On the whole we may agree with him that the rite was in its origin one of the type to which he has referred it—a final harvest festival of the Latin farm. There is yet, however, a word to be said. He does not treat it from the point of view of the Roman calendar, and thus fails to note the turn it took when Latin farmers became Roman citizens. Wissowa, on the other hand, takes the calendar as his sole basis for judging of it, and with a strange perversity, as it seems to me, brushes Mannhardt’s conclusions aside, and would explain the rite simply as a sacrifice to the god of war[1085]. Now doubtless it had come to be this in the organized city-calendar, as Mars himself began to be brought into prominence in a new light, as the _iuvenes_ of the community came to be more and more employed in war as well as agriculture, and as the Campus Martius came to be used as an exercising-ground for the armed host. The Calendars show us a curious correspondence between the beginning and the end of the season of arms, i. e. the middle of March and the middle of October, which leaves little doubt of the change which had taken place in the accepted character of the rites of the two periods by the time the Numan calendar was drawn up. This correspondence has already been noted[1086]; it may be here briefly referred to again. On March 14[1087] there was a horse-race in the Campus Martius; on the 19th (Quinquatrus) was the _lustratio armorum_ for the coming war-season, as is seen from the fact that the _ancilia_ of the Salii at least—if not all arms—were _lustrata_ on that day[1088]. So too on October 15 there was a horse-race, as we have seen, in the Campus Martius, and on the 19th we find the Armilustrium in the oldest calendars[1089], a name which tells its own tale. The inference is that the horse-races on Oct. 15 and March 14 had much the same origin, and it is just this which induces Wissowa to slight Mannhardt’s explanation of the former. He thinks that on each day the horses, like the arms, were lustrated (p. x.), i. e. before the war-season began, and after it was over. This is likely enough; but might not the same have been the case with the horses of the farm? The Roman farmer’s year began with March, and the heavy work of carrying, &c., would be over in October. I am disposed to think that we must look on organized war-material as a development later than the primitive times to which Mannhardt would carry us back, a side of Roman life which only in course of time became highly specialized.