Scanned text contains errors.
was called the gold talent, or the Sicilian talent from .its being much used by the Greeks of Italy and Sicily. This talent is perhaps connected with the small talent which is the only one that occurs in Homer. - The Italian Greeks divided it into 24 nummi, and afterwards into 12 (Pollux, ix. 6 ; Festus, s. v. Talentuni).' [Compare nummus, p. 814.]
This small talent explains the use of the term great talent (magnum talentum), which we find in Latin authors, for the silver Attic talent was great in comparison with this. But the use of the term by the Romans is altogether very inexact; and in some cases, where they follow old Greek writers, they use it to signify the old Attic or Eubo'ic Talent.
There are other talents barely mentioned by ancient writers. Hesychius (s. v.) mentions one of 100 pounds (Anrpoij/), Vitruvius (x. 21) one of 120 ; Suidas (s. v.\ Hesychius, and Epiphanius (de Metis, et Pond.} of 125 ; Dionysius of Halicar-nassus (ix. 27) one of 125 asses, and Hesychius three of 165, 400, and 1125 pounds respectively.
Where talents are mentioned in the classical writers without any specification of the standard, we must generally understand the Attic.
10. Comparison of Grecian Weights with our own. — In calculating the value of Greek weights in terms of our own, the only safe course is to follow the existing coins ; and among these (for the reasons stated under nummus, p. 811, b.), it is only the best Attic coins that can be relied on with any c r-tainty, although there are many other coins which afford valuable confirmatory evidence, after the standards to which they belong have been fixed.
Mr. Hussey's computation of the Attic drachma, from the coins, is perhaps a little too low, but it is so very near the truth that we may safely follow it, for the sake of the advantage of using his numbers without alteration. He makes the drachma 66*5 grains. [drachma: comp. nummus, p. 811, b.: for tho other weights see the Tables.]
11. Roman Weights. — The outline of the Roman and Italian system of weights, which was the same as the ancient system of copper money, has been already given under As. The system is extremely simple, but its conversion into our own standard is a question of very considerable difficulty. The following are the different methods of computing
(1) The Roman coins furnish a mode of calculating the weight of the libra, which has been more relied on than any other by most modern writers. The As will scarcely help us in this calculation, because its 'weight, though originally a pound, was very early diminished, and the existing specimens differ from each other very greatly [As], but specimens, which we may suppose to be asses iibrales, may of course be used as confirmatory evidence. We must therefore look chiefly to the silver and gold coins. Now the average weight of the extant specimens of the denarius is about 60 grains, and in the early ages of the coinage 84 denarii went to the pound. [denarius.] The pound then, by this calculation, would contain 5040 grains. Again, the aurei of the early gold coinage were equal in weight to a scrupulum and its multiples. [aurum.] Now the scrupulum was the 288th part of the pound [uncia], and the average of the scrupular aurei has been found by Letronne tc be about 17^
grains. Hence the pound will be 288 x 17£= 5040 grains, as before. The next aurei coined" were, according to Pliny, 40 to the pound, and therefore, if the above calculation be right, — 126 grains ; and we do find many of this weight. But, well as these results hang together, there is great doubt of their truth. For, besides the uncertainty which always attends the process of calculating a larger quantity from a smaller on account of the multiplication of a small error, we have every reason to believe that the existing coins do not come up to their nominal weight, for there was an early tendency in the Roman mint to make money below weight (Plin. //. N. xxxiii. 13. s. 46 ; compare As, aurum, denarius), and we have no proof that any extant coins belonged to the very earliest coinage, and therefore no security that they may not have been depreciated. In fact, there are many specimens of the denarius extant, which weigh more than the above average of 60 grains. It is therefore probable that the weight of 5040 grains, obtained from this source, is too little. Hence, Wurm and Bdckh, who also follow the coins, give it a somewhat higher value, the former making it 5053*635 grains, and the latter 5053'28. (Hussey, c. 9 ; Wurm, c. 2 ; Bockh, c. 11).
(2) Another mode of determining the pound is from the relation between the Roman weights and measures. The chief measures which aid us in this inquiry are the amphora or quadrantal, and the congius. The solid content of the amphora was equal to that of a cube, of which the side was one Roman foot, and the weight of water it contained was 80 pounds. [quadrantal.] Hence, if we can ascertain the length of the Roman foot independently, it will give us the solid content of the amphora, from which we can deduce the weight of the Roman pound. Taking the Roman foot at 11*65 inches, its cube is 1581*167 cubic inches => 5'7025 imperial gallons — 57'025 pounds avoirdupois, the 80th part of which is '7128 of a pound, or 4989 grains. But there are many disturbing elements in this calculation, of which the chief is our ignorance of the precise density of the fluid, 80 pounds of which filled the amphora.
It might, at first thought, appear that the result might be obtained at once from the congius of Vespasian, which professes to hold 10 Roman pounds [CoNGius], and the content of which has been twice examined. In 1630, Auzout found it to contain 51463'2 grains of distilled water, which would give 5146'32 grains for the Roman pound. In 1721, Dr. Hase found it to contain 52037*69 grains, giving 5203'77 grains for the Roman pound. Both these results are probably too high, on account of the enlargement which the vessel has undergone by the corrosion of its inner surface ; and this view is confirmed by the fact, that the earlier of the two experiments gave it the smaller content. (See Wurm, p. 78; Bockh, pp. 166, 167.) Again, the nature of the fluid employed in the experiment, its temperature, and the height of the barometer, would all influence the result, and the error from these sources must occur twice, namely, at the original making of the congius and at the recent weighing of its contents. We can, therefore, by no means agree with Mr. Hussey in taking the weight of 5204 grains, as obtained from this experiment, to be the nearest approximation to the weight of the Roman pound. On the contrary, if this method were followed at all, w§
% o 4 "