458 BUREAU OF AMERICAN ETHNOLOGY [bull. 28 



is the same from IV 19 to IV 17, so that there is the same interval 

 between the three days in the three numbers. The third number 

 is neither divisible by 91, 104, nor 260, and yet this is the very 

 number from which the number sought is to be obtained. However, 

 like the other two, it is at least divisible by 13, the number of week 

 days. 



Among the three large numbers the manuscript shows the now 

 familiar sign XIII 20. This means that those three numbers are 

 all to be reduced to the day XIII 20 by means of subtraction. Now, 

 the distance from XIII 20 to IV 1 is 121 ; from XIII 20 to IV 17, 

 17; from XIII 20 to IV 19, 199. The first two of these numbers are 

 directly subtracted, but the third, as is often done, is first increased 

 by a multiple of 200, which produces no alteration in the position of 

 the days. Here 197X260+199=51,419 is subtracted. These three 

 numbers, 121, 17, and 51,419, the last being in accordance with the 

 correction which I gave above, are actually provided in the manu- 

 script with the red ring, which indicates the subtrahend, and there- 

 fore stands for the minus sign with the Maya. 



By this subtraction the three following numbers are obtained: 



1. 1,272,423; that is, day XIII 20, sixteenth clay in the first luonth, year 

 .12 Muluc. 



2. 1,268,523; that is, day XIII 20, eleventh day in the seventh month, 

 year 1 Ix. 



3. 1,486,923; that is. day XIII 20, first day in the fourteenth month, 

 year 1 Kan. This day, therefore, divides the year, as was previously 

 pointed out, into a tontlamatl of 260 days and a period of 104 days. 



These numbers are not in the manuscript, but as usual in such cases 

 they must be calculated by the reader. W]\y were not 260 days less 

 deducted to obtain in this w^y the beginning of a katun, the first day 

 of the first month in the year 1 Kan ? I believe this was omitted in 

 order to avoid the unlucky new year's day. I am confirmed in this 

 opinion by the fact that the same date, 1, fourteenth month, com- 

 puted to be sure from IX 1 and in a different katun, also results from 

 the black numbers of the fourth serpent, on page 02. 



The three numbers found by computation now stand in a much 

 clearer relation to one another than those set down in the manuscript. 



1. The difference between the first and the second number is 

 3,900=15X260. 



That this difference is intentional is confirmed by the number 39,000 

 resulting from the two numbers in the serpent on page 69, which are 

 nearly ten times as large as those mentioned here. There the two 

 numbers are 12,381,728 and 12,391,470, from w4iich must be subtracted 

 the differences on page 73, 34,732 and 83,474, and the resulting re- 

 mainders are 12,346,996 and 12,307,996, whose difference is exactly 

 39,000. 



