142 BUREAU OF AMERICAN ETHNOLOGY [bull. 28 



" 3 king vulture ''. This latter day occurs 00 days before the day 

 ce Coatl, the day of Quauhtemoc's capture. 



Now, if the day of Quauhteuioc's capture was August 13, 1521, 

 the third day of the mouth Xocotluetzi, it follows, as this was said 

 to have been likewise a day ce Coatl, " 1 snake ", that the first day of 

 the month must have been the day 12 Calli and the first day of the 

 year 1 Calli. Hence it follows, as I stated above, and as can safely 

 be concluded from the dates in our manuscript, that the years of the 

 Mexicans began with the signs Acatl, Tecpatl, Calli, and Tochtli, 

 and not, as was hitherto generally supposed, with the signs Cipactli, 

 Miquiztli, Ozomatli, and Cozcaquauhtli ; and it follows, since the year 

 1521 is said to have been a year 3 Calli, that the years of the Mexicans 

 were not named for the first daA^ of the first month, Atlcaualco, as 

 has been commonly believed, but, as the computation shows, for the 

 first day of the fifth month, on whose last day the feast Toxcatl was 

 celebrated; lastly, it follows that the beginning of the month Atlcau- 

 alco in the year of the conquest did not fall on the 2d of Febriuiry, as 

 was decided after much discussion at the Indian conference held at 

 Tlatelalco in Sahagun's time,** but that it must have fallen on the 12th 

 of February. The latter result is of special importance because it 

 proves that in the forty odd years which elapsed between the year of 

 the conquest and the time when the Sahagun manuscript was com- 

 posed ^ the beginning of the Mexican year was set forward 10 days. 

 This is exactly the sum of the intercalary days, which occur in this 

 period of time, and proves that the Mexicans did not know how to 

 regulate their chronology by intercalations at short intervals. 



If this is firmly established, then we may further conclude that the 

 day of the arrival of the Spaniards, said to have been the ninth day 

 of the month Quecholli, can have been neither 8 Ehecatl (as Chimalpa- 

 hin states) nor 1 Ehecatl (as the writer of the account in the Sahagun 

 manuscript asserts), but must have been the day before 7 Cipactli 

 or 13 Cipactli. Otherwise, the month must hcLve begun with a day 

 Ocelotl, which, as we have seen, is incorrect. But if from 1 Coatl, 

 the day of Quauhtemoc's capture, we count 644 days backward in the 

 Indian calendar we do not arrive at 1 Cipactli, but at 7 Cipactli. 

 Chimalpahin's statement was, therefore, relatively correct (within 

 1 day), and the writer of the account in the Sahagun manuscript 

 made an error of 20 days. The only explanation I can give for the 

 fact that both sources agree in mentioning a day Ehecatl instead of a 

 day Cipactli is that tradition confused the day and its eve or that the 

 name of the day was not held fast by tradition, but was only recov- 



" See Sahagun, v. 7, chap. 12. 



" In tLe Sahagun manuscript of the Academia de la Historia the year orae Acatl ( = 

 A. D. 1559) is given as the year of writing down at least certain parts (the historical 

 ones) of the manuscript. 



