THOMAS] MAYA METHOD OF CALCULATION 287 



111 order to show that I'esort was had to increasing tlie added num- 

 ber to shorten the iirocess, as was done in tlie theoretic example, the 

 following example is given from plates LXX and Lxxi of the Dresden 

 codex. Ordinary numerals are used in place of the symbols, and the 

 series, which in the codex ascends from right to left, is reversed; the 

 days below the columns are also given : 



9 Oc 9 Eb 9 Ix 9 Ix 9 Ix 9 Ix 9 Ix 



It will be seen by subtracting that the difference between the first 

 and second columns and between the second and third is 1-17-2, or 

 1 ahau, 17 chuens, 2 days, ecjual to 702 days, while the difference 

 between the 3d and 4th columns is 2-18-ft-O, or 2 katuns, 18 ahaus, 

 9 chuens, days — equal to 21,060 days; and that the difference 

 between the 4th and oth, the 5th and Gth, and the 6th and 7th is, in 

 each case, 1-19-0-0, or 14,040 days. There is therefore an increase 

 of the added number or factor in passing from the 3d to the 4th 

 column. 



It will be noticed that the days below the 1st, 2d, and 3d columns 

 differ, while from this point onward they are all Ix. The change in 

 this respect requires a change in the counter. Why the counter was 

 made larger in passing from the 3d to the 4th column than between 

 the remaining columns is not clear, as the difference between the 

 3d and 4th columns would have reached the desired day, 9 Ix. It is 

 possible that the month date, though it does not appear, was here 

 taken into consideration. Assuming that the first 9 Ix (under the 

 3d column) was 9 Ix 2 Pop, year 8 Ben, the count forward of 1-19-0-0 

 would reach 9 Ix 12 Chen in the j-ear 7 Akbal, while the count for- 

 ward of 2-18-9-0 would reach 9 Ix 17 Mac, year 13 Ezanab. As the 

 first counter (702) is not a multiple of 260 or of 20, it must have been 

 ba.sed on 13, one of the factors of 260. The counters 14040 and 

 21000 are multiples of 260; and there is possibly something in the 

 fact that the former (14040) is 54 times 260 and that the first counter 

 (702) is 54 times 13. Although we are not al)le at present to solve all 

 thes(? problems, it is evident that the author of the codex increased 

 tlie counter as he proceeded, presumably to shorten the process. 



The series appears to close with two columns in the upper middle 

 I^ortion of plate LXX, the dates here having the month given. With 

 these (notwithstanding the obliterated portion of the series) we might 

 determine the true dates of the portion given above, and thus possibly 

 solve, to some extent, the problems mentioned; but unfortunately 

 there are so many errors in these two columns that it seems impossi- 

 ble to determine the true numliers and dates. Thechuen number, or 



