BABYLONU AND EGYPT 21 



verify these, on the one hand, by archseological evidence, on the 

 other, by present-day observations of backward races still in 

 their prehistoric stage. 



PRIMITIVE ASTRONOMICAL NOTIONS. On the astronomical side 

 the most obvious fact is the division of time into periods of light and 

 darkness by the apparent revolution of the sun about the earth. 

 With closer attention it must soon have been observed that the rela- 

 tive length of day and night gradually changes, and that this 

 change is attended by a wide range of remarkable phenomena. At 

 the time of shortest days, vegetable and animal life (in the north 

 temperate zone) is checked by severe cold. With the gradually 

 lengthening days, however, snow and ice sooner or later disap- 

 pear, vegetation is revived, birds return from the warmer south, 

 all nature is quickened. In the symbolism of the beautiful old 

 myth, the sleeping princess, our earth, is aroused by the kiss of the 

 sun-prince. The longest days and those which succeed them 

 are a period of excessive heat and of luxuriant vegetation, followed 

 by harvests as the days shorten, towards the completion of the 

 great annual cycle. In time, closer observers, noting the stars, dis- 

 covered that corresponding with this great periodic change are 

 gradual variations in the starry hemisphere visible at night, that 

 in other words the sun's place among the stars is progressively 

 changing, that it is in fact describing a path completed in a large 

 number of days, which after repeated counting is found to be 365. 

 It is also found that the midday height of the sun above the 

 southern horizon shares in the annual cycle. The determination 

 of the number of days in the year is a matter of very gradual ap- 

 proximation, possible only to men who have already attained 

 some command of numbers and the habit of preserving records 

 extending over a long series of years. For there is no well-marked 

 beginning of the year as of the day. An erroneous determination 

 of the number of days becomes apparent only after a number of 

 years, increasing with the accuracy of the original approximation. 

 If, for example, the year is assumed to be exactly 365 days, that 

 is, about six hours too short, the festivals and other dates will slip 

 back about 24 days in a century, and thus lose their original cor- 



