348 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 



EXCAVATIONS 



Eleven 5-foot-square test pits were excavated in the floor of the cave 

 (fig. 63) . Each of these was carried down to bedrock. Arbitrary 0.5- 

 foot levels were maintained throughout the work (pi. 54, a) . It was 

 found that the floor fill had been badly disturbed to a depth of from, 0.7 

 foot to 1.1 feet. Since no diagnostic artifacts were recovered below this 

 level, no cultural stratification could be obtained at the site. 



The fill is not more than 4.5 feet at the maximum. It consists of 

 organic material combined with roof fall. A section from the deepest 

 part of the fill, which is toward the front of the shelter and in the 

 center, revealed five layers. The top stratum, which had an average 

 thickness of 0.3 foot, consisted of a dark clay -humus. Underlying this 

 was a layer of light-yellow sandy soil with an average thickness of 0.6 

 foot. This layer was underlain by a very thin stratum of light-brown 

 sandy soil, 0.2 foot thick. The fourth layer consisted of orange-red 

 sand with thin, angular fragments of sandstone. It was 0.6 foot thick. 

 The fifth layer was gray clay, containing large fragments of slightly 

 decomposed sandstone and small lenses of organic humus with a clay 

 admixture. At a depth of 1.3 feet the color of the clay changed to a 

 deep gray-blue shade. 



PETROGLTPHS 



Of the several petroglyphs mentioned in the Andreas account, only 

 two major examples remain intact. The first of these occurs on the 

 upper surface of a large slab, undoubtedly the same one described in the 

 Andreas account. It is incised in the rock and is apparently intended 

 to represent a horse, since the tail is quite long (pi. 55, a) . Below the 

 head of the creature, which is square rather than elongated, there is a 

 line through the neck. A bilobed arrow is incised above the horse's 

 back. The treatment of this horse reminds one of the representations 

 of the mythical Underwater Panther in Plains and Woodland In- 

 dian song pictographs. Here such a line through the neck indicates 

 that the creature is emerging from beneath the earth (James, 1956, 

 pp. 345-346). 



The second petroglyph occurs on the side of a large boulder a few 

 feet south of that bearing the horse design. It seems to be a repre- 

 sentation of some growing plant, such as corn, or grass (pi. 55, h). 

 Because this petroglyph is so different in style from the others and 

 from those at other sites in the area, it may well be spurious. On the 

 other hand, it is scarcely the sort of thing a Wliite picnicker might 

 dash off on a Sunday afternoon. No tracing was made of this petro- 

 glyph because of the rouglmess of the stone on which it is inscribed. 



Of those petroglyphs described in the Andreas account that have 

 since been destroyed, one seems to merit special mention. This is the 



