The number of clans in a tribe naturally bore some proportion to the populousness of the tribe, varying from three, in the case of the Delawares, to twenty or more, as in the case of the Ojibwas and Creeks. There were usually eight or ten, and these were usually grouped into two or three phratries. The phratry seems to have originated in the segmentation of the overgrown clan, for in some cases exogamy was originally practised as between the phratries and afterward the custom died out while it was retained as between their constituent clans.[78] The system of naming often indicates this origin of the phratry, though seldom quite so forcibly as in the case of the Mohegan tribe, which was thus composed:[79]-- I. WOLF PHRATRY. _Clans:_ 1. Wolf, 2. Bear, 3. Dog, 4. Opossum. II. TURTLE PHRATRY. _Clans:_ 5. Little Turtle, 6. Mud Turtle, 7. Great Turtle, 8. Yellow Eel. III. TURKEY PHRATRY. _Clans:_ 9. Turkey, 10. Crane, 11. Chicken. Here the senior clan in the phratry tends to keep the original clan-name, while the junior clans have been guided by a sense of kinship in choosing their new names. This origin of the phratry is further indicated by the fact that the phratry does not always occur; sometimes the clans are organized directly into the tribe. The phratry was not so much a governmental as a religious and social organization. Its most important function seems to have been supplementing or reinforcing the action of the single clan in exacting compensation for murder; and this point is full of interest because it helps us to understand how among our Teutonic forefathers the "hundred" (the equivalent of the phratry) became charged with the duty of prosecuting criminals. The Greek phratry had a precisely analogous function.[80] [Footnote 78: H. H. Bancroft, _Native Races of the Pacific States_, vol. i. p. 109.] [Footnote 79: Morgan, _Houses and House-Life_, p. 16.] [Footnote 80: See Freeman, _Comparative Politics_, p. 117; Stubbs, _Const. Hist._, vol. i. pp. 98-104; Grote, _History of Greece_, vol. iii. pp. 74, 88. It is interesting to compare Grote's description with Morgan's (_Anc. Soc._, pp. 71, 94) and note both the closeness of the general parallelism and the character of the specific variations.] [Sidenote: Structure of the tribe.]