[23] Mommsen, op. cit., p. 75; Liebenam in Pauly-Wiss., VI, 1605. We must remember that this number was somewhat low; and was gradually raised by succeeding emperors. Claudius added two legions, Nero one, and Galba two, so that Vespasian had thirty, and that number seems to have been maintained until the time of Septimius Severus, who added three more. It is significant that Trajan found 30 legions quite sufficient for extensive and difficult conquests, so that 25 would doubtless have been regarded even by him as adequate for the conquest of Germany. For the evidence of the gradual increase in the army see Marquardt, op. cit., p. 448 ff.
[24] The evidence for the size of the legion at this time is conveniently summarized by R. Cagnat, “Legio,” Daremberg et Saglio, III, p. 1050 f. The most elaborate discussion of the size of the legion (especially that of Caesar) is in Fr. Stolle, Lager und Heer der Römer, 1912, pp. 1-23. He finds what he regards as evidence for legions of varying size, from 3600 up to 5000 men. The standard legion of the empire, however, can hardly have been less than 6000. Cf. also Fröhlich’s review of Stolle’s work, Berl. Philol. Wochenschr., 1913, 530 ff.
[25] A list of these is given by Liebenam, op. cit., p. 1607 ff.
[26] Calculations as to the effective strength of the standing army of Augustus vary somewhat. H. Furneaux (The Annals of Tacitus, I (1884), p. 109), gives 350,000; Mommsen (Hermes, XIX (1884), p. 4—apparently excluding the naval forces), 300,000 as a maximum figure; Seeck (Rh. Mus., XLVIII (1893), p. 618) reckons on the basis of 20 legions (which would be applicable only down to the year 6 A. D.) 132,000 citizen soldiery out of Italy: in his Gesch. des Untergangs d. ant. Welt, 3rd ed., I (1910), p. 255, on a basis of 25 to 30 legions, from Augustus to Diocletian, he calculates the total forces of the empire at 300,000 to 350,000; H. Delbrück (Gesch. d. Kriegskunst, II, 2nd ed. (1904), p. 174) counting only the 25 legions, estimates 225,000 men; if other contingents be included the total would certainly exceed 250,000 even on the basis of his extremely low estimates; Gardthausen (Augustus, I, p. 635) estimates 250,000-300,000. The figure 200,000 which he gives on p. 637 seems to refer to the conditions before 6 A. D., when only 18 legions were maintained. G. Boissier’s number, 500,000 (L’opposition sous les Césars, 3rd ed. 1892, p. 4), seems to count the auxilia three times, once in making up the number 250,000 for the legions, and again in doubling that!
[27] For historical parallels to this condition compare Miss Ellen Semple, Influences of Geographic Environment, New York, 1911, p. 215 ff.
[28] E. M. Arndt, Zeitschr. f. Geschichtswissenschaft, III (1845), p. 244, calculated a population of 800-1000 per (German) square mile, but only then on the assumption, which no man would now accept, that the Roman reports about the primitive conditions of agriculture were incorrect. On this estimate the population of Germany between the Rhine, Elbe, and the Main-Saale line, which is the part generally considered in the question of conquest, would have been roughly 1,840,000 to 2,300,000. H. Von Sybel (Entstehung d. deutschen Königtums, 1881, p. 80) estimates the Germans at 12,000,000, basing his calculation on a highly problematic series of inferences regarding the extent of territory which the Sugambri once occupied, 40,000 of whom were said to have been transferred to the west bank of the Rhine by Tiberius. Karl Lamprecht, Deutsche Geschichte (1894), I, p. 236 accepts the traditional statement that the Goths alone amounted to five-sixths of a million, a reckoning which would make the total population of Germany many times that number. Even G. Waitz, Deutsche Verfassungsgeschichte (1880), I, p. 19, takes at their face value such Roman exaggerations as 300,000 warriors for the Cimbri and Teutones, 60,000 for the Bructeri, and the like, figures which presuppose an incredibly dense population.
[29] Histoire des Institutions politiques de l’ancienne France, I (1875), “L’Invasion Germanique,” p. 310 ff. For Delbrück’s results see the chapter entitled “Zahlen,” Gesch. d. Kriegskunst, II, 2nd ed. (1909), p. 294 ff. On the actual number of the Vandals and their allies, a cardinal point in the discussion, compare H. Delbrück, Preuss. Jahrb., 81 (1895), 475 f. O. Seeck (Jahrb. für Nationalökonomie u. Statistik, III, 13 (1897), p. 173 ff.) argued unsuccessfully for the older view, but Delbrück (Gesch. d. Kriegskunst, II, 2nd ed. (1909), p. 308 f.) has completely settled this specific question.
[30] “Observations sur l’état et le nombre des populations germaniques dans la seconde moitié du IVe siècle, d’après Ammien Marcellin,” Mélanges Cagnat, Paris, 1912, pp. 247-267.
[31] “Der urgermanische Gau und Staat,” Preussische Jahrbücher, 81, (1895), p. 471 ff. The main arguments here presented (except the detailed criticism and comparison of a number of ancient estimates, p. 474 ff.) are repeated with some slight modifications in his Gesch. d. Kriegskunst, II, 2nd ed. (1909), p. 12 ff. L. Schmidt, Gesch. der deutschen Stämme, I (1904), p. 48, accepts Delbrück’s calculations indeed, though with some reserve; p. 46 f. he criticizes effectively the absurd exaggerations with which the pages of many ancient authors abound.
[32] Preuss. Jahrb., p. 482.