). This is very unusual and, so far as the writer knows, is the only example of its kind. The four dots in the numbers 4, 9, 14, and 19 never appear thus separated in any other text known.
[60] In the examples given the numerical coefficients are attached as prefixes to the katun sign. Frequently, however, they occur as superfixes. In such cases, however, the above observations apply equally well.
[61] Care should be taken to distinguish the number or figure [20] from any period which contained 20 periods of the order next below it; otherwise the uinal, katun, and cycle glyphs could all be construed as signs for 20, since each of these periods contains 20 units of the period next lower.
[62] The Maya numbered by relative position from bottom to top, as will be presently explained.
[63] This form of zero is always red and is used with black bar and dot numerals as well as with red in the codices.
[64] It is interesting to note in this connection that the Zapotec made use of the same outline in graphic representations of the tonalamatl. On page 1 of the Zapotec Codex Féjerváry-Mayer an outline formed by the 260 days of the tonalamatl exactly like the one in fig. [48], a, is shown.
[65] This form of zero has been found only in the Dresden Codex. Its absence from the other two codices is doubtless due to the fact that the month glyphs are recorded only a very few times in them—but once in the Codex Tro-Cortesiano and three times in the Codex Peresianus.
[66] The forms shown attached to these numerals are those of the day and month signs (see figs. [16], [17], and [19], [20], respectively), and of the period glyphs (see figs. [25]-[35], inclusive). Reference to these figures will explain the English translation in the case of any form which the student may not remember.
[67] The following possible exceptions, however, should be noted: In the Codex Peresianus the normal form of the tun sign sometimes occurs attached to varying heads, as (*