392 THE INSCRIPTIONS AT COPAN. 
does not conform to this practice; and although it is possible earlier hotun- 
endings may yet be found there,' the custom does not appear to have been 
followed so closely as at Copan. 
This practice, with varying modifications, continued in force down to 
the close of the New Empire in Yucatan, and is mentioned by several of the 
early Spanish and native writers.2 It was indeed one of the oldest of all 
Maya institutions, and, so far as the stele are concerned, it was the most 
important factor in determining the dates of their erection. 
Another feature of Maya chronology, possibly inaugurated at Copan, 
was the gradual abandonment of Initial Series dating and the coincidental 
increase of Period Ending dating during the last half of the Great Period. 
Of the cities of Classes I and II given on page 441, with the exception of 
Nakum, where Initial Series have not been found at all, and of Palenque, 
which may have been abandoned before this time, Copan was the first to 
discard this cumbersome but extremely accurate method of dating, the last 
Initial Series there being carved in 9.17.12.13.0 (Stela 4) as compared with 
9.18.3:1.5 at Yaxchilan (small altar near Stela 1), 9.18:5.0.0 at /icares 
Negras (Stela 12), 9.18.10.0.0 at Naranjo (Stele 8, 12, and 28), 9.19.0.0.0 
at Quirigua (Structure 1), and 10.2.0.0.0 at Tikal (Stela 11). 
A glance at the fourth column in Appendix IX, which gives the kinds 
of dates on the different monuments at Copan, 7.e., whether Initial Series, 
Period Ending, or Calendar Round, this condition appears clearly. During 
the Early and Middle Periods, Initial Series are the almost invariable rule; 
indeed, in the Middle Period there is not a single stela without one; but 
after 9.17.12.13.0 (Stela 4) they stop at Copan, the closing hotun-endings, 
9.18.0.0.0, 9.18.5.0.0, and 9.18.10.0.0 being marked by Period Ending dates 
on altars instead. 
The Initial Series method of dating required eight glyphs to express 
any single date of the Maya Era, and was correspondingly costly of space 
and effort. The Secondary Series was probably developed very early to 
escape from such a tedious process when more than one date had to be 
recorded on the same monument, and as early as 9.4.10.0.0 (Altar Q’) Period 
Ending dating, as applied to lahuntuns, was in use at Copan, and as early as 
9.7.8.0.0 as applied to any tun (inscribed peccary skull from Tomb 1). 
By the middie of the Great Period this method of dating was beginning 
to supersede the Initial Series, in some cities altogether, as for example at 

! There are two stele at Tikal, Nos. 8 and 9, which the writer believes he may possibly have deciphered correctly 
as 9.0.10.0.0 and 9.2.0.0.0 respectively. On the west side of Stela 8 (see Maler, 1911, plate 19, 1), A6 1s an ending- 
sign in which the hand element is very conspicuous, and following this in A7 appears to be the day 7 Ahau. Assum- 
ing this is a katun or lahuntun-ending, by no means certain, however, the only two positions where this day could 
have occurred in the Early Period are at 9.0.10.0.0 7 Ahau 3 Yax and 9.7.0.0.0 7 Ahau 3 Kankin, and since this 
latter date is too late on stylistic grounds to be possible, the former may be the date of thissmonument. The date 
suggested for Stela g seems more certain. ‘The west side of this stela (see ibid., plate 20, 1) opens with a hand 
ending-sign at AI; in A2 there appears to be either 4 or 2 katuns, and in a3 the day 4 Ahau. The only katun in 
the Early Period ending on a day 4 Ahau was 9.2.0.0.0 4 Ahau 13 Uo, and therefore, in spite of the fact that the 
katun coefficient in a2 looks more like 4 than 2, the latter would appear to be the better reading, and 9.2.0.0.0, 
the probable date of this stela. These two readings are so doubtful, however, that they have been disregarded in 
the comparisons and conclusions which follow. 
2 See Appendix VII. 
