478 BUREAU OF AMERICAN ETHNOLOGY [bull. 28 



4a, 5a, and 6a we always encounter a serpent with the 18 inscribed 

 within its ring, so that these four leaves readily suggest the four 

 kinds of years. 



So, too, I believe I have found a very perfect picture of the change 

 of years in Dresden codex, ]:>age 68, above on the left, in the two fig- 

 ures of gods leaning back to back and sitting on a series of astronomic 

 signs, arranged almost like the roof and wall of a house. But this 

 picture belongs to a large section, which begins on page 6.5 and ends 

 on the left side of page 69. I must here dwell more particularly on 

 this section than I could in my Erlauterungen (Dresden, 1886). 



The real nucleus of this section consists of four rows of 91 days 

 each, that is, of a year, of which the detailed explanation is found in 

 six rows of glyphs and twenty-six pictures. Now, believing that I 

 can complete the toj:) row, which is almost wholly obliterated, from 

 the still existing renniants, I read these four rows as follows : 



XII. .5 IV, 1 V, 10 II, 6 VIII, 2 X. 11 VIII. 7 II, .3 V, 12 IV, 8 XII, 4 

 III, 13 III. 



11 I, 13 I, 11 XII, 1 XIII, 8 VIII, 6 I, 4 V. 2 VII, 13 VII. 6 XIII, VI, 

 8 I, 2 III. 



11 XI, 13 XI, 11 IX, 1 X, 8 V, 6 XI, 4 II, 2 IV, 13 IV, 6 X, 6 III, 8 XI, 

 2 XIII. 



9 IX, 5 I, 1 II, 10 XII, 6 V, 2 VII, 11 V, 7 XII, 3 II, 12 I, 8 IX, 

 4 XIII, 13 XIII. 



The stud}" of these four rows shows that the end of each one of them 

 can again be very well joined to its own beginning, and also that a 

 good connection occurs between the end of the fourth and the be- 

 ginning of the third, and likewise between the end of the second and 

 the beginning of the first, also vice versa between the end of the third 

 and the beginning of the fourth, and between the end of the first and 

 the beginning of the second. But the second and third rows, on the 

 contrary, stand in no such connection. 



We further see that the final point of the first two rows is a day III, 

 that of the last two a day XIII. AMiat is more natural than to think 

 of the two days III 2 and XIII 20, which are of such great impor- 

 tance on pages 62 to 64? Our section, pages 65 to 69, then appears 

 like an introduction to pages 62 to 64 and one part of our manu- 

 script is again made to harmonize with another. 



Each row is, as we see, divided into 13 periods of time, whose 

 average duration is 7 days; the four rows therefore form 52 periods 

 of time. Now, we find 26 pictures on these pages; the half of these 

 periods of time is apparently without a picture. Thirteen of the 

 pictures are between the second and third row and 13 below the 

 fourth, but this probably has reference only to the symmetric arrange- 

 ment of the pages. 



