, “Receive me, Helioserapis,” by which the name of the vessel may be intended.
Most lamps had only one wick, but the light which they afforded must have been feeble, and consequently the number was often increased. When the number is not large, or when the body is circular (as in Plate [LXIII]. fig. 4), they project beyond the rim of the lamp, as in Class I. already described, but the lamps which have a large number are usually boat-shaped or rectangular in form (see Plate [LXIII].), and the nozzles do not then project, but are ranged along the sides, merely indicated by separate moulding underneath.[[2787]] Occasionally a conglomeration of small lamps was made in a row or group, but even in these cases the illumination given must still have been feeble. The average size of a lamp is from three to four inches in diameter across the body, the length depending on the form of the handle and nozzle, but averaging about an inch over the diameter, and they are mostly about an inch in height. The top of the lamp is almost always circular in form, occasionally oval, and rarely rectangular,[[2788]] and is usually slightly depressed, being thus shaped to enable any overflow of oil to run down through the filling-hole. Many Greek lamps, and Roman lamps from Greek sites, such as Cyprus, are convex above, with a small moulded disc on the raised centre, in which is the hole. These are either devoid of decoration, or only have an ornamental pattern or a frieze of figures on a small scale. Usually the subject is enclosed within a plain moulded rim, but in the later examples (Class IV.) especially it is more contracted in extent, and surrounded with a border of ornament, such as the egg-pattern or a wreath of some kind (see Fig. [208]).
Christian lamps, which hardly come within the scope of this work, vary very little in form; they have ovoid instead of circular bodies, a plain rounded nozzle, and a small solid handle, and the design is always encircled by a band of ornamental pattern or symbolical devices.[[2789]]
The clay of which the lamps are made is usually of a red colour, due to the presence of red ochre (rubrica), but it varies both in quality and tone according to localities; those from Greek sites, such as Athens and Corfu, are often of a pale buff colour, those from Cyprus a light reddish brown, and so on. Martial refers to the red clay of Cumae,[[2790]] a place where lamps are sometimes found, and those from Naples are usually of a dull brown or yellow colour. Lamps found in France and England are often imported from Italy, and therefore of the ordinary red clay, but those of local manufacture are of a white or yellowish tone.
FIG. 209. MOULD FOR LAMP FROM CATANIA (BRITISH MUSEUM).
The earliest undecorated examples are made on the wheel, as are those from the Esquiline and from Carthage, in which the decoration is only incised; but subjects in relief required a different technique. Occasionally they are modelled by hand, but we find that from the first century B.C. onwards they are almost invariably made in moulds, modelled from a pattern lamp, in a harder and finer clay than the pattern.[[2791]] The mould was divided into two parts, adjusted by mortices and tenons, which, in the opinion of some writers, explains the lateral projections visible on certain varieties; the lower part formed the body of the lamp, the upper the decorated discus. The two parts seem to have been marked by corresponding letters to avoid errors, and there are two or three lower lamp-moulds in the British Museum from Ephesos and elsewhere, marked with an A on the under side for this purpose.[[2792]] Other examples of moulds have been found in Greece, Italy, and Africa,[[2793]] and there are also specimens both for the upper and lower half in the Guildhall Museum.[[2794]] They were either of terracotta or plaster.
The clay was impressed into the mould with the fingers, the figured decoration being applied by means of models or stamps, as with the Arretine ware (see below, p. [439]), and the ornamental patterns probably produced with a kind of wheel or running instrument, as in Roman pottery (p. [441]). Signatures in relief were taken from the mould, those in hollow letters were impressed in the lamp itself from a stamp before baking. Important potteries must have possessed a large number of moulds; for instance, at Rome alone ninety-one different subjects are found on the lamps of one potter (L. Caecilius Saevus), eighty-four on those of C. Oppius Restitutus, fifty-one on those of Florentius, and there must of course have been many more now lost. It is clear that the same types were used by different potters; the models must, therefore, have been handed about from one to another, each potter merely adding his own name.