1. BRICKS AND TILES

The Romans divided the manufacture of objects in clay into two classes: _opus figlinum_ or fine ware, made from _argilla_ or _creta figularis_; and _opus doliare_, for tiles and common earthenware.[2395] We begin, then, as in the chapter on the Greek uses of clay, with the latter division, including the use of this material in Roman architecture, and primarily in the making of =bricks and tiles=. It must be borne in mind, however, that the structural use of bricks of clay, such as we employ at the present day, was unknown to the Romans; they only used what we should call tiles, and even these were only employed structurally, as a facing to walls and vaults of concrete; no walls were ever built of solid brick, and even in those of seven inches thickness the bricks are built on a core of concrete. Nor were the bricks allowed to appear on the outer face of the building, at least before the second century of the Empire; they were always faced with a coating of marble or stucco. Nevertheless, the general use of bricks or tiles was most extensive, and they were employed as tiles for roofing houses, as bricks for walls and vaults, and even for columns, as slabs for pavements, for furnaces and for covering graves, and in tube form for conveying water or hot air; they are found in temples, theatres, and baths, and are used for cisterns and fountains, and in aqueducts and military fortifications. They were called _lateres_, because, says Isidorus, “they were broad, and made by placing round them four boards.”[2396] The kilns were called _laterariae_, and the makers _laterarii_; to make bricks was _lateres ducere_, _fingere_,[2397] or (with reference to the baking only) _coquere_. The word _later_ seems to be employed indiscriminately for sun-dried (_crudi_) and baked bricks (_coctiles_),[2398] without the qualifying epithet, but _testa_ is also used when burnt brick is intended.[2399] The sun-dried bricks were the earlier and simpler form, used for building walls and cemented together with clay or mud.[2400] Vitruvius in his account of brick-making (ii. 3) only refers to this kind, and apparently never mentions baked bricks except in passing allusions. He describes three kinds, to which he says the Greeks gave the respective names of _genus Lydium_, _pentadoron_, and _tetradoron_ (see Vol. I. p. 95). The two latter are exclusively Greek, but the first-named, 1½ by 1 foot in dimensions, answers to the Roman _tegula sesquipedalis_.[2401] A frequent arrangement, he says, was to employ half-bricks in alternate courses with the ordinary sizes, which served to bind the walls together and present an effective as well as a stable appearance. This information is repeated by Pliny, copying almost word for word.[2402] Among the Romans two dimensions were in general use, as may be inferred from the frequent mention in inscriptions or elsewhere of the _sesquipedales_ and of _bipedales_,[2403] or two-foot bricks, as we shall have occasion to show later. Being very flat and thin in proportion to their size, these bricks rather resemble tiles, as has been already noted; they are generally square, or at least rectangular. But there were also _tegulae bessales_ or bricks measuring two-thirds of a foot square, _i.e._ about 8 inches, and triangular bricks, equilateral in form, with a length varying from 4 to 14 inches. The latter are the kind used in all existing Roman walls of concrete with brick facings. The thickness varies from 1¼ to 2 inches. They are not always made with mechanical accuracy, the edges being rounded and the sides not always parallel. In military works they were often used alternately with flint and stone (see below, p. 337), as we see them in England, at Colchester, Dover, Verulam, and many other places.[2404] At Verulam the tiles are arranged in three horizontal layers at intervals of about 4 feet, with flint and mortar between. They were also used for turning the arches of doorways, and for this purpose _tegulae bipedales_ were cut into pieces, so as only to tail a few inches into the concrete which they cover. Complete squares were introduced at intervals to improve the bonding.[2405] The pillars of the floors of hypocausts were formed of _tegulae bessales_, and sometimes also of two semicircular bricks joined so as to form a circle, varying from 6 to 15 inches in diameter.[2406] Occasionally the upper bricks diminished in size, in order to give greater solidity to the structure. The bricks or tiles forming the upper floors were from 18 to 20 inches square; in some cases, as at Cirencester,[2407] these were flanged tiles (see below). The general size of Roman bricks was, in the case of the _sesquipedales_, 1½ by 1 Roman foot; but variations are found, such as 15 by 14 inches. For the _bipedales_ Palladius recommends 2 feet by 1 foot by 4 inches. The great building at Trier known as the Palace of Constantine is built of burnt bricks, 15 inches square by 1¼ inch thick.[2408] Prof. Middleton notes tiles in Rome of 12, 14, and 18 inches square,[2409] and Marquardt[2410] states that bricks found in France measure 15 by 8 to 10 inches; others (the _bessales_) 8 by 8 by 3 inches. A complete circular brick, measuring 7½ inches across by 3¼ inches thick, and impressed with the stamp of the eleventh legion, was found at Dolae near Gardun, and is now in the museum at Spalato.[2411] Vitruvius[2412] gives elaborate instructions about the preparation of the clay for sun-dried bricks, and counsels in the first place a careful choice of earth, avoiding that which was sandy or stony or full of loose flints, which made the bricks too heavy, and so liable to split and fall out when affected by rain; it also prevented the straw from binding properly. Clay which was either whitish or decidedly red (from a prevalence of ochre) was preferred, and that combined with coarse sand (_sabulo masculus_) made light tiles, easily set. The process of manufacture was a very simple one. The clay was first carefully cleaned of foreign bodies, and then moistened with water and kneaded with straw. It was then moulded by hand or in a mould or frame of four boards, and perhaps also pressed with the foot.[2413] The bricks were then dried in the sun and turned as required, the usual process also adopted in the modern brickfield. Some bricks actually bear the marks of the feet of animals and birds which had passed over them while the clay was soft, and there is one in the Shrewsbury Museum with the imprint of a goat’s feet. Others at York and Wiesbaden show the nails of a boy’s shoes.[2414] These impressions of feet (where human) may also be referred to the practice of using the feet to knead the bricks. The bricks were then ready for use, but were kept for two years before being employed, otherwise they were liable to contract, which caused the stucco to break off and the walls to collapse. At Utica, Vitruvius tells us, they had to be kept five years, and then could only be used if passed by a magistrate. Altogether, much care was taken in their preparation, and it was generally considered that spring and autumn were the most favourable times for making them, probably because they dried more slowly and were less liable to crack during the operation. In summer the hot sun baked the outer surface too fast, and this appeared dry while the interior was still moist, so that when the inside dried the outside contracted and split.[2415] It was also, of course, advisable to avoid seasons of rain and frost. But the bricks could not be properly tested until they had undergone some exposure to the weather, and for this reason Vitruvius recommends the employment of old roof-tiles where possible in building walls.[2416] For baked bricks the processes must have been much the same, with, of course, the addition of the baking in the furnace. Existing Roman bricks are nearly always of well-tempered clay and well baked; but the clay exhibits a great variety of colour—red, yellow, and brown. The paste is remarkably hard, breaking with an almost vitreous fracture, and sometimes shows fragments of red brick (_pozzolana_) ground up with it to bind it together, and prevent warping. This may be seen in the Flavian Palace on the Palatine, and in an archway in the Aurelian Wall near the Porta Latina. As an instance of varieties of brick found in the same building, Nero’s Aurea Domus may be cited.[2417] The durability of Roman tiles is ascribed to their careful preparation and seasoning, which give them a much longer life than modern tiles; hence they were frequently used up again in early mediaeval buildings and in Romanesque churches in England, as at St. Albans, St. Mary-in-Castro, Dover, and St. Botolph’s and Holy Trinity, Colchester.[2418] During the period of the Republic private houses and public buildings alike were built of unburnt brick in Rome, as we learn from the words of Dio Cassius,[2419] Varro,[2420] and Cicero[2421]; Varro speaks of _domus latericiae_, and Cicero of “the brick (_latere_) and concrete of which the city is constructed.” After the Republican period this material was still employed outside Rome with burnt-brick cornices,[2422] but even this was exceptional. Pliny mentions walls of sun-dried bricks at Arretium and Mevania.[2423] Henceforth, then, burnt brick was employed more and more as Rome grew more populous.[2424] In Vitruvius’ time (the beginning of our era) the materials used for building were stone for substructures, burnt brick (_structura testacea_) for the outer walls, concrete for the party-walls, and wood for the roofs and floors. He explains the cessation of the use of unburnt brick as due to the legal regulations of his time, which prohibited party-walls of more than 1½ foot in thickness, and unburnt bricks could only support one story above them in that size.[2425] Baths, either public or private, walls and military fortifications, were built of bricks, the latter being thus better able to resist attacks than if they were of stone. Temples, palaces, amphitheatres, the magnificent aqueducts and the cisterns with which they communicated, were also usually of this material. Of these, numerous remains exist in Rome and other places, such as Cumae and Pozzuoli. The aqueduct made by Nero from the Anio to Mons Caelius is of brick, that of Trajan partly so; the _aqua Alexandrina_ of Severus Alexander (A.D. 229) and that existing at Metz are wholly of brick, and so are the _castella_ or reservoirs made by Agrippa when he constructed the Julian conduit over the Marcian and Tepulan.[2426] It is true that Augustus boasted that he had found Rome of brick and left it marble[2427]; but it must be remembered, firstly, that Suetonius uses the term _latericiam_, which may denote unburnt brick; secondly, that the phrase is probably to be limited to public buildings and monuments, in which there was an increased use of marble for pillars and roofs. For walls brick and concrete continued to be used, as in private buildings, with a covering of stucco in place of marble incrustation. In the first century of the Empire brick-making was brought to perfection, and its use became universal for private and public buildings alike; the mortar of the period is also of remarkable excellence. The Romans introduced brick-making wherever they went; and even their legions when on foreign service used it for military purposes. But of pure brick architecture, as we see it, for instance, in the Byzantine churches of Northern Italy, there was no question until comparatively late times. It was always covered over with marble or stucco until the second century of the Empire. Examples of sepulchral buildings wholly in brick, of the time of Hadrian, may be seen in the tomb before the Porta San Sebastiano at Rome, known as the temple of Deus Rediculus. This has Corinthian pilasters with a rich entablature, red bricks being used for architectural members, yellow for the walls; the capitals are formed of layers of bricks. Of Hadrian’s time are also the guard-house of the seventh cohort of Vigiles across the Tiber, of which a small part remains, and the _amphitheatrum castrense_ on the walls of Aurelian.[2428] One of the most remarkable instances of Roman brick construction is the Pile Cinq-Mars, as it is called, a tower still standing on the right bank of the Loire, near Tours. It is about 95 feet high and 13 feet square, expanding at the base, being built of tiles to a depth of 3 feet each side, with a body of concrete; the tiles are set in mortar composed of chalk, sand, and pounded tiles. On one side there are eleven rectangular panels with tile-work of various patterns, like those on the flue-tiles (see p. 348), and as also seen on the Roman wall at Cologne; the patterns include squares, triangles, and rosettes. The history and purpose of this building are quite unknown.[2429] At Pompeii bricks are used only for corners of buildings or doorposts, and sometimes for columns, as in the Basilica and the house of the Labyrinth.[2430] There are also late examples of brick columns with capitals in tiers of bricks as in the tomb mentioned above. Brick walls are not found, but bricks occur as facing for rubble-work. These are less than an inch thick, triangular in form, with the hypotenuse (about 6 inches long) showing in the face of the wall. Sometimes fragments of roof-tiles are used (cf. p. 334). The earlier bricks contain sea-sand, and have a granular surface; the later are smooth and even in appearance. Later, what is known as _opus mixtum_ (see below) is used, as in the entrance of the Herculaneum gate; this implies courses of stone and brick alternating,[2431] which, as we have seen, was common in military works, as in the Roman walls in Britain. In this country, owing to the absence of good material for concrete, the use of stones or brick throughout for building was general from the first; hence, too, the bricks are always flat and rectangular in form (_bipedales_).[2432] The arrangement of triangular bricks (made by dividing a medium-sized brick into four before baking), laid flat in regular horizontal courses, is characteristic of the earliest examples of Roman methods. It is found in the Rostra (44 B.C.) and in the Regia (35 B.C.), the earliest existing examples.[2433] The back wall of the Rostra is of concrete faced with triangular bricks 1½ inch thick, the sides 10 inches long. The same arrangement may be seen in the Pantheon, in the Thermae of Diocletian, and in some of the aqueducts (see below). The brickwork in the Pantheon was formerly thought to belong to the building of Agrippa in 27 B.C., but has been now shown to belong to the second century.[2434] At Ostia, in the temple of Honos and Virtus, the walls are built of triangular bricks or with red and yellow bricks with moulded cornices. [Illustration: Section of Angle. FIG. 187. CONCRETE WALL, Faced with (A) _opus incertum_, (B) _opus reticulatum_. C shows the horizontal section, similar in both. ] [Illustration: FIG. 188. CONCRETE WALL (VERTICAL SECTION), FACED WITH BRICK.] About the year 80 B.C. the method known as _opus reticulatum_ was introduced, in which the bricks presented square faces (about 4 inches each way), and were arranged diagonally to form a network pattern (Fig. 187). At Pompeii the _opus reticulatum_ dates from the time of Augustus; it is laid on concrete, and the bricks are small four-sided pyramids with bases 3 to 4 inches square.[2435] This method lasted down to about A.D. 130 in Italy. It should, however, be noted that it was commoner in stone than in brick, the latter material not having come into general use for building at the time when it was employed.[2436] But even when tufa was used for the reticulated work, bricks or tiles were used for quoins at the angles, and for bonding courses through the walls, as well as for arches and vaults (Fig. 188). This combination of _opus reticulatum_ and brickwork is well illustrated in the palace of Caligula.[2437] In the case of vaults, indeed, the use of brick seems to have been general, as in the baths of Caracalla, and many other buildings (cf. Fig. 189). Vitruvius[2438] advises the use of _tegulae bipedales_ to protect the wooden joists over the vaults from being rotted by the steam from the hot bathrooms; they were to be placed over the whole under-surface of the concrete vault, supported on iron girders, which were suspended from the concrete by iron clamps or pins. Over the whole was laid a coating of cement (_opus tectorium_) in which pounded pottery was the chief constituent, and this was stuccoed.[2439] [Illustration: FIG. 189. CONCRETE ARCH; HALF WITH BRICK FACING REMOVED.] The _opus mixtum_ (the term is not classical) prevailed regularly under the later Empire, from the fourth to the sixth century; the earliest example which can be dated is the circus of Maxentius. It is also used in work of the time of Theodoric.[2440] The method of construction is shown in Fig. 190. The reason for the limited use of brick in Rome may have been the scarcity of wood for fuel for the kilns. But in any case the pointed backs of the bricks made a good bonding with concrete, and presented a large surface with a comparatively small amount of clay. The secret of the wonderful durability of Roman buildings is that each wall was one solid coherent mass, owing to the excellence of the concrete. In the Pantheon the concrete of the dome is nearly 20 feet thick, the brick facing only about 5 inches. The character of the brick facing often indicates the date of a wall, the bricks in early work being thick and the joints thin; later, the reverse is the case. But caution must be exercised in dating on this principle, owing to the great variety of methods employed during the same reign, and even in the same building.[2441] [Illustration: From _Blümner_. FIG. 190. DIAGRAM SHOWING CONSTRUCTION OF WALL OF OPUS MIXTUM. ] * * * * * The word for a tile, _tegula_, is derived from _tegere_, to cover, or, as Isidorus says, they are so called _quod aedes tegant_[2442]; the curved roof-tiles were known as _imbrices_ because they received rain-showers (_imbres_). The maker of roof-tiles was known as _tegularius_[2443] or _figulus ab imbricibus_.[2444] _Tegulae_ or flat roof-tiles were usually made with vertical flanges (2½ inches high) down the sides, and these flanges, which fitted into one another longitudinally, when placed side by side served to hold the covering-tiles placed over them. There were also roof-tiles known as _tegulae deliciares_[2445] and _colliciares_, which formed the arrangement underneath the surface of the roof by means of which the water was collected from the _tegulae_ and carried off in the front through spouts in the form of lions’ heads.[2446] Besides the various rectangular forms we find triangular tiles used, either equilateral or right-angled; semicircular or curved tiles, used for circular walls, ovens, tombs, and cornices, or other parts of buildings; cylindrical tiles (_tubuli fictiles_),[2447] which were used for drains and conduits; and, finally, the rectangular hollow flue-tiles, employed for hot air in hypocausts.[2448] Another form was the _tegula mammata_, a plain square tile with four knobs or breast-like projections (_mammae_), which was often used in party-walls with the object of keeping out damp.[2449] The tiles were inserted by the points of the projections into the concrete, thus leaving a space between in which the warm air could circulate freely. Existing examples of tiles are composed of a compact dense clay, less fine than that of the bricks, and of a pale salmon or light straw colour when baked. They were probably made in moulds—but these may only have been a couple of boards placed together—and after being dried in the sun were baked in kilns. The flanged tiles were, of course, produced by turning up the edges before drying. Besides the arrangement described above, it is probable that roofs were sometimes tiled in the manner prevalent in the present day, with flat or curved tiles overlapping like scales; and for this purpose the tiles seem to have been pierced with holes at one corner, and so attached to one another. The same method obtained in the Roman villas in Britain, except that Stonesfield slate was used in place of tiles. An inscription found at Niederbrunnen in Germany speaks of _attegia tegulicia_, or huts roofed with tiles, erected in honour of Mercury.[2450] Tiles with turned-up edges or flanged tiles were principally employed, as has been indicated, for roofing; but some were also placed in walls where required, especially where a space was required for the passage of air.[2451] They were also employed for the floors of bath-rooms, in which case they were laid on the _pilae_ of the hypocaust in an inverted position, and the cement flooring was laid upon them. The flanges are generally about 2¼ inches higher than the lower surface of the tile; they are bevelled on the inner side in order to diminish the diameter of the _imbrex_, but have no holes for nailing to the rafters. The ends of the sides were cut away in order that the lower edge of one tile might rest on the upper edge of the one adjoining. Those found in France are said to be distinguished by the sand and stones found in their composition.[2452] There are flange tiles of red and yellow clay from the Roman Thermae at Saintes in the Museum of Sèvres, and others from ancient potteries at Milhac de Nontron, as well as tiles of red clay from Palmyra.[2453] In the military _castra_ in England flange tiles of a red or yellow colour have been found, the latter with fragments of red tiles mixed in the clay. They are also often found in the ruins of villas. A flange tile from Boxmoor, Herts, now in the British Museum, measures 15½ by 12 inches, the flange being 2¼ inches high; and it will be seen that these dimensions correspond roughly with the _tegulae bipedales_. Flanged tiles with holes in them appear to have been used at Pompeii for lighting passages, the flanges serving to keep out rain.[2454] The _imbrices_ or covering-tiles which held the flat tiles together, thus rendering the roof compact, were quite plain, with the exception of the end ones over the gutters. These were in the form of antefixal ornaments like the Greek examples (Vol. I. p. 98), an upright semi-oval termination ornamented with a relief or painted pattern, with an arched support at the back. Many examples exist at Pompeii (see below), Ostia,[2455] and elsewhere; but artistically they are far inferior to the Greek examples, and of simpler design. Most of them have a simple palmette or acanthus pattern in low relief, but on or below this an ideal head or the head of a deity is sometimes added, such as Zeus Ammon, Medusa, a Bacchic head, or a mask, or even a figure of Victory. Of the last-named there is a good specimen in the British Museum (D 690 = Fig. 191); she carries a trophy from the battle of Actium, and stands on a globe from which spring two Capricorns (the symbol of Augustus).[2456] [Illustration: FIG. 191. ROMAN TERRACOTTA ANTEFIX: VICTORY WITH TROPHY (BRITISH MUSEUM). ] No better example of the various uses of ornamental tiles in architecture can be selected than the remains found at Pompeii, which are exceedingly numerous. Terracotta seems to have been used here especially for such parts of the decoration as were exposed to wet, as well-mouths, gutters, and antefixal tiles.[2457] A characteristic feature of the decoration of Pompeian houses was the trough-like gutter which surrounded and formed an ornamental cornice to the _compluvium_ or open skylight of the atrium and peristyle, through and from which the rain-water was collected in the _impluvium_ or tank sunk in the ground below. These were adorned with spouts in the form of animals’ heads or foreparts,[2458] usually lions and dogs, with borders of palmettes between; the gutter behind was virtually a long tank of square section. Antefixes and gutter-cornices, where they occur, must always be regarded as serving ornamental rather than necessary purposes. All early work in terracotta at Pompeii is of coarse clay, but good execution; later, the reverse is the case. The only public building in which many remains of terracotta tiles and cornices have been preserved is the temple of Isis; but the Basilica may also have had terracotta decoration. Many fragments also remain from private houses, some actually _in situ_, having been neglected by early explorers as unimportant. In the house of Sallust a kymation cornice from one of the garden courts has scenic masks forming the spouts; this is not earlier than the rebuilding of the house A.D. 63. There is also much terracotta work in the house of the Faun.[2459] Comic masks were used both as spouts and as antefixes, the exaggerated mouth of the mask serving admirably for the former purpose.[2460] These date from the reigns of Nero and Vespasian, and all seem to be from the same fabric, although there is considerable variety in the types; the use of masks for these purposes is not earlier than Nero’s reign (cf. the house of Sallust, above). Besides the ornaments above mentioned the patterns on the cornices include palmettes and floral scrolls, dolphins and Gryphons. The roof-tiles were of the usual kinds, flat oblong _tegulae_ with flanges, measuring 24 by 19 by 20 inches,[2461] with semi-cylindrical _imbrices_. They were laid in lines parallel to the long ridges of the roofs, so that the water converged into the gutter-tiles at the angles, whence it fell into the impluvium. These gutters, however, were not confined to the angles of the openings, but were sometimes ranged along the whole length of the sides, as we have seen; those at the angles only seem to be earlier in date. They are not found on the exteriors of buildings. The front of the gutter was usually in the form of a vertical kymation moulding, but was sometimes simply chamfered. Antefixal ornaments terminating the covering or ridge-tiles are not invariable, but are found at different periods. The earliest examples are in the form of palmettes, but the later exhibit a great variety[2462]: comic masks, a head in low relief on a palmette, or a head surmounted by a palmette. Of the latter class thirty-eight were found in 1861. In the Augustan period ideal heads of gods and demi-gods are sometimes found.[2463] Von Rohden, in summing up (p. 14), is of the opinion that terracotta roof-decoration at Pompeii was _comparatively_ rare. In the whole record of excavations only twenty-three water-spouts are mentioned, though it is probable that many were never registered. In scarcely more than twelve private houses have as many pieces been found as would suffice for the whole of the atrium and peristyle roofs, and nearly all of these are of late date. The discovery of isolated pieces in a house seems to show that they were used up again in the restorations after the earthquake of A.D. 63. There are also some good examples of roof-tiles among those which have been found at Ostia, both in baths and private houses; some of the latter came from a house of which the brickwork bore inscriptions with the names of consuls of Hadrian’s reign. The arrangement of the roof-tiles is that described on p. 341; the antefixal ornaments are usually in the form of palmettes or acanthus leaves, with maeander below; but heads of deities, such as Venus and Neptune,[2464] or of Medusa, and tragic masks were also found. Two exceptional examples had groups in relief of Neptune drawn over the sea by hippocamps, and of the statue of Cybele in the ship drawn by the Vestal Virgin Claudia.[2465] * * * * * Tiles of the size known as _bipedales_ are also used for lining the walls of rooms. They are found in Roman villas in Britain, and are ornamented on one side with various incised patterns, made with a tool in the wet clay. On some found at Ridgewell in Essex the decoration consists of lozenges, rosettes, and other ornaments,[2466] like those on the Pile Cinq-Mars already described; they are often found covered with the stucco with which the walls were plastered. At Pompeii, Orvieto, and elsewhere the stucco-painted walls were constructed with _tegulae mammatae_ placed edgewise, and connected with the main walls by leaden cramps, the brick lining being thus detached from the walls by a narrow interval which served as an air-cavity.[2467] This was a frequent proceeding, and was also contrived with flanged tiles; it corresponds with the system prescribed by Vitruvius for keeping damp from the painted walls of rooms.[2468] It was also largely employed in baths and bathrooms, the object being both to keep the walls dry and to allow hot air to circulate from the hypocausts and warm the rooms. In the cold climate of Britain the Romans found this a universal necessity, and instances may be observed in many of their villas; but, as far as can be observed, the general method of warming was by an extensive system of pipes under the floors rather than up the walls.[2469] These tiles are pierced with holes, by means of which they were attached to the walls by plugs or nails of lead. In the _castrum_ at Jublains a chamber is yet partly standing with one of its sides coated with tiles of this kind.[2470] [Illustration: From _Middleton_. FIG. 192. METHOD OF HEATING THE BATHS IN THE THERMAE OF CARACALLA. ] A A Concrete wall, faced with brick, shown in vertical and horizontal sections. B Lower part of wall, with no brick facing. C C _Suspensura_, or upper floor of Hypocaust, supported by pillars. D D Another floor, with support only at edges. E E Marble flooring. F F Marble plinth and wall lining. G G Under floor of Hypocaust, paved with large tiles. H H Horizontal and vertical sections of flue-tiles lining wall of Calidarium. _a a_ Iron hold-fasts. J J Socket-jointed flue-pipe of Tepidarium. K Rain-water pipe (in horizontal section). L L Vaults of crypt, made of pumice- stone concrete. More commonly, however, a peculiar kind of tile was used for warming the hot rooms (_sudationes_) of baths, and in villas when required. They were hollow parallelopipeds, known as _tubi_, with a hole in the side for the escape of the air which traversed them, the usual dimensions being about 16 by 6 by 5 inches.[2471] Seneca speaks of pipes inserted in walls, which allowed the warmth to circulate and warm both the upper and lower stories equally[2472]; and the younger Pliny mentions the air-holes (_fenestrae_) in the pipes which warmed his bedroom, by means of which the temperature could be regulated at pleasure.[2473] Sometimes, as in the baths of Caracalla and the house of the Vestals, the whole side of a wall was composed of flue-tiles covered with cement,[2474] which was made to adhere by scoring the sides with wavy or diagonal lines, as in the flat tiles described above, and as is often done in modern building. The whole system of heating, which may be seen in the baths of Caracalla, is very instructive (Fig. 192): the walls were of concrete with brick facing, through which a system of flues of socket-jointed tiles passes upwards from the hypocaust below, effectually warming every part.[2475] [Illustration: FIG. 193. FLUE-TILE WITH ORNAMENTAL PATTERNS.] The hollow tiles often assume a more ornamental appearance (as in Fig. 193), the patterns scratched on them taking the form of lozenges and diapers, chevrons, chequers, and rosettes, as may be seen in a Roman villa at Hartlip in Kent, where other tiles are simply scored with squares.[2476] This villa is remarkable for the extensive use of tiles throughout; even the staircases are constructed with them. Others found in Essex and Surrey have dogs, stags, and initial letters among foliage; one found in London had among the wavy lines of pattern the letters Px Tx[2477]; and another, from Plaxtol in Kent, the local maker’s name, CABRIABANTI.[2478] These hollow tiles, which are generally of the same clay as the roof-tiles, were also occasionally used as pillars of hypocausts,[2479] but for this purpose columns of _tegulae bessales_ were more usual, as Vitruvius implies.[2480] Many examples may be seen in the Roman villas of Britain, as at Cirencester, Chedworth, Lympne, and Wroxeter. In a villa found at Carisbrooke, Isle of Wight, the whole bath was constructed of tiles, the floor supported by _pilae_ of the same.[2481] At Bath the hollow tiles are actually used as _voussoirs_ for arches and vaults.[2482] Through these chimneys—for this is what they practically were—the hot air circulated and gave an imperfect warmth to the rooms, the heat radiating from the walls or penetrating through the air-holes.[2483] The pipes standing close to one another virtually made up the wall; but the exact method by which the warming was accomplished, without great inconvenience to the occupiers of the rooms, is not quite clear. It is not difficult to imagine that the tiles would have warmed rooms merely by the introduction of hot air circulating through them, even though covered with stucco. On the other hand, the apertures for admitting the air into the rooms, if of any size, must also have admitted smoke from the hypocausts, and interfered with the ventilation. It may be that they were not made for this purpose at all, but only for fastening the pipes together or to the walls. Another difficulty is the method in which the flues made their exit into the open air. It has been suggested, partly on the analogy of a mosaic found in Algeria, that they ended above in an arrangement like a chimney-stack. There is, moreover, a terracotta roof-tile in the Museo delle Terme at Rome with a circular pipe, 8 inches in diameter, projecting from its upper surface.[2484] Terracotta pipes, or _tubuli_, of cylindrical form, were sometimes employed by the Romans for conveying or distributing water, but the more usual material for this purpose, especially for drinking-water, was lead; the latter were called _fistulae_.[2485] The Venafrum inscription, an edict of the Emperor relating to the water-supply of the town, mentions _canales_, _fistulae_, and _tubi_.[2486] Vitruvius calls the _canales structiles_, implying that they were of masonry.[2487] Pliny speaks of _tubi fictiles_ used for conduits from fountains,[2488] and Vitruvius recommends the use of terracotta pipes (_tubuli fictiles_) in aqueducts.[2489] Examples of clay piping are preserved in the Museo delle Terme at Rome. At Marzabotto, near Bologna, terracotta pipes were used for carrying off the water from the roof of a house, by means of a straight tube through the wall fitting into another which curved upwards inside.[2490] These date from the fifth century B.C. Other examples have been found in Rome and Italy,[2491] and specimens found on the Rhine were 21½ inches long, of which ¾ inch was inserted into the adjoining pipe, and 3½ to 4½ inches in diameter. Terracotta was also used for cisterns, as at Taormina,[2492] and for aqueducts; but Lanciani has pointed out that its use in these ways was confined to irrigating purposes. The Campagna of Rome was formerly extensively drained with these tiles, and owed to that circumstance much of its ancient healthfulness. Of the use of tiles in pavements there is frequent mention in Roman writers.[2493] For this purpose complete tiles were seldom used, at any rate in Italy; but in Britain it was not at all uncommon, as in the villa at Hartlip already mentioned. On the other hand, hypocausts were regularly paved with tiles, as in the Baths of Caracalla (Fig. 192 above),[2494] and in an example found at Cirencester, where the tiles are flanged.[2495] But in another form tiles played a considerable part in Roman methods of paving. Pliny and other writers[2496] speak of _pavimentum testaceum_ or _opus signinum_ as the usual pavement for rooms, especially those liable to damp, such as kitchens and outbuildings, or for baths and cisterns. This was made of a layer of fragments of tiles stamped and pounded into a firm solid mass, combined with mortar. It corresponds to the _nucleus ex testis tunsis_ of Vitruvius, which (to a depth of six inches) was laid on the _rudus_ or coarser concrete. On this was laid the flooring, consisting either of tiles or marble slabs, or more generally of mosaic. The Baths of Caracalla again afford a good illustration of the process.[2497] In the mosaics too fragments of clay were often used, especially for producing red or black colour.[2498] Vitruvius and other writers allude to this practice,[2499] and the former also speaks of _testacea spicata_, a kind of false mosaic made with small bricks about 4 inches by 1 inch, set on edge to form a herring-bone pattern. In the Guildhall Museum is part of a tesselated pavement of concrete, faced with small bricks about an inch square. One of the most interesting uses of tiles by the Romans is in connection with their tombs. Not only are they used in the construction of the more magnificent edifices (cf. p. 336), but they were also often employed (as in Greece) for the humbler graves. For the latter, three, or sometimes six, _tegulae bipedales_ were set up in the form of a prism, one forming the floor, the other two the gabled covering which protected the body from the superincumbent earth. Within this were laid the _ollae_ or sepulchral urns which held the ashes of the dead, and other vases. A tomb found at Litlington in Cambridgeshire was covered with a large flanged tile, which protected the pottery buried underneath[2500]; and at Eastlow Hill in Suffolk a tomb was found roofed with twelve rows of flanged tiles, each side in rows of four.[2501] In some of the tombs of Greece belonging to the Roman period semi-cylindrical tiles were used for this purpose. In the provinces the tiles often have impressed upon them in large letters the names of the legions which garrisoned the various cities. The tiles of Roman tombs at York are inscribed with the names of the sixth and ninth legions which were quartered there: as LEG · VI · VICT · P · F, _legio sexta victrix pia fidelis_; LEG · IX · HISP (or VICT), _legio nona Hispana_ (or _victrix_).[2502] At Caerleon (Isca Silurum) the bricks bear the name of the second or Augustan legion: LEG · II · AVG.[2503] The stations of the twentieth legion may also be traced at Chester in this manner; the tiles are inscribed LEG · XX · V · V.[2504] They were placed at the foot of the tomb like tombstones, in order to indicate who was buried beneath, the inscriptions being written across the breadth of the tile. They are of very different dates, some of those in Britain being apparently as late as the introduction of Christianity. * * * * * The extent to which bricks and tiles were used in Roman buildings under the Empire may be gauged by the number of those with inscriptions which remain; a whole section of the Latin _Corpus_ (see below) is devoted to those found in Rome alone, numbering some two thousand. Many of them have been removed to the museums from the principal edifices, such as the Pantheon, the Coliseum, the Circus Maximus, the Baths of Titus and Caracalla, the Basilica of Constantine, and the Praetorian Camp. Other inscriptions have been found on tiles removed from such buildings and used to repair the roofs of churches in Rome. Such places as Bologna, Cortona, Tibur, and Ostia have also produced numerous inscribed tiles of this class. The use of such stamps was to guarantee the quality of the clay. To the topographer, as will be seen, these stamps are often of great value; and had the custom of placing on them the names of the buildings for which they were intended been less rare, they might often have afforded valuable evidence as to doubtful sites. Besides their topographical value, the tiles also help to settle the succession of consuls, and throw great light on the economy of the Roman farms and the possessions of the great landed proprietors. The uninterrupted series, extending from the times of the Caesars to the age of Septimius Severus, of names of proprietors, potters, and estates, tells much of the internal condition of Italy, and of one of the sources of revenue to the Roman nobility.[2505] The stamps found on bricks and tiles are of four kinds—rectangular, semicircular, circular, and crescent-shaped. The inscriptions are in raised letters in all cases, but instances are also known of incised inscriptions, written without frames across the tile. After the time of Diocletian the only forms found are square, circular, and octagonal; the square stamps always have straight inscriptions. On the circular stamps the inscriptions are placed in a circle, in one or two lines, and the beginning is determined by a small cut-out circle at the edge of the stamp, thus 20[22]orbiculus known as the _orbiculus_; apart from this its object is uncertain. In later stamps the inscription often reads backwards, or certain letters are reversed. The letters were cut straight in a mould and lie in the plane of the surface, being of rectangular section, not wedge-shaped, as in inscriptions on marble. During the Republican period and the first century of the Empire a plain “block” type is used; then the letters become smaller and more elegant, with bars at the ends of the _hastae_, as 20[54]E, M etc. Finally they show a tendency about A.D. 200 to become broader and shorter: 20[76]E, M, S At and after the time of Diocletian the forms become very varied. Punctuation in the best period takes the form of a 15[17]triangle afterwards the mark becomes vague in form. Ligatured letters are rarely found after the time of Diocletian, but are common in the best period; sometimes more than two are combined.[2506] The stamps with which the letters were made were usually of wood or bronze, but have not been preserved. In the centre of the stamp it was customary to place an emblem or device of some kind, perhaps in view of a law which obliged brick and tile makers to affix distinctive marks or emblems on their bricks; but the devices are not peculiar to individual workshops, and some potteries, such as the Terentian (see below), used several. They may be compared with the countermarks or small adjuncts on the coins of the Republic, and the seals and stamps on the wine-amphorae of Thasos (Vol. I. p. 158). Figures of gods, such as Mars, Cupid, and Victory, animals, and even groups of figures, occur, and after the third century Christian emblems are often found. It is most probable that they were merely ornamental and without significance, except in certain cases of canting or punning allusions. Thus M. Rutilius Lupus has a wolf; Flavius Aper a boar; Aquilia an eagle; C. Julius Stephanus a wreath; and Aelius Asclepiades a serpent, with reference to the god Asklepios.[2507] [Illustration: FIG. 194. STAMPED TILE (BRITISH MUSEUM).] The most complete stamps have the date of the emperor or the consulship, the name of the estates (_praedia_) which supplied the clay, that of the pottery where it was baked (_figlinae_ or _officina_), and that of the potter who prepared it; sometimes even of the slave who moulded the tile, and even its very dimensions. Two typical examples may be given from the British Museum collection,[2508] of which the first (Fig. 194) is said to have been found in the Catacombs at Rome. It has in the centre of the stamp a figure of Victory, round which is the inscription in two lines, beginning with the outer band: OPVS DOL(_iare_) DE FIGVL(_inis_) PVBLINIANIS (_ex_) PREDIS AEMILIAES SEVERAES “Pottery[2509] from the Publinian works, (the clay) from the estate of Aemilia Severa.” The other has no device, but the last word of the inscription is in the centre: IMP ANTONINO II E(_t_) BALBINO COS D P Q S P D O ARABI SER(_vi_) “The Emperor Antoninus for the second time and Balbinus consuls; from the estates (_de praediis_) of Q. Servilius Pudens, pottery (_doliare opus_) from the hand of the slave Arabus.” The earlier stamps exhibit more method and precision; the later betray comparative carelessness. In the latter the name of the emperor sometimes occurs alone, and unusual expressions are introduced. Contractions are invariable at all periods, and even the consuls are sometimes only mentioned by initials; but by comparison of examples it is possible to place them in the right order. Those found in Rome cover the period from the reign of Trajan to that of Theodoric (A.D. 500), but in other parts of Italy they are found dating as early as 50 B.C. We are told that Theodoric, when he repaired the walls of Rome, made a present of twenty-five thousand tiles for the purpose,[2510] and on the tiles bearing his name he is styled “The good and glorious king,” with the additional exclamation, “Happy is Rome!”[2511] The estates on which the clay for the tiles was produced are called _possessiones_; _privata_ (private property); _rationes_ (shares); _insulae_ (blocks); or more generally, _praedia_. The latter word, indeed, is almost invariably used down to the third century, the others being more characteristic of the time of Diocletian. The _praedia_ not only provided the clay, but in some cases also contained the potteries. On some tiles _fundus_, which means a country farm, is found. The proprietors of these estates were imperial personages, persons of consular dignity or equestrian rank, and sometimes imperial freedmen. Many tiles give merely the name of the imperial estates, without mentioning the reigning emperor; in the later ones, as in the Basilica of Constantine, it is usual to find the expression OFF · AVGG ET CAES NN, _Officina Augustorum (duorum) et Caesarum (duorum) nostrorum_.[2512] Several names of the Antonines occur; also Annius Verus and his wife Domitia Lucilla, the parents of M. Aurelius. Septimius Severus owned many _praedia_ which supplied bricks for his palace on the Palatine.[2513] The Empress Plotina was evidently a large landed proprietor, and we also find the names of Aelius Caesar (Hadrian’s adopted heir), M. Aurelius, Faustina II., and Julia Procula. Among the names of inferior proprietors, unknown to fame, occur Q. Servilius Pudens, T. Statilius Severus, and L. Aemilius Julianus, priest of the sun and moon.[2514] Such names as Q. Agathyrsus, Rutilius Successus, and Sulpicius Servandus seem to denote imperial freedmen; the first-named styles himself AVG · LIB.[2515] A remarkable fact in connection with these inscriptions is the prevalence of feminine names, the quantity of tiles on which these are found being enormous. The causes are various,—partly the renunciation by emperors of their private fortunes in favour of their female relations; partly the proscriptions which, from the failure of male heirs, caused estates to devolve upon women; partly the gradual extinction of great families. The important position held by freedmen under the Empire is well known to the student of Roman history. The potteries of the tile-makers were of two kinds—_figlinae_ and _officinae_; but the former seems to be a wider and inclusive term—that is to say, that one _figlina_ included several _officinae_ or workshops. In the inscriptions, ex figlinis is usually followed by the name of the owner, _ex officinis_ by the name of the potter (_officinator_). The former expression is by far the commoner, and the latter (OF or OFFIC) is more usually found on lamps and vases, although after the third century it is invariable on the tiles. The _figlinae_ are always mentioned in a subordinate manner to the _praedia_, when both are mentioned, as is usually the case. The potteries were mostly outside the city, even at some distance. Localities are not often mentioned, but we have the Salarian potteries on the _Via Salaria_,[2516] and also mention of the _Via Nomentana_,[2517] and such expressions as _Ad Aureliam_, _Ad Mercurium felicem_, or _Ad viam triumphalem_. Stamps found in the walls along the Appian and Latin ways show that potteries existed in the direction of the Alban and Tusculan hills, and in other parts of Latium, as at Praeneste and Ostia. On the north side they extended as far as Narnia and Ocriculum[2518] on the Tiber. They are also found in Etruria and Campania. Tiles from Latium were exported to Liguria, the Adriatic, Sardinia, Africa, Gaul, and Spain. Usually a descriptive epithet is associated with the word _figlinae_, either of a geographical or personal character. Examples of the former are Macedonianae, Rhodianae, and Oceanae. The latter give either the name of an emperor, as Neronianae, Domitianae; or a Gentile or family name, as Favorianae,[2519] Furianae, Publinianae, Terentianae, or Voconianae. One of the names which occurs most frequently is that of L. Brutidius Augustalis, a freedman; others are stamped EX FIGLINIS PRIMIGENI SERVI DNI NOSTRI IMP—“From the potteries of Primigenius, slave of our lord the Emperor.” Imperial slaves owned many potteries, and others were owned by the emperors or other wealthy proprietors, and administered by freedmen or slaves. The _officinae_ served to distinguish the functions of the different _figlinae_. Thus the establishment of M. Publicius Januarius, a freedman, is styled _doliariae officinae_; or they are distinguished by separate names, as Claudianae, Domitianae, and so on. The tiles from the potteries of Asinius Pollio bear the name of C. Cosconius as maker, as do those of Julia Procula’s potteries, being further distinguished as _doliares_, _bipedales_, and _sesquipedales_.[2520] It would appear that the potteries of private proprietors were under the direction of freedmen, while those of the imperial estates were chiefly managed by slaves, from whose labours large revenues were obtained. There were many private potteries in Gaul and Germany.[2521] In the neighbourhood of Saarbrück many tiles have been found with the maker’s name, L. Valerius Labeius. Others with private names have been found at Trier, one with the stamp of the _colonia_. Several potters with Gaulish names are known, and probably FIDENATIS on a tile at Zulpich, SECVNDANVS F(_igulus_ or _fecit_) and PACATVS F from Seligenstadt, refer to craftsmen of that nationality.[2522] Often the master’s name only occurs, of which possible instances are BELLICIANVS on a tile from Caerwent, and PRIMV(_s_) on another from Colchester.[2523] In the British Museum are tiles with the initials T · P · F · A, T · P · F · C, T · P · F · P, from Rodmarton in Gloucestershire.[2524] Tiles found in the provinces also have the maker’s name simply, without indications of date or the owner of the pottery, as on those from Seligenstadt already cited. The makers must in all cases have been of inferior condition, as implied in the example already quoted of the slave Arabus (p. 354); and other names—Daedalus, Peculiaris, Primigenius, Zosimus—belong to the same rank of life. Yet the occurrence of a single name for a private individual is everywhere very common. On the other hand, imperial slaves usually have two names given, and freedmen three.[2525] On the tiles of the freedmen of the Gens Domitia (dating about the reign of Hadrian) is frequently stamped the formula VALEAT QVI FECIT, “May he who made it prosper,” with the name of the representative of the family in the genitive.[2526] On other tiles we find such expressions as VTAMVR FELICES, “May we use it and be happy”[2527]; FORTVNA COLENDA, “Fortune is to be worshipped” (a second-century tile)[2528]; and on others of post-Diocletian date, VRBIS ROMAE, “The city of Rome”[2529]; SECVLO CONSTANTINIANO, “The age of Constantine”; FELIX ROMA (on the tiles of Theodoric), “Happy is Rome.”[2530] Even on sepulchral tiles of late Imperial times are stamped such aspirations as, VTI FELIX VIVAS, “May you live happily.”[2531] [Illustration: FIG. 195. INSCRIBED TILE FROM LONDON (GUILDHALL MUSEUM).] Again, memoranda are found incised on the tiles, as on one at Hooldorn in Holland, KAL · IVNIS · QVARTVS LATERCLOS N(_umero_) CCXIIII, “Quartus (made) 214 tiles on the first of June”; and on another, found in Hesse in 1838, STRATVRA TERTIA LATERCVLI CAPITVLARES NVM · LEG · XXII, “In the third layer large tiles of the number of the twenty-second legion.”[2532] A tile found in Hungary had scratched upon it two metrical lines in cursive writing: _Senem severum semper esse condecet_ _Bene debet esse povero_ (sc. _puero_) _qui discit bene_[2533]; and on others names such as Tertius, Kandidus, Verna, were incised.[2534] Idle boys in the brickfields often seem to have scratched the alphabet or other words in the soft clay, and complete Roman alphabets are found at Hooldorn[2535] and Stein on the Anger[2536]; the letters I K L M on one at Winchester[2537]; on another at Silchester is ... E PVELLAM.[2538] On a tile in the Guildhall Museum (Fig. 195), found in Warwick Square, E.C., are the words AVSTALIS | DIBVS · III | VAGATVRSIB | COTIDIM, of which no satisfactory translation has been given, but it has been usually regarded as the gibe of a fellow-workman at a devout individual.[2539] On another, now at Madrid, the first two lines of the _Aeneid_ are written in excellent cursive characters of the first century after Christ.[2540] The Roman tiles, if rightly used, are found very useful for judging the dates of buildings. For instance, a study of those in the Pantheon showed that the walls were neither the original ones nor those built by Agrippa in 27 B.C., but were restored in the second century or supplied then with new brickwork. On the other hand, the stamps from the Flavian amphitheatre and Thermae Antoninianae confirm the dates of those buildings. Those tiles which bear the name M. Aurelius Antoninus as consul[2541] seem to be the Emperor Caracalla’s. In the time of Diocletian the dates cannot be definitely ascertained, but before his time the shape of the stamp is a good criterion. Rectangular stamps are found in the best period, and in the first century B.C. only one line of inscription is usual. Two lines denote the period 50–100 A.D. or later; semicircular or lunate forms came into use under Claudius, and lasted to the end of the first century; perfect circles belong to the same period. The type with the cut-out _orbiculus_ came in about Nero’s reign, and the size of the _orbiculus_ gradually diminishes down to that of Severus, while the inscriptions gradually increase in length.[2542] A considerable number of the Roman tiles are inscribed with the names of the consuls of the current year in which they were made, presenting a long and interesting series, from the consulship of L. Licinius Sura and C. Sosius Senecio (A.D. 107) to that of Severus Alexander (A.D. 222). Many of these consulships do not, however, appear to have been recorded in the regular _fasti consulares_ or official lists, and they were probably _suffecti_, whose names were not recorded after their temporary elevation. It seems likely that the occurrence of consuls’ names implies that such tiles were destined for public buildings, and were so marked to prevent their being stolen with impunity. They are fewer in number than those which have merely the names of _praedia_ or potteries, but are yet sufficiently numerous to be an invaluable aid in tracing the succession for upwards of sixty years. Inscriptions of this class are only found on _opus doliare_, and chiefly in Italy. Their appearance is probably due to some law passed by the Senate about the reign of Trajan to regulate the potteries. As an example may be given a tile from Hooldorn in the Netherlands, inscribed SVB · DIDIO · IVLIANO · COSS[2543]; the date is A.D. 179, the name being that of the future emperor (COSS is a mistake for COS). The following examples are taken from Dr. Dressel’s scheme of the chronological order of the stamps,[2544] and show the style of inscription characteristic of the different periods: I. First century after Christ.