their different clans offered to bring into the field on a certain occasion at 306,000. I think that the above sum total was intended to express the proper militaris ætas of their nation; and consequently to be a fourth of their whole population. Strabo tells us the Belgæ consisted in his time of fifteen clans or ovn, and once could bring into the field an army of 300,000 soldiers; in which he seems to have had his eye on this passage in Cæsar. Cæsar himself informs usi that the Aduatuci, who promised on the same occasion 28,000 soldiers, consisted only of 53,000 + 4000, or 57,000 in all. Whence it appears that they promised half their population or all their males, excepting children. In like manner, the Nervii, who had promised 50,000 soldiers, had in fact but 60,000 in all: out of which number he tells us they lost all but 500, or as the Epitomizer of Livy has it, all but 300; the whole of their adult male population. On this principle, Belgium contained a population of about 1,200,000: and the whole of Gaul, if four times as great, contained one of 4,800,000.

Diodorus Siculus indeed has a statement k that the greatest nation in Gaul contained a population of nearly 200,000 males, and the least, one of 50,000: between which the average would be 125,000. But that this statement is erroneous, either in restricting these numbers to the male population only, or in the numbers themselves, or in both, may be rendered very probable. Cæsar mentions an instance, in which the gross population of five nations was 368,000; which was but 73,000 and upwards, male and female, to each. The Aduatuci, as we saw, were but 57,000 in all. Belgium with 15 nations, according to Strabo, (cf. Cæsar De B. G. ii. 4.) had a population of 1,200,000 in

Lib. iv. cap. 4. §. 3. 56, 57. i De Bello Gallico, ii. 4. 28, 29. 33. Cf. Livy, lib. civ. kv. 25. 1 Lib. i. 29.

all; which is 80,000 apiece. The average of Belgium, I should consider to be a very fair average for the whole of Gaul. Now Gaul contained, according to Strabo", 60 nations; and according to Servius ad Æneidem, i. 286, 64; and that the first of these numbers may be looked on as true, we may infer from a variety of passages in Strabo ". Among these, the 20 nations of the Aquitani, he observes, were μupà kaì ädoğa, in his time; which also must contribute to discredit the statement of Diodorus. Assuming then the number of evŋ at 60, and the average rate of population at 80,000, we obtain the sum total of the inhabitants of Gaul, 4,800,000: a conclusion exactly the same as before.

A Roman census took an account of all the members composing the family of a Roman citizen; male and female, adult or non-adult, alive or dead, bond or free. The published results of such accounts, indeed, did not comprehend the sum of all, but only of the free portion of the whole. This free portion included the women and children, who possessed the rights of citizens, as well as the men P: and there is no reason why they should not be considered to be comprehended in the joint amount of the cives Romani at a given time, as well as the men*. That they were so comprehended in this instance of the census in the time of Claudius appears from the following fact, which is on record in reference to it.

There is extant an ancient inscription to the pur

* We find the orbi and orbæ, the pupilli and viduæ, sometimes expressly excepted, as Livy, iii. 3. and lib. lix. Epitome; which implies either that at other

times they were included, or that women not viduæ, and minors not orbi or orbæ, ordinarily were so.

m Lib. iv. 3. §. 2. 44. Cf. Pliny, H. N. iii. 24. n Lib. iv. 1. §. I. 4: 2. §. I. 37, 38: 4. §. 3. 56, 57. Cf. Geographi Min. i. 46. 48, 49, 50. Marciani Periplus, ii. o Dionysius Hal. iv. 15. ix. 25. Cf. Frontonis Opera inedita, Pars ii. 444. Epistolæ Græcæ, vii. p Cf. Pliny, Epp. x. 4. 107.

port that, Temporibus Claudii Cæsaris, facta hominum armigerum ostensione in Roma, (reperta sunt) septies decies* centena millia lxxxvii. If this inscription be an authentic document, it refers to the census in the eighth of Claudius, U. C. 801: and the authenticity of the inscription is strongly confirmed by the proportion which it asserts between the armigera pars, and the gross amount of the census: viz. 1,700,087: referred to 6,944,000. This is as nearly as possible the proportion of one to four: and such, it is calculated, is the proportion which the part of a given population, at a given time, fit for war bears to the whole. Of this proportion, we may adduce the following instances, which will illustrate the truth of the assertion.

Cæsar himself informs us that out of 368,000 Helvetii, the militaris ætas amounted to 92,000; that is, just to one fourth of the whole.

Strabo mentions" that when the Salassi were reduced by Augustus, U. C. 729, out of 36,000 in all, 8000 were able to bear arms. This is not quite a fourth; but the deficiency may be explained by supposing that they had lost 1000 of their soldiers, before they were reduced.

Velleius Paterculus s tells us that in the revolt of Pannonia and Dalmatia, U. C. 760, out of a population of 800,000 and upwards, 200,000 and upwards took the field; that is, one fourth of the whole.

As then we perceive an excess of nearly three millions in the census, U. C. 801, above that in U. C. 767; so, if the population of Gallia Comata was taken into account in the former census, but only that of Gallia

* Septies decies, that is, 17, dici sine et conjunctione, et not 70. Cf. Varro, Fragmenta,

198: Quintum tricesimum diem

similia.

q Lib. i. 29.

r Lib. iv. 6. §. 7. 84. Cf. Dio, liii. 25.

s Lib. ii. 110.

Braccata in the latter, we perceive that the difference between them is accounted for. Gallia Comata, in point of extent, might be more than three fourths of the whole of Gaul; but the Provincia Romana, in point of population, contained perhaps one third of the inhabitants of Gaul. For the Provincia had never suffered from those destructive wars of Cæsar, by which one third of the rest of the inhabitants of Gaul had been cut off. Hence if the population of all Gaul was about five millions, the Provincia might contain nearly two millions of these, and the rest of the country the remainder. A census, then, which took in these last, as well as the former, would exceed one which comprehended only the former, by nearly three millions.

Again, among the other criteria for determining the amount of the population of Rome, the numbers of the plebs urbana, or of the duos, properly so called, would seem to be one, if those numbers could be ascertained with any thing like precision. Under this denomination, Diot includes the commonalty of Rome as such; that is, all the free population of the city, with the exception of the knights and the senators. Neither of these latter classes in particular was at any time so numerous as to make much difference in the total amount of Roman citizens, whether reckoned inclusively or exclusively of them. The number of senators, even when greatest, never exceeded 1000; and U.C. 736, was permanently reduced by Augustus to 600". And as to the amount of the equestrian order, though greater than that of the senatorian, yet it might be shewn from the accounts of the numbers of their

t lii. 28. 30. u Vide Plutarch, viii. 21, De Garrulitate: 1 Mace. viii. 15: Livy, lx: Cicero, Oratio post reditum ii. 10: Appian, B. C. i. 59. 100. ii. 30: Dio, lii. 42. liv. 13, 14. 17. 35: Suetonius, Augustus, 35. Cf. Aurelius Victor, De Vespasiano.

body, which perished in the proscriptions, at a time when the greatest part of them were cut off at once, that they probably did not exceed two or three thousand V.

The most effectual means of ascertaining the number of the citizens of Rome, as such, is the account of the several congiaria of various kinds, which were distributed to them at different times; in some instances of which the numbers who partook of them are actually specified, and in others may very probably be conjectured. I shall produce examples of these congiaria, not only during the reign of Augustus, but before and after it; from which it will appear that the number of those who were entitled to partake in such gratuities, preserves a remarkable uniformity through a period of two centuries and upwards.

Lucullus, on his return from Asia, U. C. 688, distributed among the people of Rome, 100,000 cadi of Chian wine w.

The cadus congiarius is considered by Arbuthnot an uncertain measure. But we may suppose it was nearly the same as the Attic xoûs; that is, it contained something more than six pints of our measure, or six Roman sextarii.

If we refer to the passages cited below, we shall conclude that two sextarii, or about a quart of our measure, would be no improbable allowance to each recipient on such an occasion as this*. If so, the

* In the Greek Anthology there is an epigram of Posidippus,

which, as it appears from the context, reckons three choës of wine

▾ Cf. Appian, B. C. i. 103. iv. 5. w Pliny, H. N. xiv. 17. xv. 30. Velleius Pat. ii. 33. x Cf. Eckhel, Doctrina Numorum Veterum, v. 5. Suidas, indeed, voce Xous has a gloss, that the Xoûs was equal to two sextarii or έéσrat, and the Xocus to six; in which case the xoeùs was the same with the Cadus. But in fact, as Kuster, in locum, observes, xoûs and xoeùs are the same thing. y Thucydides, iv. 16. vii. 87: Livy, vii. 37: Plutarch, Lycurgus, 12: Dicæarchus apud Athenæum, iv. 19: Horace, Sermonum i. i. 74: Juvenal, vi. 426, 427: Vopisci Tacitus, II.