This design survives among the Ottawa and Pottawatomi Indians of Northern Michigan, but among the latter a single line only is drawn across the breast, shown in Figure 145. This corresponds, also, with one of the gesture-signs for the same idea. [Illustration: FIG. 145.--Starvation. Ottawa and Pottawatomi.] Figure 146, from the record of Battiste Good for the year 1826-’27, signifies “pain.” He calls the year “Ate-a-whistle-and-died winter,” and explains that six Dakotas, on the war path, had nearly perished with hunger when they found and ate the rotting carcass of an old buffalo, on which the wolves had been feeding. They were seized soon after with pains in the stomach, their bellies swelled, and gas poured from the mouth and the anus, and they died of a whistle, or from eating a whistle. The sound of gas escaping from the mouth is illustrated in the figure. The character on the abdomen and on its right may be considered to be the ideograph for pain in that part of the body. [Illustration: FIG. 146.--Pain. Died of “whistle.” Dakota.] SYMBOLISM. The writer has, in a former publication, suggested the distinction to be made between a pictorial sign, an emblem, and a symbol; but it is not easy to preserve the discrimination in reference to ideographic characters which have often become conventionalized. To partly express the distinction, nearly all of the characters in the Dakota Winter Counts are regarded as pictorial signs, and the class represented by tribal signs, personal insignia, etc., is considered to belong to the category of emblems. There is no doubt, however, that true symbols exist among the Indians, as they must exist to some extent among all peoples not devoid of poetic imagination. Some of them are shown in this paper. The pipe is generally a symbol of peace, although in certain positions and connections it sometimes signifies preparation for war, and again subsequent victory. The hatchet is a common symbol for war, and closed hands or approaching palms denote friendship. The tortoise has been clearly used as a symbol for land, and many other examples can be admitted. If Schoolcraft is to be taken as uncontroverted authority, the symbolism of the Ojibwa rivalled that of the Egyptians, and the recent unpublished accounts of the Zuñi, Moki, and Navajo before mentioned indicate the frequent employment of symbolic devices by those tribes which are notably devoted to mystic ceremonies. Nevertheless, the writer’s personal experience is, that often when he has at first supposed a character to be a genuine symbol it has resulted, with better means of understanding, in being not even an ideograph but a mere objective representation. In this connection, the remarks on the circle on page 107, and those on Figure 206, on page 246, may be in point.