(3) I come to the last and most important point; indeed it was this which suggested the whole inquiry. On what principles, or by what evidence, did Hippias fix on the year 776 B.C. as his starting-point? We need not plunge into the arid and abstruse computations of years and cycles which make early chronology so difficult to follow and to appreciate. For one general consideration is here sufficient. Even had we not shown from Plutarch's words, and from the silence of all our authorities, that Hippias _could_ not have determined it by counting _upwards_ the exact number of duly recorded victories, it is perfectly certain that he _would_ not have followed this now accepted method. All the Greek chronologists down to Hippias' day (and long after) made it their chief object to derive historical families and states from mythical ancestors, and they did this by reasoning _downwards_ by generations. They assumed a fixed starting-point, either the siege of Troy, or the return of the Herakleids. From this the number of generations gave the number of years. Thus we may assume that Hippias sought to determine the date of the 1st Olympiad by King Iphitus, who had been assigned to the generation 100 Olympiads--a neat round-number--before himself. Hippias thus fixed the date of both Iphitus and Lycurgus. The Spartan chronologers would not accept such a date for Lycurgus. His place in the generations of Herakleids put him fully three generations earlier. Other chronologers therefore sought means to accommodate the matter, and counted twenty-eight nameless Olympiads from Lycurgus to Corœbus (and Iphitus). Others imagined two Iphiti, one of Lycurgus' and one of Corœbus' date. But all such schemes are to us idle; for we may feel certain that the number of Olympiads was accommodated to the date of Iphitus, and not the date of Iphitus to the number of Olympiads.