32 BUREAU OF AMP:RICAN ethnology [bull. 28 



glyphs; also by the fact that, in coiijuiiction with the element '"tree"''', 

 it is used to denote honey and honey wine (// and o, tigure l>), and 

 that it a[)pears vicariousl}' for kin. ''sun", and is tsometinies replaced 

 l)}^ the hieroglyphic expression for th(> latter. According to this, 

 indeed, we should have the four colors, yellow, red, white, and black, 

 in t to w, tigure 1, and in the same order of succession as they are 

 given by Landa for the four cardinal points. 



But these elements, which I call kan, chac, zac, and ek, are not, in the 

 above-mentioned passages, as we should suppose, assigned to the east, 

 north, west, and south, but, in the same way as Landa — though, as 

 we must assume, incorrectly — refers the variously colored Bacabs and 

 their years to the cardinal points, the}^ are assigned to the south, east, 

 north, and west. I must confess that this fact disturbed me for a long 

 time, until it gradually became clear to me that in this instance other 

 ideas were decisive in referring the rain god, Chac, to the cardinal points, 

 and hence other colors were necessarily chosen to express that refer- 

 ence than those chosen for the Bacabs prevailing in the different years. 

 Wherever the Bacabs themselves and the difl'erent years and the cere- 

 monies performed before the beginning thereof are represented in the 

 Dresden manuscript, especiall}" on the familiar pages 25 to 28, there the 

 elements of iigures f to ui are not coordinated with d, a, 5, c, but with 

 rt, h, e, d (tigure 1) — that is, actually with the east, north, west, and south. 

 This can not, indeed, be noted on all four pages, the upper parts of 25 and 

 27 being unfortunately too far destro3'ed. But we can still see that 

 on all four pages in a certain place on the upper part there was a per- 

 vading hieroglyph, which contained the elements of t to in as a varia- 

 ble constituent part. The same is retained on two pages, 2<) and 28 

 (see r and *•, tigure 1), and there we actually see that the elements of u 

 and v^ — that is, as I assume, red (chac) and black (ek) — are allotted to 

 the north and south. That yellow (kan, t) and white (zac, v) are also 

 correspondinglv arranged is, I think, as good as certain. And these 

 assumptions are confirmed by corresponding passages in the Troano 

 codex. There the various Chacs are represented, pages 30 and 29^, 

 beginning with that of the west, c. And the elements ek, kan, chac, 

 zac answer to the directions of c, d^ a, h. On pages 31 and 30c7, on 

 the contrary, the various Bacabs are represented, beginning with that of 

 the east (chac and hobnil). And here, as comparison with the Cortes 

 codex, pages 41 and 42, show the elements kan, ek, zac, chac correspond 

 to the directions of a, d, c, h — that is, east, south, west, north. Thus, 

 that which I think I have discovered in regard to color nomenclature 

 agrees with the old Schultz-Sellack idea that a to d represent hiero- 

 gl3'phically the cardinal points — east, north, west, south, or likin, 

 xaman, chikin, nohol. 



Now if we turn with this, as 1 believe, (-ertain knowledge to pages 

 25 to 28 of the Dresden manuscript, on which the various yea»'s are rep- 



