On comparing all the Winter Counts it is found that they often correspond, but sometimes differ. In a few instances the differences are in the succession of events, but they are usually due to an omission or to the selection of another event. When a year has the same name in all of them, the bands were probably encamped together, or else the event fixed upon was of general interest; and when the name is different the bands were scattered, or nothing of general interest occurred. Many of the recent events are fresh in the memory of the people, as the warriors who strive to make their exploits a part of the tribal traditions proclaim them on all occasions of ceremony, count their coups, as the performance is called. Declarations of this kind partake of the nature of affirmations made in the invoked presence of a supposed divinity. War shirts, on which scores of the enemies killed are kept, and which are carefully transmitted from generation to generation, help to refresh their memories in regard to some of the events.
The study of all the charts renders plain some points remaining in doubt while the Lone-Dog chart was the only example known. It became clear that there was no fixed or uniform mode of exhibiting the order of continuity of the year-characters. They were arranged spirally or lineally, or in serpentine curves, by boustrophedon or direct, starting backward from the last year shown or proceeding uniformly forward from the first year selected or remembered. Any mode that would accomplish the object of continuity with the means of regular addition seemed equally acceptable. So a theory advanced that there was some symbolism in the right-to-left circling of Lone-Dog’s chart was abandoned, especially when an obvious reproduction of that very chart was made by an Indian with the spiral reversed. It was also obvious that when copies were made, some of them probably from memory, there was no attempt at Chinese accuracy. It was enough to give the graphic or ideographic character, and frequently the character is better defined on one of the charts than on the others for the corresponding year. One interpretation would often throw light on the others. It also appeared that, while different events were selected by the recorders of the different systems, there was sometimes a selection of the same event for the same year and sometimes for the next, such as would be natural in the progress of a famine or epidemic, or as an event gradually became known over a vast territory.
A test of the mode of selecting events for designating the Winter Counts may be found in a suggestion made by the present writer in his account of Lone-Dog’s chart, published in 1877, as follows:
The year 1876 has furnished good store of events for the recorder’s choice, and it will be interesting to learn whether he has selected as the distinguishing event the victory over Custer, or, as of still greater interest, the general seizure of ponies, whereat the tribes, imitating Rachel, weep and will not be comforted, because they are not.
It now appears that two of the Counts made for 1876 and observed by the writer several years later have selected the event of the seizure of the ponies, and that none of them make any allusion to the defeat of Custer.
After examination of all the charts it is obvious that the design is not narrative, that the noting of events is being subordinated to the marking of the years by them, and that the pictographic serial arrangements of sometimes trivial though generally notorious incidents having been selected with special adaptation for use as a calendar. That in a few instances small personal events, such as the birth of the recorder or the death of members of his family, are set forth, may be regarded as interpolations in or unauthorized additions to the charts. If they had exhibited a complete national or tribal history for the years embraced in them, their discovery would have been in some respects more valuable, but they are interesting to anthropologists because they show an attempt before unsuspected among the northern tribes of American Indians to form a system of chronology.
While, as before mentioned, it is not now necessary to recapitulate the large amount of matter before published concerning the Winter Counts of the Dakota, it has been decided to present in an abbreviated form the characters and interpretations of the Lone-Dog chart as being the system which was first discovered, and the publication of which occasioned the discovery of all the other charts mentioned. The Winter Count of Battiste Good has not hitherto been published, and it possesses special importance and interest apart from its chronology, for which reason it is inserted in the present paper, see infra.
The several charts of The-Flame, The-Swan, American-Horse, and Cloud-Shield, published in the Fourth Annual Report of the Bureau of Ethnology, are omitted, but selections from all of them are presented under the headings of Ideography, Tribal and Personal Designations, Religion, Customs, History, Biography, Conventionalizing, Comparison, and in short are interspersed through the present paper where they appropriately belong.
The reader of the Lone-Dog and Battiste Good charts may find it convenient to note the following brief account of the tribal names frequently mentioned: