^{page 256 note 1} C.I.L. vii, 1201, 1202, 1211. The old view of 1201 as a ‘trophy’ has died hard (cf. Balch, , Wookey Hole, 1914, p. 96Google Scholar).

^{page 256 note 2} Archaeologia lvii, 1901, p. 402, Table iii.

^{page 256 note 3} E.g. Haverfield, , Vict. Co. Hist. Somerset i, 1906, pp. 340–343Google Scholar; Besnier, , Revue arch. Ser. v, xiii, 1921, pp. 40–45Google Scholar; Cough, , Mines of Menaid 1930, pp. 20 ff.Google Scholar

^{page 256 note 4} I am indebted to Dr. James Chambers, Super intendent of The Priory, for generous permission to examine and photograph these pigs: to Dr. Brown of The Priory for valuable aid in photography and weighing: and, for information received, to Dr. Bristoe of Wrington, Somerset, and Dr. Thuinam of Nordrach-on-Mendip, Somerset.

^{page 257 note 1} It should be noted, in view of the sequence implied in V.C.H. Som. i, p. 341, no. 6, that when Scarth's original communication to the Society of Antiquaries (dated March, 1877) appeared in the volume for that year, the pig had already been published by Hübner in Eph. Epigr. iii. Scarth, whose notes on the details of Eph. Epigr. iii, 121, a, b, and d, seem to have been very inadequate indeed (cf. below pp. 260, n. 3, 261), was responsible for these inaccuracies in the first instance.

Since the first published accounts of 1877 there is no record of any fresh examination of the pig, though some of the details then given were admittedly in dispute, nor has any Thotograph of this pig hitherto been available, Haverfield's account in V.C.H. Somerset (published in 1906) was not given from first-hand knowledge, though details and a rubbing had been obtained (l.c. p. 341, no. 6).

^{page 257 note 2} Where, as Scarth himself pointed out (Soc. Ant. Proc. l.c), this is wrongly termed a ‘fragmentum’ by Hübner. This pig also came into Dr. Woo' hands, and was later presented by him to Bristol Museum, where it now is.

^{page 257 note 3} Gowland (Archaeol. l.c.) makes the unprecedented error of printing the main inscription as IMP. VESP. AVG.

^{page 257 note 4} The final stop is also found in C.I.L. vii, 1209, c. and d.; and there are traces of an initial (foliate ?)stop in 1209. a.

^{page 258 note 1} The genitive of the imperial name seems to have been the recognised usage on British pigs as else where, though noteworthy exceptions are the two Domitianic pigs from Yorkshire. Traces of the bottom of an ‘s’ are, I think, to be seen after CAESARI on the curious pig (C.I.L. vii, 1212) preserved at Chester.

^{page 258 note 2} In both instances this stands in contrast with the squareness of the ‘G’ on the Clausentum pigs. The Bristol Museum pig; of course, bears no side inscription.

^{page 258 note 3} Brit. Arch. Assoc. Journ. l.c.

^{page 258 note 4} Gowland's doubts remained even after the Clausentum discovery. See the discussion referred to in note 6, below.

^{page 258 note 5} As recently as 1921 this explanation was at first accepted by Besnier in his account of the pig(Revue Arch. Ser. v, xiii, p. 41, no. 23, a.), but in view of the Clausentum discoveries was later with drawn (ibid. xiv, p. 119, no. 7c).

^{page 258 note 6} See Soc. Ant. Proc. xxxi, pp. 36–9 for a description of these pigs, now preserved in the Tudor House Museum at Southampton. Fate seems to be against us in our attempt to reach the truth in the inscriptions concerned. On p. 38 ad loc. (the ensuing discussion) by a curious double misprint ERG (=ARG?) appears instead of VEB; and in view of the photographs there reproduced (p. 37) it is surprising to be told more recently that the two pigs resemble the one under discussion except that the whole inscription ‘runs continuously on the top surface,’ and that ‘on them there appeared to be a third letter after the VE, which looks rather like a B, but it is much blurred.’ (Gough, Mines of Mendip, p. 28, n.)

^{page 259 note 1} E.g. Gowland, Archaeol. l.c., p. 399; Haverfield, App. II, p. 178, in Tac. Agric. (ed. Furneaux and Anderson); Collingwood, , Arch. of R. Brit., 1930, p. 173.Google Scholar It should be added that no stop is visible between EX and ARG either on this or on the Clausentum pigs.

^{page 259 note 2} C.I.L. vii, 1202, which, by its apparently early date (A.D. 49), can hardly be regarded as seriously affecting the issue.

^{page 259 note 3} The whole question still needs more careful inquiry. The low silver content revealed on analysis was noted in 1886 by Perks, (Trans. B'ham and Mid. Inst. xiii, p. 11)Google Scholar, though without details being given. Since then at least sixteen of the British pigs have been analysed, the results supporting the view taken above. It is significant that a fragment of lead taken from the remains of a Roman lead furnace at Pentre (Flint) showed on analysis 9 ozs. 16 dwts. of silver to the ton—the highest figure yet recorded for Romano-British lead (Atkinson, and Taylor, , Flints. Hist. Soc. x, pt. I, 1924, p. 20).Google Scholar This furnace from associated pottery was dated by the excavators to approximately the same period as the extant Flintshire pigs; none of these are marked EX ARG, yet the Chester specimen only showed about 6 dwts. per ton (ibid. pt. 2, 1925, p. 27), and one of the Hints Common specimens 14 dwts. (Gowland l.c., p. 402–3, no. 50). The ‘lAgricola’ lead pipe discovered in Chester, which belongs to the same period (A.D. 79) and probably came from the same mining area, yielded 11 dwts. per ton. (Soc. Ant. Proc. xviii, p. 98; Reliquary vi, 1900, p. 114).

^{page 259 note 4} Cf. also the inscription on the ‘Agricola’ lead pipe referred to in the previous note.

^{page 259 note 5} Arch. Journ. xxxiii, 1876, p. 353.

^{page 259 note 6} Eph. Epigr. l.c.

^{page 259 note 7} So Gowland, Haverfield and other authorities.

^{page 260 note 1} Soc. Ant. Proc. l.c.

^{page 260 note 2} Brit. Arch. Assoc. Journ. l.c.

^{page 260 note 3} It may be desirable to point out here that this figure, though accurate for this pig, seems to have been misapplied elsewhere by Scarth. In Soc. Ant. Proc. (loc. cit.) Scarth gives the weight as 223 lbs. and adds the curious statement that the pig ‘is supposed to have lost 1 Ib. by erosion.’ Thus armed with a total of 224 Ibs. he seems to have applied it by mistake to the Bristol Museum pig, which he states to be 2 cwt.(Soc. Ant. Proc. vii, 1877, p. 159). This seems to be the explanation of a figure which puzzled Haverfield (V.C.H. Som. i p. 341, no. 7), and it is the source of the 224 lbs. which we find in Gowland's list (Archaeol. loc. cit., no. 19). Elsewhere, by some devious process, Scarth gives the weight of the Bristol pig as 2 cwt. 2 qrs. 16 lbs. (=296 lbs.). In Eph. Epigr. iii no weight is given, but the words fragmentum simile can only mean that, in spite of the absurdity, Hübner (like Watkin) regarded 121, a. and 121, b. as weighing 143 lbs. each.

^{page 261 note 1} C.I.L. vii, 1209 d. With the exception of C.I.L. xv, 7915 (apparently a pig, though of unusual shape), found in the Tiber at Rome and bearing an inscription of doubtful meaning, it is also the heaviest pig found anywhere in the Empire.

^{page 261 note 2} It must date after the acceptance of the title Pater Patriae in 139. We have no pig of Antoninus giving any other form of inscription by reference to which we might argue (as we did with the Vespasianic pig) that M. Aurelius might be in cluded after 146. At least, the date cannot be 138–9 as sometimes said (by Hübner, e.g.).

It should further be borne in mind in considering these inscriptions that there will always be a more. or less marked time-lag involved. The reorganisation, however trifling, necessitated by the changing of the moulds at the mines of a distant province cannot be expected to follow immediately upon events at Rome.

^{page 261 note 3} It need hardly be said that what we term the ‘base’ of the pig represents the top surface of the pig as actually cast, the inscription being in reverse in the bottom of the mould.

^{page 261 note 4} See below pp. 263 and 264.

^{page 262 note 1} C.I.L. ii, 6247, 2, (delphinus: gubernaculum), 6247, 3 (caduceus); Eph. Epigr. viii, 254, 1 (caduceus), 254, 2 (cycnus);Besnier, Revue Arch., Ser. v., xii, 1920, p. 233, no. 4 (delphinus).

^{page 262 note 2} C.I.L. iii, 8080; Soc. Ant. proc. xx, 1904, pp. 93–4; Num. Chron. 4th Ser., xv, 1915, pp. 489 ff.; Num. Zeit. xx, 1888, pp. 19 ff.

^{page 262 note 3} In association, e.g., with the inscription FL(AVIVS) FLAVIANVS PRO(BATOR) SIG(NAVIT) AD DIGMA (imperial mint of Sirmium).

^{page 262 note 4} It should be added that in discussing the palm-branch as a symbol of the chief official of various provincial mints (including London) in the fourth century (cf. Kenner, , Num. Zeit. xx pp. 37Google Scholar, 40). Sir Arthur Evans suggested a close connecton between treasury officials and the silver-producing mines of the Mendips, (Num. Chron. 1915, pp. 488Google Scholar; 498 ff). The inclusion of the emblem in a mould for pigs of lead might, therefore, conceivably be regarded as a certificate of previous or intended desilverisation under official supervision for treasury purposes. It might consequently under certain circumstances be regarded as an equivalent for EX ARG. The relative dates involved remain a difficulty (cf. p. 264 below). In the two other pigs of Antoninus' reign found at Charterhouse (Brit. Mus.; Bristol Museum), no such symbol appears, though they have certainly both been desilverised (Gowland l.c., p. 423, no. 15; V.CH. Som. i, p. 343; Friend, and Thorneycroft, , J. Inst. Metals, vol. xli, 1929, no. 1, p. 113).Google Scholar

^{page 262 note 5} By Mr. R. G. Collingwood, to whom I submitted a photograph and a cast.

^{page 262 note 6} Experiments (undertaken at Mr. Collingwood's suggestion) showed that a blow upon very coarse material laid upon the cold metal resulted in a series of dots, representing the points at which the threads crossed. A really heavy blow produced no better result and in addition broke the fabric used. These results cannot, of course, be regarded as conclusive, but they are suggestive in view of the external evidence provided by the Shropshire pig.

It may be added that although, as noted above (p. 260), there is damage at the right-hand side of the inscribed face, this is in an entirely different plane and cannot have been due to the same cause as this depression. The corner of a pig would also probably cause more of a gash and much more extensive damage.

^{page 263 note 1} Brit. Arch. Assoc. Journ. xvi, 1860, p. 350. It takes the form of a note on the exhibition of a drawing of a lead pig seen by the British Archaeological Association at a visit in 1860 to Linley Hall, Shropshire.

^{page 263 note 2} With regard to the explanation put forward by the writer of the note, it should be said that there is no evidence that the Romans used sand in this way for casting the pigs. Perhaps they used moulds of stone, but more probably, in view of the character of the pigs and their inscriptions, of clay.

^{page 263 note 3} J. D. la Touche in Intellect. Obs. xi, 1867, pp. 354–5: an interesting article not only for its good discussion of the Linley pig but as containing the first evidence for the alleged Roman candles later mentioned by Wright (Uriconium, 1872, p. 266–7: cf. Haverfield, , V.C.H. Shropshire I. 1908, p. 264Google Scholar).

^{page 264 note 1} Beginning with these statements in 1860 and 1867 it reappears in Pulsifer, (Notes for a Hist, of Lead, New York, 1888, p. 30)Google Scholar and even in Haverfield, (V.C.H. Shropsh., i, p. 264)Google Scholar, though with some doubt. The explanation lies in a misunderstanding of the ’stratification’ of the pig, which admittedly bears some resemblance to the grain of oak. But this is observable in almost any specimen found in this country.

^{page 264 note 2} Though Haverfield apparently knew of the former passage, since a reference to it occurs in V.C.H. Shrops., i, p. 264.

^{page 264 note 3} Permission to examine this pig was kindly given to me by Prof. W. S. Boulton, Head of the Geology Department in the University of Birmingham, and I am indebted to Mr. W. H. Laurie, of the same Department, for a photograph.

^{page 264 note 4} I have to thank the Liverpool City Museum authorities for affording me access to their pig. The mark is there so badly worn as to be almost indistinguishable. Mr. C. F. Hawkes generously made an examination of all the British Museum pigs and informs me that the branch appears on the side next to the top of the inscription only. It is in good relief and is inverted, running downwards from the top of the pig.

Note that all extant Shropshire pigs are from different moulds (cf. Way, in Arch. Journ. xvi, p. 34Google Scholar). It may be added that in my own view, which I hope to elaborate later, nos. 1209, a, and 1209, b, are one and the same pig.

^{page 264 note 5} We cannot date exactly any known pig of Hadrian, all those so far discovered giving us the minimum of title. The Bath and all recorded Shropshire specimens read IMP . HADRIANI . AVG . one from Derbyshire IMP . CAES . HADRIANI . AVG . MET . LVT; so, too, the one found near Cheshunt (Herts.) but without MET . LVT (cf. C1L, x, 8073, 2, from Sardinia). Of the one found in Yorks., and now missing, we only know that it was ‘stamped with the name of Hadrian.’ This latter, like its elder brother of Trajan's reign, is not so well recognised as it should be in the history of Roman mining in Yorks, having been passed over in favour of the well-advertised Domitianic pair. For the Hadrianic pig, see Speight, H., Romantic Richmondshire, 1897, p. 207Google Scholar (wrongly stating it to be in the Brit. Mus.), Edwards, W., Early History of N. Riding, 1924, p. 28Google Scholar, Raistrick, A., Trans. Newcomen Soc. vii, 1927, p. 81.Google Scholar For the Trajanic pig, Lucas, J., ‘Grassington Lead mines’ in Old Yorkshire (2nd Ser.), ed. Wheater, W., 1885, pp. 49Google Scholar, 50, A. Raistrick, loc. cit.

^{page 264 note 6} Cf. the two tiny stamps on the rim of no. 1209, a (Haverfield, , V.C.H., eph. Ep. ix, p. 643Google Scholar).

^{page 264 note 7} See above, p. 262, n. 4.