lORSTKMANN.] CENTRAL AMERICAN TONALAMATL 529 



rated by 5, 12, and 6, but relatively by 65, 52, and 26, since the week 

 day indicated above them in always included. However, I have 

 explained this somewhat at length in my Erlauterungen. 



The three kinds mentioned include the entire number of tonala- 

 niatls contained in the manuscripts, with the exception of a few 

 anomalous examples, and it is quite worth while to learn in what pro- 

 portion the three kinds occur in the two manuscripts. 



Dresden Troano-Cortesianns 



Tonalaniatl of four parts 13 44 



Tonalamatl of five parts 43 132 



Toualamatl of ten parts 8 40 



63 316 



Both manuscripts, otherwise differing so greatly from one another, 

 agree in this, that the division is by far the most frequent into peri- 

 ods of 52 days, into those of 65 days less so, and into those of 26 days 

 least frequent of all. Indeed, the ratio of the tonalamatls of five 

 parts to the entire number is surprisingly alike in both manuscripts : 

 in the Dresden codex, 1 to 1.5; in Codex Troano-Cortesianus, 1 to 1.6. 

 It is more a matter of chance with the other two kinds, OAving to the 

 smallness of the numbers; nevertheless the figures expressing the 

 ratio of the periods of four parts do not differ very greatly: in the 

 Dresden codex, 1 : 5.2 ; in Codex Troano-Cortesianus, 1 : 4.9. Who 

 will be the one to discover the reason for this wonderful similarity? 



But there are still other remarkable coincidences observed. "\Miile 

 we have just seen that the division of only the first quarter, fifth, or 

 tenth of the tonalamatl is carried out in detail, and it is left to the 

 reader to apply this arrangement to the other sections, in isolated 

 cases a tonalamatl of four parts (and only such a one) shows uniform 

 treatment throughout. The Dresden codex offers three examples of 

 this: 



1. On each of the four pages 31b to 34b 46 days are separated into 

 periods of 9, 9, 9, 2, 4, 9, and 4 days, and 19 days are designated as the 

 distance of each one of these groups from the next; thus, 260=4 

 (19-f46). 



2. On pages 33c to 39c the division into 9, 11, 20, 10, and 15—65 

 days occurs four times in succession with great uniformity of detail; 

 thus, 260=4 (9+11+20+10+15). 



3. On pages 42c to 45c (the end of the first division) four repeti- 

 tions of 17+6X8=65 days give rise in each case to a special row of 

 glyphs and a special representation; thus, 260=4 (17+6X8)- 



I can quote two examples from Codex Troano-Cortesianus, which 

 correspond perfectly : 



1. In Codex Cortesianus, pages 13b to 18b, four horizontal rows, 

 each of 52 days, follow each other in close succession, the last being 

 7238— No. 28—05 34 



