lack of bones or teeth of large pelycosaurs in the extensive collections of the University of Kansas from the Fort Sill locality.

If Thrausmosaurus is, in fact, adult, the genus is an unusually small sphenacodontid, and of significance both on that account and because of the resemblance of the teeth presently known to those of its far larger relatives.

The Fort Sill Locality.—Peabody (1961) suggested that the fissures of Fort Sill had been used as dens by predatory animals in the early Permian, and that the unusually abundant bones in the fissures were the remains of animals eaten there by these occupants. Evidence now known to me affords an alternative explanation that is presented here as a preliminary to a more complete study of the fauna and paleoecology of these deposits currently being undertaken.

The suggestion that the skeletal material found in the fissures is the remnant of the prey of other animals is questionable because of:

1. The absence of tooth marks on the fossils.

2. The recovery from the matrix of skulls and portions of articulated skeletons that are undamaged or damaged only by pressure after burial.

3. The rarity in the deposits of animals of larger body size than Captorhinus, the exceptions being a few limb fragments and skull fragments of labyrinthodont or pelycosaurian nature.

4. The absence of coprolites in the matrix.

If the fissures were the dens of predators, at least some and probably many of the bones would show tooth marks. A predator feeding on other animals would be expected to leave some evidence of its habits on the bones of its prey. No such evidence is known to me, either from my own examination of several thousand bones or from the reports in the literature by others who have studied aspects of the early Permian fauna of Fort Sill.

If the predators were larger than Captorhinus and occupied the fissures for a long enough time to account for the accumulation of the tremendous numbers of individuals that are represented, a considerable amount of the skeletal material of the larger animals would be present in the fissure deposits. Even if for some reason the predators died in areas other than within the fissures, thereby accounting for the absence of large bones, coprolites should appear in the deposits if, in fact, the fissures were feeding places. In view of the nearly undamaged condition of many of the bones recovered from the fissures, it is reasonable to expect that fecal material would be preserved.

The character of the matrix of the deposits varies from a homogeneous clay to clay interrupted by layers of soft, limey, conglomeratic rock, to a hard, well-cemented, calcareous conglomerate. In general the bone in each kind of matrix is colored characteristically and exhibits a characteristic degree of wear. The bones entrapped in the homogeneous clay are relatively few, black, usually disarticulated, little worn and not unduly fragmented; consequently the discovery of undamaged limb bones, for example, from this kind of matrix is not unusual. The bones found in the stratified portion of the matrix are more numerous within the layers of conglomerate than between. The bones are black, brown or white, highly fragmented and waterworn to a variable degree. The fragments recovered from the hard, calcareous matrix are numerous, range in color from white through various shades of brown, to black, are highly fragmented, and are usually worn by water.

These categories for bone and matrix, however, are not mutually exclusive, since bones of any of these colors and exhibiting any degree of wear and fragmentation are found in any of the kinds of matrix described above. That water was the agent of wear is suggested by the highly polished appearance of the worn bones and pebbles that are found in the matrix.

The variability of