466 BUREAU OF AMERICAN ETHNOLOGY [bull. 28 



bers of the manuscript incorrect. Moreover, the 20 errors are lessened 

 b}^ the fact that six of them ai^e really one and the same, for in all 

 of these six cases, where the difference is 178, the writer has over- 

 looked this and mechanically' Avritten down the usual 177, although 

 the numbers and days of the series quite correctly indicate 178. 

 Moreover, the three errors in groups 58 and 59 are only one, for the 

 author had confounded the differences 177 and 148, and was, there- 

 fore, obliged to write the number 10,089 instead of 10,068, which will 

 find confirmation later. From this it follows besides that the writer 

 was at the same time the computer, consequently the actual author. 



I must further call attention to the regular position of the differ- 

 ences 178 and 148. In the three periods of 1,742 -days 178 is always 

 in the sixth place, in those of 1,034 days it is always in the fourth 

 place. It appears, therefore, in groups 6, 14, 29, 37, 52, and 60, that 

 is to say, at intervals of 8, 15, 8, 15, 8 groups; in the periods of 1,210 

 days it is wholly wanting. The difference 148 in the nine divisions is 

 alwaj^s in the third place, that is to say, always close to the pictures, 

 of which we shall presently speak; therefore, in groups 3, 13, 19, 26. 

 36, 42, 49, 59, 05, that is, at intervals of 10, 6, 7, 10, 6, 7, 10, 6 groups. 

 We can not yet look further into the causes of this curious fact. 



But I must refer to a pregnant error. Groups 22 and 23 quite 

 correctly have the difference 177, but the writer in this single place 

 sets down 178 and consequently computes the three days belong- 

 ing here as VII 11, VIII 12, "iX 13, instead of VI 10, VII 11, 

 VIII 12, and from here to the end he is always one day in advance, 

 so that group 69 on page 58 closes wirh the days X 3, XI 4, XII 5, 

 which ought to be IX 2, X 3, XI 4. 



Now it is important to determine the zero point belonging to this 

 series, for every series of this manuscript conceals it. It must be 177 

 days before the first group, that is, before daj^s VI 1, VII 2, and 

 VIII 3. which leads to\he days XI 4, XII 5, and XIII 6. 



Of these days the middle one, XII 5, is by far the most important ; 

 it occurs on the upper half of page 51 six times, on page 52 four times. 



On page 51, in the first column on the left, we first find the normal 

 date and starting point of the computation, the da}^ IV 17, as the 

 eighth da}^ of the eighteenth month in the year 9 Ix, but under it 

 our day XII 5. Below the latter there is an 8, beneath this number 

 the character kin ('' sun ", "day "), and combined with the' latter the 

 character imix, with a sign above it clearly denoting " combination ", 

 " union ■'. In the Zeitschrift fiir Ethnologie, 1891, page 152, I have 

 already ascribed the meaning of a katun (18,980=52X365 days) to 

 this combination, and I still hold this opinion. 



This group may, therefore, signify the 8 days which elapse between 

 IV 17 and XII 5, but it may also denote the period of 8X18,980= 

 151,840 days; probably it signifies both at the same time. 



