are several times referred to in extant orations. While they could be used at any time of the day or night, they required constant attention, and were by no means accurate. Generally the sun and the stars were depended on when they could be seen; for in the climate of Greece and the adjoining lands there are fewer cloudy days and nights than in the more northerly regions. In modern Athens about one half the days of the year are entirely cloudless, and only thirty are noted as cloudy. The Greeks used daylight almost entirely for business and rose very early. A decree of Solon is often referred to which forbids teachers to open school before daylight. For longer divisions of time the Greeks, like most of the people of antiquity, depended on the moon, but they never got the lunar months to correspond exactly with the facts. They reckoned the month at twenty-nine and a half days, or one twenty-nine, the next thirty. Their months, however, were not divided like ours and the method of counting them so as to make them correspond with the year was very complex, and the result unsatisfactory; there had to be frequent corrections to make the seasons come at the same time of the year. Yet nowhere in Greece was there ever discovered any way to obviate the inherent defect of their clumsy system. In different parts the months had different names, but were not divided like ours. There is a passage in the "Clouds" of Aristophanes in which the moon is represented as complaining of ill treatment because the Athenians had allowed their calendar to fall into confusion to such an extent that the gods were disappointed in their feasts. This made them angry with the moon—very unjustly, since the confusion in their reckoning was the people's fault. The case is very much as if we allowed our fourth of July to drift about until it ultimately came in cold weather. The lack of a fixed date for determining events gradually became generally recognized; consequently, as is generally supposed, Timæus, a Sicilian Greek, proposed the Olympiads as an era. The Olympiads, however, do not correspond with the era employed in Christian countries. Hence we have to use a rule like the following: "Multiply the complete Olympiads by four, and deduct the total from 776 for events of the autumn and winter, or from 775 for events of spring and summer." Although Timæus flourished as late as 300 b.c., earlier dates were made to correspond to his method of reckoning as well as it could be done. It is probable that much of the older chronology is erroneous. By means of observations taken on the star Sirius, both in Egypt and Babylon as early as the fourteenth prechristian century, the year was' found to be about 36514 days in length. Those old-time astronomers also reckoned by a lunar year of twelve months of 29 and 30 days alternately. This was merely a concession to custom. The moon is such a convenience for measuring periods longer than a day and shorter than a year that the incongruity between its phases and the sun's motions
