4:76 BUREAU OP AMERICAN ETHNOLOGY [bull. 28 



at 365 whole days. In earlier times such confusion was perhaps not 

 possible, because the chronology was probably not based then on 

 the solar year, but on the period of 260 days, the tonalamatl, pos- 

 sibly also on a period of 400 (20X^0) days. To obviate this con- 

 fusion I think they did what has been done under similar circum- 

 stances by other peoples; that is, they intercalated 17 days; and, 

 instead of Imix, which had hitherto begun the series of days, Kan, 

 which had already passed, was reintroduced with the days which fol- 

 lowed it. Traces are found in Codex Troano-Cortesianus of this 

 older arrangement, for instance, in Cortesian codex on page 31a, and 

 in Troano codex on page 31, whether this is older than the Dresden 

 codex (which my correspondents will not admit), or has been copied 

 from an older manuscript, or was })roduced in some other region 

 which still preserved the Aztec arrangement. But Landa, who un- 

 questionably spoke of his own time, is thoroughly trustworthy when 

 he gives Kan as the first day, especially as the Dresden codex gives 

 precedence to that day. I need only recall the eight highest figures in 

 this manuscript, those in the serpents on pages 61 and ()2, which are 

 all counted from a day Kan. In this way I explain number 1. 



Numlier 2 may also be very simply explained. Before the cor- 

 rection of the calendar that eighth clay of the eighteenth month, 

 from which all computation of time proceeded, was the twenty- 

 fifth; that is, the last day of the eighteenth month, and therefore 

 of the whole year. At least this was the case every four years. The 

 Maj^as therefore reckoned how many days had elapsed since this day 

 as the zero i)oint. The years which followed a year closing with 

 Ahau quite properly began with Imix, the first day of the series ; the 

 others, with Cimi, Chuen, and Cib (according to my notation 3, 

 8, 13). It would be interesting if we could discover anything to 

 indicate that these three days had once been of especial imjxjrtance 

 (see, for instance, Codex Cortesianus, pages 13b to 18b, where four 

 rows of 52 successive days begin with these very four days, each row 

 with one of them). 



New light is now also thrown on number 3. From this starting 

 point of all chronology, this last day of the year beginning with Cib, 

 the period of 24 years then beginning (which was also the period of 

 15 apparent Venus years) was always computed. The fourth ahau, 

 for instance, began with the year 5 Imix, and each ahau in the same 

 Avay with this first day until everything was displaced by the intro- 

 duction of the 17 days. It looks like a modification of this abrupt 

 change that in the j)lace of Imix, " maize bread ", its synonym, Kan, 

 " maize kernel ", was used, the two glyphs occurring countless times 

 closely connected in the manuscript. 



While the first three points are thus explained by my theory of a 

 correction in the calendar, the other three may be explained by an 



