[448] Considerable time was required for the establishment of the earliest known meaning of classis before the second and third divisions were added.
[449] This is a conjecture of Bruncke, in Philol. xl (1881). 362, favored by Delbrück, Gesch. d. Kriegsk. i. 222.
[450] P. 79, 86.
[451] Usually scholars (cf. Domazewski, in Pauly-Wissowa, Real-Encycl. iii. 1953 f.; Delbrück, Gesch. d. Kriegsk. i. 227; Smith, Röm. Timokr. 39) assume fifteen centuries for the fifth rating, on the authority of Livy i. 43. 7; Dion. Hal. iv. 17. 2; vii. 59. 5. But our knowledge of the phalanx is only inference, which to be acceptable must have at least the merit of possibility. The number fifteen is wrong because it could not have been divided evenly between the two legions; and on the other hand it will be shown later (p. 208) that in all probability the fifteenth century was not military but was added in the make up of the comitia centuriata.
[452] Müller, in Philol. xxxiv (1876). 129, is right in supposing that the legion was strengthened between the time of Servius and 387, but it was not in the way he assumes. The tradition of a legion (half phalanx) of 4000 men is preserved in Livy vi. 22. 8.
[453] Polyb. vi. 20.
[454] Cf. Smith, Röm. Timokr. 121 ff.
[455] Livy iv. 46. 1: “Dilectum haberi non ex toto passim populo placuit: decem tribus sorte ductae sunt. Ex his scriptos iuniores duo tribuni ad bellum duxere.” If this passage does not state a historical fact, at least it gives the idea of the writer as to the custom of earlier time.
[456] P. 72, 76.
[457] Cf. Smith, Röm. Timokr. 51 ff.