probably in clay, with patterns of leaves, geometrical designs, or human figures posturing, etc., and with these he would stencil running friezes, not into each bowl, but into a mold which could serve for the production of hundreds of bowls. The designer was a trained craftsman who could model in clay and who had some taste in the composition of patterns, but we need hardly suppose that he was an original artist like the men who so frequently produced exquisite work in the famous Greek vases; for in Arretine ware the patterns were usually borrowed from those of silver plate.

To judge from the instances in which we can actually apply a test to the form of the signature,1 the designer was usually a slave or a freedman. If the designer was a slave we may be fairly sure that the ordinary laborers were. The owners of the factories are of course Roman citizens, but it is surprising how frequently they bear a foreign cognomen, a fact which implies that they or their ancestors of no remote date had come up from slavery. Indeed some of the owners appear to be the very persons who designed the patterns of an earlier style, an indication that the slave-artists sometimes secured their freedom and a sufficient competence to gain possession of their masters' factories.

The extensive proportions of some of the factories are proved beyond a doubt. So, for instance, the ware of certain firms has been found, not, to be sure, over the whole Roman world-for each firm seems to have supplied the regions opened by the natural arteries of trade-but at least over half of the Mediterranean basin. Indeed for one period it is true that the potteries situated in three districts, i.e., near Puteoli, at Arretium, and in the valley of the Po, supplied the whole demand for moderately good table-ware throughout the Empire, excepting only the southeast. The scale of production is also indicated by the great number of workmen engaged in certain firms. That of Cornelius, for instance, has provided the names of some forty designers; to be sure, they were not all con

1 See Am. Hist. Review, 1916, p. 693, for criteria; also Oxé, Rheinisches Museum, 1904, p. 28. By way of contrast it may be noted that the Calenian ware of two centuries earlier seems to be designed almost wholly by the shop-owners, who were free citizens. See Pagenstecher, Die calenische Reliefkeramik, pp. 148 ff.

2 See Notiz. degli Scavi, 1896, p. 455, for the description of a large workroom. The mixing-vat had a capacity of 10,000 gallons. The accounts of the Gallic potteries also prove large-scale production; see Déchelette, Les vases céramiques de la Gaule, p. 91.

temporaneous, but at any rate a single designer could keep a large number of mixers, potters, and furnacemen busy, since he presumably merely made the molds and touched up the designs. Calidius Strigo had at least twenty designers, Perennius as many, and there were a dozen other firms of goodly proportions at Arretium. Finally, mass production with a view to extensive trade is disclosed in the establishment of branch factories in Gaul and elsewhere, the purpose being, of course, to save what was in that day the heavy item of freight. Indeed the home factories were eventually put to rout by these new ones, whether because the clays of Gaul were better, the makers more enterprising, or the provincial market more conservative when fashions began to change in Italy. So, for instance, a consignment of red ware that had reached Pompeii shortly before the eruption-the box had not yet been opened-contained more Gallic pieces than Italian, although the box had apparently been packed at Rome.1

In this industry, then, we find the machinery of an extensive factory production of articles intended for wide distribution. Of course the student of Roman society sees in this instance an exception to, rather than an example of, the usual rule, but it is apparent that conditions here favored the development of large-scale production. Two elements were of prime importance. One was the quasi-tradesecret involved in the making of the paste, for, though there was no copyright and this particular clay could be and in fact was manufactured in several places, exact knowledge of a rather intricate formula was after all essential. Secondly, a designer of some skill, training, and taste was required; consequently competition could not spring up overnight, and the expense of keeping a skilled designer naturally suggested the advisability of gathering under him enough unskilled labor to occupy his time. Hence it is that this industry developed in a way that was rather unusual in the Roman world.

By way of contrast both in workmanship and in conditions of production it is interesting to compare the manufacture of another article of pottery, namely, the ordinary clay lamp,2 millions of which

1 See Atkinson, Journal of Roman Studies, IV, 27.

2 See CIL, XV, 784; Fink, Sitzungsb. Akad. München, 1900; Loeschcke, Keramische Funde in Haltern, p. 210.

must have been manufactured every year and sold for a very few cents apiece. Many of these lamps have a little decoration, but seldom does a pattern show any real worth. They were turned out in molds by an ordinary potter, and the clay paste was little better than that used in good roof-tiles. Furthermore, they were so cheap that it would hardly have been worth while to ship them any considerable distance. To be sure, the recurrence in all parts of the world of certain types of lamps bearing a well-known firm-name succeeded until recently in deceiving archaeologists into thinking that certain firms commanded the trade over wide areas. But it has now been proved1 by measurement that the greater number of these lamps came from local potteries that simply used various shapes successively popular at some center like Rome, importing the originals and using them, firm-name and all, as molds. Since, then, in the absence of protective copyright, there was here no difficult formula or trade-secret to aid in excluding competition and no great economic inducement for gathering considerable labor and a definite requisite overhead expense, the industry scattered in such a way that local potteries usually supplied the needs of each locality. Concerning the class of labor used we have some indications. The firm-names are usually in briefest form, a cognomen alone, though in early examples good Roman gentile names occur. Judging from the frequency of Greek cognomina we may suppose that the potteries which produced these cheap wares fell into the hands of the class that in general managed Rome's industries, at least in the early Empire, i.e., the freedman.

Certain developments of the glass industry in Augustus' day2 bring us back to conditions not unlike those of the red glazed pottery. Glass-making apparently grew out of the art of surface-glazing in Egypt at a very early age, and in Roman times the glassware of Alexandria, chiefly mosaics of varicolored glass pastes, was shipped the world over. It is likely that there were very large factories in Egypt, but since the ware bears no trade-mark and since it was successfully copied at Rome and elsewhere we are quite unable to deter

1 See Loeschcke, op. cit., p. 210.

2 Kisa, Das Glas im Altertum, pp. 261 ff., 702 ff.; Eisen, in Am. Jour. Arch., 1916, p. 143; Morin, La Verrerie in Gaule, 1913; CIL, XV, 871.

mine the proportions of monopolistic production. There is, however, a translucent glass of Rome, usually figured and signed, and apparently made with the blow-pipe, that provides some little information of value. When Strabo says that in his day certain new inventions at Rome had greatly increased the production of glass and brought down the price to a cent or two per article he may well be referring to the discovery of the process by which a bubble of glass paste was manipulated by means of the blow-pipe. It is obvious why this method revolutionized the production of clear glass. Hitherto for the making of bottles and for many shapes of beakers a new mold had to be shaped for each individual article, a labor-consuming process, and one which, because of the sand and clay of the mold, left the article far from clear. With the blow-pipe a permanent outer mold was used which might contain the figured pattern— figures were the fashion in all wares of Augustus' day-and the glass-blower with the use of his pipe could force the paste to fill the mold and assume the pattern desired. The product was clearer and smoother and the work was far more rapidly done than by the old method. It is not surprising that the maker of the new glass, inartistic though it was, showed such enthusiasm over the new process as to put his name in prominent letters upon the pattern. The ware, which can readily be distinguished, is not only found widely distributed, but the maker's name is printed in Greek as well as in Latin, apparently on the supposition that the articles would find a wide sale.

Here again, as in the case of Arretine ware, conditions favorable to monopolistic production existed. Whether or not modern methods can extract glass paste with ease out of the sands and pozzolanas of Italy everywhere, it is clear from Strabo and Pliny that the ancient glass-maker had great difficulty in finding a tractable sand. This alone prevented much competition. Moreover, the new invention made a peculiar distribution of specialists necessary. Hitherto the workman who handled the hot glass paste at the furnace must also be skilled at molding it quickly into the desired pattern. By the new process one designer could shape any number of exterior molds, and any number of glass-blowers might produce the articles on these molds, given only the special skill in glass-blowing. Thus

again it was good economy to gather labor into one place and about one designer.

There is one additional fact of interest here that deserves mention in passing. The manufacturers of this signed glass bear Greek names and call themselves natives of Phoenician Sidon. It may be that some of the factories were in Sidon; at least Ennion's work is found mainly in that region. On the other hand, the work of Artas, Neikon, and Ariston appears mostly at Rome. Either we are dealing with an eastern product that captured the trade of Rome, or, what is more likely, we are dealing with skilled artisans and manufacturers who, realizing that Rome offered the best market, set up their main factories there or nearby.

The brickmaking industry was another which tended toward factory and monopolistic methods at Rome, though for a very different set of reasons. During the Republic the industry found little encouragement. Public buildings were largely made of tufablocks, and when concrete was introduced it came to be lined with stone, large blocks or small squares set in cement. Clay was burnt chiefly for roof-tiles. In the early Empire till Claudius' day the stone facing and opus reticulatum still continued in vogue, though broken roof-tiles were also introduced for the facing of concrete walls. In the reign of Claudius, however, brick-facing became more general so that brickyards had to supply new forms in addition to the rooftiles. It was in Nero's reign, especially when, after the famous fire, a great part of the city had to be rebuilt, that the brick industry came to its own. It is evident that the brick kilns then in existence had to supply an inestimable quantity of material for the facing of concrete walls, and brick-faced concrete remained the standard material for construction thenceforward.

There were of course both centrifugal and centripetal forces in this industry at Rome as elsewhere. The recipe was centuries old, and by no means a secret. Furthermore, good clays for bricks were abundant. To be sure, the excellent Pliocene shales behind the Vatican that now feed the great kilns of Rome seem not to have been exploited to any great extent in ancient times, but the alluvium of the Anio and of the Tiber that combines the limestone silt of the 1 CIL, XV, 1 ff.; Van Deman, Am. Jour. Arch., 1912, pp. 247 ff.