against one of the outer walls and serviced from
the outside, was dispersed into this chamber and rose from
there in vertical flues imbedded in the walls. In the other
type, hot air was taken from the furnace through a trench
beneath the floor to the center of the room and then diverted
radially to the four corners into a channel running all the
way around the room, at the bottom of the walls, from which
point it rose into the wall flues, as in the hypocaust from a
building in block II of the Romano-British city of Silchester
(fig. 380).[262]

THE CHANNELED HYPOCAUST OF THE
MONKS' DORMITORY

On the Plan of St. Gall only those parts of the hypocaust
are shown which lie outside the warming rooms, namely
the furnaces and the chimneys (fig. 381). We are told
nothing about the layout of the ducts and flues that distributed
the heat in the rooms themselves. To do this in the
calefactory of the monks would have been impossible, as it
would have obscured the layout of the beds in the Monk's
Dormitory, which the drafter of the Plan considered to be
of greater importance. But in the warming rooms of the
Novitiate and the Infirmary, with no story above, he could
have gone into detail. Since he chose not to do this, it then
becomes clear that in his day the construction of a hypocaust

[ILLUSTRATION]

381.A PLAN OF ST. GALL. FIRING CHAMBER & SMOKE STACK OF HYPOCAUST

MONKS' WARMING ROOM

see INDEX TO BUILDING NUMBERS OF THE PLAN VOL. I page xxv, VOL. III page 14

The hypocaust services a two-storied
structure, 40 feet wide and 85 feet long,
which contains on the ground floor the
Monks' Warming Room and on the upper
level the Dormitory. Of the heating system
itself only the firing chamber and the smoke
stack are shown. Their existence at two
opposite ends of the building postulates the
presence of a system of connecting ducts in
the subfloor which provides for draft to
distribute the heat and allows the smoke to
escape through a smoke stack placed at a
safe distance from the main structure.

was a matter of common knowledge, one that required
no further instruction.

The furnaces are designated by the words fornax
(cloister walk in front of the Monks' Calefactory), caminus
ad calefaciendū (firing chamber of Monks' Calefactory),
and camin' (Novitiate); the smoke flues, by euaporatio fumi
(Monks' Calefactory) and exitus fumi (Novitiate). There
is no reason to assume that the furnaces were meant to lie
directly beneath the floor of the rooms they heated.
Nowhere else on the Plan has the architect designated
anything at the side of a building which was meant

[ILLUSTRATION]

382.A PLAN OF ST. GALL: BAKING OVENS

MONKS' BAKE AND BREWHOUSE

The daily need for bread of the monastery's planned permanent inhabitants was 250 to 270 one-pound loaves. The oven in the Monks' Bakery
with a diameter of 10 feet, could produce 300 to 350 loaves in one cycle of firing and baking (see p. 259 n. 26 for more detail).

to be beneath it. External firing chambers, moreover, were
traditional. Nor can there be any doubt about the location
of the smoke flues. They are meant to be where they are
shown: at a considerable distance from the outer walls of
the buildings they served, in order to keep the smoke away
from the windows of the dormitories and to reduce the
danger of their roofs being ignited by glowing cinders.

The hypocaust is superior as a heating system to the
hearth or the corner fireplace when large rooms and many
people are involved, since it is capable of distributing the
heat evenly throughout the width and length of the building.
On the Plan of St. Gall this heating system was provided
for the rooms that served as general work and reading
areas for the monks, the novices, and the sick.[263]

Since in the Carolingian cloister, the dormitory was
usually above a room warmed in this manner, the heat
could be transmitted to it through openings in the ceiling
or through wall flues.

[ILLUSTRATION]

382.B BAKE AND BREWHOUSE FOR DISTINGUISHED GUESTS

see INDEX OF BUILDING NUMBERS OF THE PLAN VOL. I, page xxiv; VOL. III, p. 14

The oven of the House for Distinguished Guests, with a diameter of
7½ feet, could have produced easily in one cycle of firing and baking
200 to 250 one-pound loaves. The oven could therefore have
accommodated the needs of the emperor and his complete entourage
who might from time to time be expected to visit the monastery.

As I do not know of any evidence for the existence in
the Middle Ages of pillared hypocausts, I am inclined to
assume that the hypocausts of the Plan of St. Gall belonged
to the channeled type. The Carolingian hypocaust of the
monastery of Reichenau, at any rate, belonged to this type,
and this holds true also of the tenth-century hypocaust
unearthed by Seebach at Pfalz Werla (fig. 209A-C).[264]

V.7.4

WINDOWS

There is only one place on the Plan of St. Gall where
windows are actually shown in the drawing: in the walls of
the Scriptorium (fig. 99) where they are functionally of
vital importance. The symbol used there, two short parallel
strokes intersecting the wall at right angles, is identical
with the one which the draftsman used to designate the

[ILLUSTRATION]

382.C BAKE AND BREWHOUSE FOR PILGRIMS AND PAUPERS

The number of visitors expected to lodge each night in the Hospice
for Pilgrims and Paupers may not have exceeded twelve (see below,
p. 144). But if customs prevailing at the monastery of Corbie
reflected general conditions, these transient guests were issued bread
rations considerably larger than were allotted to the monastery's
regular inhabitants: 3½ pounds per person upon arrival, and half
that amount on departure. For a complement of twelve guests this
distribution would amount to 63 pounds of bread each day. During
the great religious festivals these amounts rose steeply as the number
of travellers increased; even under ordinary conditions it would have
been necessary to add to the needs of the overnight guests those of
the transients who might stop for a noon meal and then move on.
This explanation would account for the large size of the Hospice
oven, its 7½-foot diameter being identical to the oven of the House
for Distinguished Guests.

presence of doors, except that in the case of the Scriptorium
the line that defines the course of the walls is not interrupted
by these strokes as it is in the majority of the entrances,
doors, and exits. In other buildings, such as the Church,
the Abbot's House, or the large complex that contains the
Novitiate and the Infirmary and their chapels, windows
were so clearly an integral part of the building type that the
draftsman felt it unnecessary to go into any detail in this
matter.

In the case of the guest and service buildings, however,
conditions were different. Traditionally, this type of house
was not provided with windows. Lighting, ventilation, and

[ILLUSTRATION]

LANGENBECK, HARBURG, GERMANY

384.A

384.B

LANGOBARDIC BAKING OVEN

[after Adrian, 1951, 69, fig. 2]

This later oven, retaining the daubed wattle shell of its ancestor (fig. 383),
reveals a sophisticated awareness of heating insulation, with its boulder-lined
fire and baking chamber sunk below the surface of the ground.

[ILLUSTRATION]

385. ROME

MONUMENT OF THE BAKER EURYSACES. DETAIL. FIRST
CENTURY B.C.

[after Singer et. al, II, 1956, 118, fig. 88]

The frieze appears on the funerary monument of a wealthy plebeian master
baker who prospered during the final years of the Republic. The oven appears
in a form it has retained into modern times.

[ILLUSTRATION]

MESOPOTAMIA, PALACE OF MARI. BAKING OVEN

386.A

386.B

BEGINNING OF 2ND MILLENNIUM B.C.

smoke control, as we have seen, were regulated by an opening
in the roof. This does not categorically preclude the use
of wall windows. As the need for higher degrees of privacy
led to the installation of corner fireplaces as secondary
sources for heat in rooms that were segregated from the
common hall by internal walls and ceilings, so the separation
of these rooms from the central source of light made necessary
the installation of supplementary devices for the
admission of light and air. This could take the form either
of wall windows or of dormer windows.

In the guest and service buildings of the Plan of St. Gall,
accordingly, we may have to take into account the entire
range of possibilities, from complete absence of windows
on the lower levels of dwellings (the houses of the serfs and
the houses of the animals and their keepers) to the presence
of some windows in the rooms of the higher-ranking monastic
officials (such as the Gardener, the Physician, the Porter,

and the Master of the Outer School), and more elaborate
fenestration on the highest level of dwelling (such as the
House for Distinguished Guests; or in the Outer School,
where supplementary light inlets were a functional necessity).
The small squares in the cubicles for the students of
the Outer School, as we shall subsequently show, must
probably be interpreted as symbols for dormer windows.

V.7.5

BAKING OVENS

On the Plan of St. Gall there are three baking ovens (fig.
382A-C): one in the Monks' Bake and Brew House
(caminus); one in the Bake and Brew House for Distinguished
Guests (fornax); and one in the Bake and Brew
House for Pilgrims and Paupers (fornax). They have diameters,
respectively, of 10 feet, 7½ feet, and 7½ feet.

[ILLUSTRATION]

387. KARKÓW, POLAND. BEHAIM CODEX (1505). BAKEHOUSE WITH OVEN, KETTLE, AND KNEADING TABLES

JAGIELLONIAN UNIVERSITY LIBRARY, CODEX PICTURATUS, fol. 246. [after Winkler, 1941, Pl. 4]

The Behaim Codex is a compilation of privileges, oath formulae, and guild ordinances written in German. It is named from an annotation on
the title page by the clerk and notary of Kraków, Balthasar Behaim (d. 1508), which reads: Anno domini 1505 consummatum. Written
and illuminated in strong and radiant colors by a local artist, doubtlessly of German descent, the manuscript has stylistic roots in a school of
illumination that flourished in Augsburg and Nürnburg, and was strongly influenced by the work of Albrecht Dürer. The representation of the
bakehouse appears on folio 246. Its title is written in bold red letters. (Winklers' color plate suggests that the roof may have been sheeted in
copper.)

[ILLUSTRATION]

388. TÜBINGEN, WURTTEMBURG, GERMANY. UNIVERSITY LIBRARY, CODEX M. D. 2, fol. 29V

This detail shows the planet-children of Saturn in furious, cheerful dispatch of the tasks of baking: milling, kneading dough and setting loaves
to rise, loading the familiar domed oven. Pastries may have been the fate of the caged birds. (Panofsky and Saxl treat the complex iconography
of this illumination in Dürer's "Melencolia I", Leipzig, 1923, 61 [Studien der Bibliothek Warburg]. The codex dates to the late 15th century.

The baking of bread is one of the most ancient of human
arts. Calcined remains of unleavened bread made from
crushed grain were found in Swiss lake dwellings that date
from the early Stone Age.[265] Reference by implication to the
custom of leavening (i.e., admixing to the dough a substance
that produces gases, thus causing the bread to rise)
is made in Genesis, where it is said of Lot that "he made
them a feast, and did bake unleavened bread."[266] One very
early baking method, perhaps the first devised, was that of
placing the dough on a heated flat or convex stone and
covering it with hot ashes.[267] The size and number of loaves
that could be baked in this manner was limited by the shape
of the stone. To bake in quantity required the invention of
the oven, a round or ovoid chamber that held the heat and
allowed it to be distributed over a wider surface. One of the
earliest Central European ovens was excavated in the Stone
Age settlement of Taubried, on Lake Federsee, Germany
(fig. 383).[268] The walls of the baking chamber were made of
daubed wattle. The opening in front was covered with a
removable shutter, probably of wood and cloth.

Between the third millennium B.C. and the middle of the
nineteenth century A.D., neither the shape of the oven nor
the method of baking changed significantly. A circular oven
of baked brick, dating from the beginning of the second
millennium B.C., was found by André Parrot[269] during his
excavation of the Palace of Mari, Mesopotamia, (fig. 386)
in a bathroom of the quarters of the superintendent of
the palace (cf. fig. 372). A Roman oven shaped exactly like
this one is shown on a frieze of the monument of the baker
Eurysaces at Rome, dating from the first century B.C. (fig.
385).[270] On the left, the baker is placing the loaves in the
oven. On the right, four men are kneading dough on a table.

[ILLUSTRATION]

389. A, B, C, D PLAN OF ST. GALL. KITCHEN STOVES AND BREWING RANGES

In each building the stove is indicated as a square, with or without openings for pots; in all probability the "fornax super arcus" of the Monks'
Kitchen was the type in the other kitchens. The brewing ranges show corner openings for vats. The association of baking and brewing facilities
under one roof is traditional and consistent on the Plan; functional interdependence of the two crafts can be shown (see below, pp. 249ff).

Primitive clay ovens of the Taubried type (fig. 383) are
still in use today[271] and were unquestionably common in
medieval times. Figure 384 shows the reconstruction of a
Langobardic oven of this type from the first century A.D.[272]
A handsome illustration in the Behaim Codex in Krakow
(fig. 387)[273] shows the baker placing the loaves in the oven,
his helper shaping them, and a woman throwing some salt
or herbs into the dough rising in a kettle on the floor in
front of the oven. The oven is built into the corner of this
copper-roofed shed. The smoke rises from a round hole in
the top of the oven and passes through a dormer window
in the roof out into the open. Figure 388 shows a baking
scene that occurs among the representations of the planet
children of Saturn in a manuscript in the University Library
at Tübingen.[274] Here again the smoke escapes through circular
openings in the top of the baking chamber. Ovens of
this same design are found in other medieval illuminations.[275]

[ILLUSTRATION]

389. E, F, G PLAN OF ST. GALL. KITCHEN STOVES AND BREWING RANGES

But an oven with a chimney is shown in the December
representation of the Grimani Breviary.[276]

In conformity with the pictorial tradition reviewed above,
we have reconstructed the St. Gall ovens as more or less
circular chambers, the larger one with a smoke flue, the two
smaller ones without (cf. figs. 402 and 394).

V.7.6

KITCHEN STOVES AND KETTLES

Square cooking ranges, resting on arches, with firing chambers
beneath and openings for pots on the surface are
designated for the kitchens on the Plan. Only one of them,
the stove in the Monks' Kitchen, has an explanatory title
(fornax super arcus) (fig. 389A). It is 7½ feet square. The
other kitchen stoves—House for Distinguished Guests
(fig. 389B), Hospice for Pilgrims and Paupers (fig. 389C),
Novitiate and Infirmary (figs. 389D,E)—are about 5 feet
square. We may assume, I think, that their design was the
same as that of the stove in the Monks' Kitchen discussed
earlier.[277] This type of stove was also used in the brew-house
(fig. 389F,G).

It is possible that a cooking area for the serfs and laymen
is to be found in the living room of the House for Horses
and Oxen and Their Keepers, and that the H-shaped
symbol was intended to mean an open fireplace with kettles
suspended on cranes.

The circles around the larger stoves (fig. 389A,F,G) were
undoubtedly meant to indicate tubs or kettles, and these
may have belonged to any of the varieties shown in figures
210, 390, and 400.