FORSTEMANN.] CENTRAL AMERICAN TONALAMATL 531 



Contrariwise, 26, 52, and 65 are never divided, respectively, into 

 13 sections of 2, 4, and 5 days; that has unquestionably been avoided. 



Thus it is doubtless intentional, not accidental, that these three 

 periods are often divided into the greatest number of equal parts, to 

 which one or two more unequal parts are added or between which 

 they are inserted in order to complete the sum. I here give the cases 

 which have come to my knowledge: 



1. 26=4x4+10 (Troauo codex, page 25*f) =4x0+0 (Troano codex, 

 page 28*c) =3X7+5 (Dresden codex, page 21b, also Troano codex, page 

 23*d). 



2. 52=4x<!+28 (Troano codex, page 29*a)=8X<>+4 (Troauo codex, 

 page 15*c) =5x8+7+5 (Troano codex, page 24*d) =5x9+7 (Dresden 

 codex, page 8c, and Troano codex, page 31*e) =4x10+3+9 (Dresden 

 codex, page 40c) =4x10+12 (Troano codex, page 8c) =3x11 + 10+9 

 (Dresden codex, page 19c) =4x11+8 (Troano codex, page 31b) =4+6x8 

 (Ti'oauo codex, page 23*b). 



3. 65=6X10+5 (Troano codex, page 35a) =5x12+5 (Dresden codex, 

 page 23b) =3x16+17 (Cortesian codex, page 20d). 



The varieties of intentional regularity are entirely exhausted by 

 these examples, and I should waste space if I were to cite more. I 

 will only add one from the Dresden codex, pages 4a to 10a, where the 

 period of 52 days is divided into not fewer than 20 parts of from 1 

 to 4 da^'s each without any intelligible order. All these 20 parts have 

 a common superscription, consisting of two glyphs. And, besides, 

 each part has belonging to it the picture of a god and a glyph closely 

 connected Avith the latter. I have given a thorough study to this one 

 tonalamatl and have reall}' found nnich that is curious, which, how- 

 ever, is not yet readj^ for publication. 



Let us noAv attempt to approach these tonalamatls from a third 

 side, proceeding from the initial days. If the arrangement here were 

 left to chance, we should, on an average, find each of the so-called 

 month days in one-twentieth and each of the week days in one- 

 thirteenth of all the cases. But this does not accord with the actual 

 facts in two points in which the two manuscripts agree with each 

 other in a veiy remarkable manner. 



(1) Among the month days both give decided prominence to the 

 seventeenth day (Ahau, " lord "), which was by far the most exalted 

 day, and the one most in use among the Mayas and also the beginning 

 of their entire computation of time. Ahau stands 14 times at the 

 beginning of the tonalamatl in the Dresden codex and 50 times in 

 Codex Troano-Cortesianus, thus in betAveen a fourth and a fifth in- 

 stead of in a twentieth of all the cases. 



(2) Among the week days, the first and tlie last, I and XIII, were 

 greatly preferred. They appear in the Dresden codex 9 and 11 

 times; in Codex Troano-Cortesianus 27 and 25 times, respectively, 



