¹ Dubs, H. H., trans., History of the Former Han Dynasty, vol. III,Google Scholar ch. 99 C: 25a, b (To be published. Chapter and paging according to the Chinese edition of Wang Sienchien, Han-shu Bu-ju).

² Wang Mang's possession of a great treasure is independently attested by the size of his betrothal present to the family of his second wife (whom he married in A. D. 23). This present consisted of 30,000 catties of gold (235,000 oz. or 7,320,000 g.), besides chariots and horses, male and female slaves, variegated silks, jewels, and treasures, worth approximately as much again (Dubs, III, 99 C: 20a).

For a study of these Chinese standard weights and measures, cf. , Dubs, op. cit., vol. I (1938), 276–280.Google Scholar A study of Ban Gu's History shows that its author was unusually careful. He seems to make no important statements without some documentary authority. When the capital, Ch'ang-an, was captured and the usurper Wang Mang killed in A. D. 23, the victorious generals undoubtedly reported the circumstances to their Emperor, for they expected high rewards. It would have been only natural for them to include in this report an account of the treasure captured. Such memorials were preserved in the imperial archives, to which Ban Gu later had free access. It is thus highly probable that Ban Gu's statement came from such a report, so that these figures were taken from a contemporary account.

It would be difficult to appraise the reliability of such a memorial, even if we had the original. The round number of chests mentioned, “sixty,” is itself suspicious, for it seems to represent an estimate, rather than an enumeration. The captors may have exaggerated the number of chests, in order to increase their reward. But they would hardly have exaggerated very much, since they would expect to be held accountable for what they had secured. The important buildings in the imperial Wei-yang Palace had been burned in the attack upon the usurper, but Ban Gu says specifically that, after Wang Mang's death, when order was immediately restored, the treasuries were still intact (Dubs, III, 99 C: 29a). It is moreover possible that some of these chests were not quite full or were even empty. Wang Mang was, however, stingy with his money, and even when attacked was very reluctant to disburse any large sums (Dubs, III, 99 C: 25b). I think it is rather likely that Wang Mang actually possessed at least a large proportion of the gold with which Ban Gu credited him.

³ Found in the Shzh-ji, 30: 11 (written about 100 B. C), translated by Chavannes, E., Memoires historiques, III, 553.Google Scholar This information is very likely from the imperial treasurer's report. It probably summarizes the expenditures from the beginning of the reign in 135 to and including 123 B. C.

⁴ The only very large ancient treasure known to me is that of the Persian empire. Herodotus states that the regular revenue of Darius I was valued at 14,560 Euboic talents, but almost all of this amount (as well as the talents in which it is valued) was silver, not gold (Herodotus, III, 89–97). The only Persian province remitting in gold was India, which contributed 360 talents of gold dust yearly (ibid., III, 94), equivalent to 9,072,000 g. (taking the drachm at 4.2 g., according to Hill, G. F., in Cambridge Ancient History, IV, 133).Google Scholar Since much of this revenue was expended, we cannot tell how much accumulated.

Darius III drew largely upon his wealth for the war with Alexander the Great and carried off with him 8,000 talents [of silver] in his flight; nevertheless Alexander is said to have obtained from Susa, Persepolis, and Pasargadae coined and uncoined metal to the value of 1,800 [silver] talents; but most of this metal was silver. At Susa he obtained 40,000 silver talents and gold darics to the value of 9,000 silver talents (Dio-dorus Siculus, bk. xvii, ch. 7 [cf. Booth's trans.]; Gray, G. B., in Cambridge Ancient History, IV, 199).Google Scholar This amount of gold (calculated at the ancient rate of 13½:l [cf. G. F. Hill, ibid., 135]) was only 1,608,320 g. If the proportion of gold to silver secured at Susa was maintained elsewhere, Alexander obtained only a few million grams of gold, so that the Persian gold treasure did not compare with the approximately 150,000,000 g. possessed by Wang Mang.

Roman conquests seem also not to have secured any large quantities of gold. During the whole period 200–150 B. C. (which seems to be the only period during which we have many figures concerning booty and products of mines), Rome received a total of only 31,735 pounds of gold in booty from the east and 6,316 pounds from Spain (including the products of Spanish mines) (Frank, T., Economic Survey of Ancient Rome, I, 138).Google Scholar Taking 327.45 g. to the pound (Paully-Wissowa, Real-Encyclopaedia, art. “libra”), this 38,051 pounds amounted to only 12,307,272 g. It should be added, however, that Rome much preferred silver to gold and chose the former when an indemnity might be paid in either metal. Roman specie was drained off to the east, and there seem to have been no very large accumulations of gold in Roman territory.

⁵ Joseph Kitchin, Encyclopaedia of the Social Sciences, article entitled “Gold.

⁶ Hamilton, E. J., American Treasure and the Price Revolution in Spain, 42Google Scholar, computes it as 181,333,180 g. (I owe this and preceding reference to Professor R. S. Smith of Duke University.)

⁷ Dubs, I, 111, n.3.

⁸ Dubs, II, ch. vi, app. iii.

⁹ Dubs, III, ch. 99, app. ii, n. 4.

¹⁰ Herodotus, I, 215.

¹¹ For a more extended account of Wang Mang's many economic experiments, cf. , Dubs, “Wang Mang and His Economic Reforms,” T'oung Poo (Leiden, Holland), v. 35 (1939–1940), livr. 4, 219–265Google Scholar.