110 MEXICO AND CENTRAL AMERICA 



The true length of the year was probably obtained by 

 observations of sunrise or sunset on summer or winter 

 solstices. From some fixed point of observation, such 

 as the doorway of a temple, the extreme point on the 

 horizon reached by the sun in its northward or south- 

 ward march could be accurately determined. Over a 

 period of years the average solstitial period (tropical 

 year) could be readily obtained if only the days were 

 recorded and the intervals compared. 



Although we ourselves depend mostly upon the year 

 count rather than the day count we must remember 

 that the annual calendar was only one of several that 

 the Mayas brought into relation to the inviolable count 

 of days. The lunar and Venus calendars will be con- 

 sidered presently. 



The Lunar Calendar. The revolution of the moon 

 around the earth was used by the Mayas in what may be 

 called the lunar calendar. It has already been ex- 

 plained that an early lunar period of thirty days seems 

 to have been arbitrarily changed to a notational one of 

 twenty days. Now the exact duration of a lunar revo- 

 lution is 29 days, 12 hours, 44 minutes, 2.87 seconds. If 

 the customary period of 29.5 days is taken for con- 

 venience there is an error of about two full days in five 

 years. Such an error was too great to pass the Mayan 

 calendar makers. On pages 51 to 58 of the Dresden 

 Codex their solution is recorded unmistakably. A suc- 

 cession of 405 lunar revolutions, or nearly 33 years, is 

 calculated by the addition of groups of five and six revo- 

 lutions, the former given as 148 days and the latter as 

 either 177 or 178 days. This method of calculation 

 may have been a device to carry fractions or it may have 

 been based upon ecliptic data, probably the latter. 

 The steps of the calculations are put down in a sort of 



