554 THE INSCRIPTIONS AT COPAN. 
forms, has three extremely constant elements: (1) the elbow element with crossed 
bands at the angles E 6 Sy (2) the small oval element always < 5EjTe > to be found 
somewhere within ithe angle formed by the eee [Sind (3) an 
ending prefix or UWsuperfix, which takes several forms. aire little animal 
head usually found in the angle of the elbow clement appears in at 
least one instance (Stela 11 at Piedras Negras) emerg- ing from Glyph 
=] and on the basis of this association, as 
well as the fact a Hat it always has an{‘y4 ending-sign, and finally because of its 
position, always immediately preced- (ing Glyph A, the writer has suggested 
for it the general meaning: ‘‘here ends the count of the moon’’ or “next follows 
the current lunar month.’’ This concludes the non-numerical signs, which, as 
already mentioned, doubtless had little real effect upon the meanings of the Sup- 
plementary Series. 
Turning next to the numerical glyphs, the most important of these is prob- 
ably Glyph A, the last sign of the Supplementary Series (see figure 79, a—h), and the 
key by means of which the general meaning of the whole count was first worked out. 
To begin with, the glyph itself, which is the moon-sign, is very constant in its 
normal form (Gieare 79, a—€) as well as in its head variant (figure 79, f); 1n fact, there 
is only one other variant known (figure 79, g and /), and that occurs but thrice.! 
The most constant characteristic about this glyph, however, and the one which 
gave the first clue as to its meaning, is its coefhicient, which is always 9 (figure 79, 
a, b, e, and h) or 10 (figure 79, c, d, f, and g), and which is always attached to the 
right of the glyph (figure 79, b, d, e, and f) or at the bottom (figure 79, a, c, g, and h); 
that is, never at the left or above, as in the case of all other numerical coefficients.” 
Férstemann had shown long before, in his study of the Dresden Codex, that 
the moon-sign there has a numerical value of 20, and in 1915 Professor R. W. Will- 
son, of Harvard University, suggested to the writer that Glyph A of the Supple- 
mentary Series, which was nothing more than the moon-glyph with a coefficient of 
9 or 10, was a sign for the 29 and 30 day month respectively, the nearest approxima- 
tions possible in terms of whole days of the exact length of alunation. He further 
suggested that the close resemblance of the moon element in Glyph A to those forms 
so often found in the Dresden Codex, where the moon-glyph is used as a numerical 
sign for 20, when taken into consideration with these coefhcients of 9 or 10, is of itself 
convincing proof that the Maya once used a lunar calendar consisting of alternate 
months of 29 and 30 days—such an arrangement as 1s in use in the Mohammedan 
calendar. And in a recent letter he refers the writer to a similar usage in the 
Babylonian lunar calendar in which the months were labeled 1 or 30 accordingly 
as they contained 29 or 30 days.’ It then became apparent why these coefficients 
of g and 10 were attached to the right of or below the moon-sign, instead of in the 
usual positions at the left or above. This was done in order to indicate thereby 
that they were added to the moon-glyph, giving totals of 29 (1. e., 20-+9) and 30 
(7.e., 20+ 10) instead of being multiplied by it, giving totals of 180 (1. ¢., 20X49) 
and 200 (20X10) as Maya coefficients do when they stand in the regular posi- 
tions to the left or above. 
The Maya had no fractions, and the only way they could keep the lunations 
correct in terms of whole days was to have some months composed of 29 days and 
A, CF i.e., the moon-sign proper, 


1(1) Stela A, Copan; (2) Stela N, Copan; and (3) the Hieroglyphic Stairway at Naranjo. 
“There are only four exceptions to this known, the last three texts at Quirigua, Stela I, Stela K, and Structure 1, 
9.18.10.0.0, 9.18.15.0.0, and 9.19.0.0.0 respectively, and on the west jamb of the north doorway of Temple 
11 here at Copan: 9.16.12.5.17. When the first three were inscribed, however, the purpose for which this 
differentiation of position had been devised was so well known that no mistake in meaning could arise about 
them, and their coefficients were allowed to go back to the regular positions for other coeficients. And for the 
explanation of the irregularity of the last, see pages 311-313. 3See Ginzel, 1906-1914, vol. I, p. 124. 
