IgO BUREAU OF AMERICAN ETHNOLOGY [bull. 118 



the left-hand side, looking tov* T ard the entrance. The hard-clay sub- 

 soil sloped from the right-hand to the left-hand side of the cave ; which 

 fact undoubtedly offers an explanation of this peculiarity. If the 

 bodies had been tossed in from above all would have rolled toward 

 the left wall of the cave. 



Not one skeleton was found entirely in anatomical order, as shown 

 in plate 122, a. It was, however, possible to determine that corpses 

 had been deposited in the cave rather than bundle reburials or the 

 remains from another mortuary. In one instance more than half 

 of the skeleton of a child was found in anatomical order. Associ- 

 ated with it were several olivella-shell beads. Occasionally femora, 

 tibiae, humeri, radii, ulnae, and ribs were in recognizable anatomical 

 order. Of note, also, was the fact that the skulls, for the most part, 

 were farther down the declining slope than the long bones which 

 appeared to be associated with them. This would lead to the 

 assumption that the corpses were cast into the cavern head fore- 

 most. Under such conditions, skulls, when detached, would roll 

 to the lowest part of the cavern floor. 



Two male crania recovered from what was roughly estimated to 

 be about 20 burials were definitely dolichocephalic, by casual 

 observation. 



No artifacts were found other than the beads previously men- 

 tioned, the two discoidals, and the bone implement illustrated in 

 the top row of plate 122, h. 



Site No. 21.— CKAWFOKD FAEM MOUNDS 



The farm of Mr. Samuel Crawford lies along the north side of 

 Clinch Kiver, near Scarboro, Anderson County, Tenn. About a 

 quarter of a mile north of the river and just opposite to Copper 

 Kiclge in a cultivated field were two mounds. These mounds, the 

 larger one designated No. 1 and the smaller one No. 2, have never 

 been cultivated. They once had large trees growing on them. 

 Some 10 years ago these trees were cut, and only large stumps 

 remain. Second growth, watersprouts, and weeds still covered the 

 mounds, as shown in plate 123, a. The difference in size of the two 

 mounds is shown in plate 123, &, which was taken after the mounds 

 had been cleared of all undergrowth. Sometime about 1926 the 

 larger mound, No. 1, was disturbed by residents of the vicinity. A 

 trench, which reached almost to the center of the mound, was dug 

 into it from the south side. 



Mound No. 1 was 45 feet in diameter and Mound No. 2 was about 

 35 feet in diameter. The centers of these mounds were about 60 feet 

 apart, Mound No. 2 being southwest of Mound No. 1. Because of 



