304 to 220 would best explain all the known facts.[1281] As no description of the reformed organization has come down to us, we are obliged to reconstruct it from the scant references of various writers. It is to be noted first that the five classes continued in the new system.[1282] They were still based on the census,[1283] and were called to vote in their order as before.[1284] The distinction between juniors and seniors was retained;[1285] and as these comitia were still called centuriata, the centuries necessarily continued as the voting units.[1286] But the reform brought them into direct relation with the tribes, which now served as a basis for the division into centuries and for their distribution according to age and class. On this point Livy[1287] remarks, “We ought not to wonder that the arrangement which now exists after the tribes have been increased to thirty-five, their number being doubled in the centuries of juniors and seniors, does not agree with the total number instituted by Servius Tullius; for he divided the city into four parts, ... which he called tribes.... Nor did those tribes have any relation to the distribution and number of the centuries.” From this passage we may infer (1) that in the reformed assembly the number and distribution of the centuries depended closely upon the tribes—a conclusion supported by other citations to be given hereafter, (2) that the number of centuries was changed, although we are not distinctly informed whether by diminution or increase. According to one interpretation the number of tribes was doubled by the number of centuries of juniors and seniors, and there were therefore seventy of these centuries, thirty-five juniors and as many seniors, each century forming a half tribe. This view is supported by passages in which the century bears the name of the tribe, as Aniensis iuniorum,[1288] Voturia iuniorum,[1289] Galeria iuniorum,[1290] as well as by those which in a more general way refer to voting or the announcement of the votes by or according to tribes in the centuriate assembly.[1291] It accords perfectly with other evidence that the century was an integral part of the tribe.[1292] This is the view adopted by Niebuhr.[1293] It is open, however, to the fatal objection of abolishing the classes, which in fact continued through the republic, as has already been shown.[1294] He does indeed allow for a first class comprising the country tribes and a second class made up of the others;[1295] but this hypothesis is overthrown by those citations which imply the continuance of all five classes,[1296] as well as by those which make the census an element of the later organization.[1297] Huschke,[1298] who places the reform in the earliest times of the republic, adopts Niebuhr’s view as to the number of centuries; but maintaining the continuance of the five classes,[1299] he considers them to be groups of tribes, the seventeen old rural tribes