The signatures are usually abbreviated, the full form being ex officina (officinatoris), the name being consequently in the genitive. On a lamp from Rome is EX · OF · AIACIS, ex officina Aiacis.^[3050] Sometimes, but rarely under the Empire, the nominative is used: A.B. fecit, or more commonly A.B.f. Thus we have AVGENDI, ATIMETI, C . IVLI NICEPHORI, or ASPRENAS, FELIX, TROPHIMVS. But where a single name occurs it is rarely full enough to show the case. On a lamp at Dresden the potter Diomedes calls himself LVCERNARIVS.^[3051] From the second century down to the time of Augustus the name may be either in the nominative or genitive, either the praenomen and nomen, or the nomen or cognomen only; these signatures were all incised while the clay was moist. In the period represented by the third class (see p. 401) nearly all the signatures are cognomina simply, as ATIMETI, COMMVNIS, FORTLS, STROBILI, all in the genitive. In the fourth class, or lamps of the second century, the nominative is very rare; the names are usually abbreviated, and one (cognomen), two (nomen and cognomen), or three may be found. Potteries were, as we have seen, often owned by women, hence female names are not uncommon. Abbreviations of a particular name vary considerably; for instance, L. Caecilius Saevus appears as L · CAEC · SAE, L · CAE · SAE, L · CA · SAE (see below, p. 428); L. Fabricius Masculus as L · FABRIC · MASC, L · FABRIC · MAS, L · FABR · MASC, FABRIC · MAS, and so on.^[3052] Or the praenomen may vary, and for C · OPPI · RES we find L · OPPI · RES; or, again, the cognomen, as in the case of C. Junius, where it may be Alexis, Bitus, or Draco,^[3053] or of L. Munatius, found with Adjectus, Restitutus, Successus, Threptus, and Philemo.^[3054] The variations in the names may denote potteries in connection, or successive holders of one business. In one instance the name of a workman PVLCHER occurs with that of Fabricius Masculus, in another that of PRIMVS with C. Oppius Restitutus.^[3055] Greek names, where they occur, seem to imply that the potters were freedmen, as in the case of Dionysius, Phoetaspus, and others.

The following list gives the names most frequently found, with the localities in which they occur^[3056]:—

Annius Serapiodorus (ANNI · SER): Rome, Ostia.

C. Atilius Vestalis (C · ATILI · VEST): Rome, Italy, Gaul, Britain.

Atimetus: Italy, Gallia Narbonensis, Pannonia.

L. Caecilius Saevus (L · CAE · SAE): Rome, Southern Italy, Sicily, Sardinia, Gallia Narbonensis, Britain.

Clodius Heliodorus (CLO · HEL): Italy, Africa, Spain, Gaul.

C. Clodius Successus (C · CLO · SVC): Rome, Gaul, Sardinia, Africa.

Communis: Rome, Pompeii, Gallia Cisalpina, Pannonia.

Crescens: Gaul, Pannonia.

L. Fabricius Masculus (L · FABR · MASC): Rome, Gallia Cisalpina, Africa.

Florentius (FLORENT): Rome, Italy, Sicily, Tunis, Gaul, Germany, Britain.

Fortis: Rome, Italy, Sicily, Dalmatia, Germany, Gaul, Britain.

Gabinia: Italy, Sardinia, Africa, Gaul.

L. Hospidius Crispus (L · HOS · CRI): Gaul.

C. Julius Nicephorus (C · IVLI · NICEP): Italy, Gaul.

C. Junius Alexis: Rome, Campania, Sicily, Sardinia, Africa.

C. Junius Bito: Italy, Sicily, Gaul.

C. Junius Draco: Rome, Italy, Sicily, Sardinia, Africa, Gallia Narbonensis.

L. Mar. Mi.: Rome, Campania, Sicily, Spain, Gallia Cisalpina.

L. Munatius (with various cognomina): Rome, Africa.

N. Naevius Luc. (N · NAEV · LVC): Italy, Sardinia, Spain, Gaul.

M. Novius Justus (M · NOV · IVST): Rome, Naples, Sicily, Sardinia, Africa, Gallia Narbonensis.

C. Oppius Restitutus (C · OPPI · RES): Rome, Italy, Sicily, Sardinia, Africa, Gallia Narbonensis, Cyprus.

Passenus Augurinus (PAS · AVG): Italy, Gaul.

Phoetaspus: Italy, Gaul, Pannonia.

Strobilus: Rome, Italy, Africa, Pannonia, Dalmatia, Gaul, Britain.

Vibianus: Gaul, Pannonia.

C. Viciri Agathopus (C · VICIRI · AGAT): Italy, Sardinia, Gallia Cisalpina.

It will be noted that nearly all are found at Rome, but that the others fall into geographical groups; the same name is seldom found both in the north and south of the Empire. Thus Fortis is not found in Africa, Oppius Restitutus only rarely in Gaul. Certain names are entirely localised, as Annius Serapiodorus at Rome and Ostia, L. Hos. Cri. and Marcellus in Gaul, Q. Mem. Kar. and Pudens in Sardinia. The name of Vindex, a maker of terracotta figures at Cologne (see above, p. 383), is found on lamps at Trier and Nimeguen.^[3057]

The distribution of the Fortis lamps in particular is remarkable. They have been found in several places in Gallia Cisalpina, such as Aquileia^[3058]; at Lyons, Aix, Orange, and elsewhere in France^[3059]; at Nimeguen in Holland^[3060]; at Trier, Cologne, Mainz, and Louisendorf in Germany^[3061]; in London^[3062]; in Spain^[3063]; and over the region of Dacia, Pannonia, and Dalmatia,^[3064] as well as in Rome and Italy.^[3065] The most natural conclusion to be drawn from these results is that the majority of the lamps seem to have been made in Italy, and it has been thought probable that there were three principal centres of fabric whence exportation went on in different directions—Rome and its environs, Campania for the lamps found in Southern Italy, Africa, and the Mediterranean, and Gallia Cisalpina for those found in Central Europe.^[3066] It has also been suggested that the last-named fabric centred in Mutina (Modena) and that this was the place where the lamps of Class III. (see p. 401) were chiefly made.^[3067] Outside Italy there may well have been manufactures in North Africa, where lamps are so plentiful, and in Gallia Narbonensis, to which region some signatures are peculiar. Evidence of a lamp-manufacturer in Africa seems to be afforded by the mention of praedia Pullaenorum in an inscription from Tunis,^[3068] the lamps of Pullaenus occurring in Sardinia and Africa. Local fabrics of very poor lamps were doubtless numerous.

A certain number of Roman lamps have Greek signatures, not differing in character but only in alphabet from the Latin inscriptions. The most curious instance is that of ΚΕΛΣΕΙ ΠΟΜΠΕΕΙ for Celsi Pompeii, which is found on lamps in Southern Italy^[3069]; Πομπιλίου is also found at Naples, and even Ἀβασκάντου and Πρείμου, which are usually associated with lamps made in Greece (see Vol. I. p. 108), occur on some found in Italy.^[3070] In Sicily we find the signatures of Apollophanes of Tyre (ΑΠΟΛΛΟΦΑΝ ΤΥΡΙΟ) at Himera and Proklos Agyrios (ΠΡΟΚΛ ΑΓΥΡ) at Gela and Catania^[3071]; Ῥήγλου for Regulus occurs at Tarentum.^[3072] Greek names are often found in Cyprus,^[3073] and conversely a large number of lamps found at Knidos by Sir Charles Newton bore the signature ROMANE(n) ROMANE(n)SIS, in Latin letters with the S reversed, apparently suggesting that the lamps were made by a Roman abroad.^[3074] Greek signatures are even found in Gaul and Germany.^[3075]

Mention must also be made here of the recent researches of Herr Fink^[3076] with the object of ascertaining the chronological succession and general distribution of the signatures on lamps of the Imperial period. Starting with the four main classes of forms which have already been laid down as the basis^[3077] (the distinction resting mainly on the various forms of the nozzle), he has obtained, by comparison chiefly of the lamps in the British Museum, Berlin, and Munich collections, the following interesting results.

Certain stamps appear to be peculiar, or almost peculiar, to each class: thus, in Class I. only, we find P. Cessius Felix and L. Munatius Successus; in Class II. only, L. Fabricius Masculus; in Class III. only, Atimetus, Fortis, Phoetaspus, and other single cognomina; in Class IV., which contains by far the larger number of stamps, Clodius Helvidius, C. Junius Bitus,^[3078] L. Munatius Threptus, and C. Cornelius Ursus. The lamps of the Gaulish potter L. Hospidius Crispus are all of one peculiar form, a transition between Fink’s I. and IV.^[3079] Cross-instances are very rare, but C. Junius Draco is found in Classes I. and IV., C. Oppius Restitutus in Classes II. and IV., Florentius and Celsus Pompeius in Classes III. and IV. It is also interesting to note that there are lamps in Class IV. with the Christian monogram and the figure of the Good Shepherd. In Class I., generally speaking, signatures are very rare; in Class III. they are almost invariable, but the total number of lamps is relatively small. Another curious result is that certain signatures, such as L. Caecilius Saevus, Bassus, Cerialis, Sextus Egnatius Aprilis, and Romanensis, are not confined to one type of lamp, but in these cases it is to be noted that each type has a variation of signature: thus, in Class I., L·CAEC·SAE; in II., L·CAE·SAE; in III., L·CA·SAE; while in IV., L·CAE·SAE occurs no less than 140 times.

His conclusions are that one workshop did not necessarily set itself to produce only one form, but that the differences in form are merely due to changes of fashion. In Class I. Greek technical instincts are still strong as regards form and choice of subjects, but in ornament the taste of Southern Italy prevails; the subjects are mainly mythological. In Class II. the typically Roman motives appear: gladiators, combats, and hunting-scenes; this form, according to Fink, is more developed than Class I. Evidence which has been obtained from Regensburg shows that Class III. belongs to the time from Augustus to Hadrian, and, as we have seen, it is chiefly confined to the north of the Apennines. Where provincial potteries can be traced, as at Westerndorf and at Westheim in Bavaria, the lamps are usually of this form, but it was doubtless imitated in Italy. Form IV. is essentially Italian, but is also found in Central Europe, and is evidently of late date.