106 Spencer and Walcott : 



example, Fig, 3, Plate XXXVII., represents the impression nia(ie 

 in clay by the hinder part of the cutting edge of the lower right 

 premolar. If the curve of the cut edge in the model be com- 

 pared with the same in Figs. 1, 2, 4, 5, Plate XXXVII., the 

 likeness between them is striking. 



Continuing the experiment, and cutting right through the 

 clay with the tooth, the piece (Fig. 15, Plate XXXVII.) yielded 

 by the inner side of the cutting edge is, so far as the curve is con- 

 cerned, practically identical with the bone specimen (Fig. 16, 

 Plate XXXVII.).' 



Other specimens again, especially those on which the cuts 

 occur in series parallel to the length of the bone, are not suffi- 

 ciently definite to give reliable results. 



The concavities in Fig. 14, Plate XXXVI., and Figs. 17, 18, 19, 

 Plate XXXVII., we quite failed to reproduce experimentally, and 

 could only conclude that, as there is no good reason to doubt the 

 identity of the agentof causation, and the concavities, as before men- 

 tioned, were not clear cuts, fracturing had largely influenced 

 the contour in question. One of the concavities in Fig. 19, 

 Plate XXXVII., especially supports this view, and might have 

 been made by the rounded anterior end of the upper premolar, or 

 by the formidable incisors. In connection with these, a feature in 

 the large fragment (Fig. 1, Plate XXXVIII.) may be noticed. It 

 is seen at one end of the specimen, and consists of a curved cut (a), 

 through the cortex of the bone to a depth of about 3 mm., by 

 which the cortex ha.s been removed entirely between the cut 

 and the fractured surface of the bone. Just close to this, and 

 towards the more distant end of the specimen, a small concave 

 piece (b) has also been taken out of the cortex to the same depth. 

 The curve which forms the margin of the larger cut is in 

 general agreement with that seen in Fig. 4, Plate XXXVII. , but, 

 so far as we can tell, it appears to have been made by the upper pre- 

 molar, the anterior part of which for about 18 mm. fits re- 

 markably well into the line of curvature. The curve must have 

 been completed by a continued fracture. The small concavity 

 (b) immediately behind the large curve (a) could easily be made 

 by the anterior raised end of the cutting ridge of the tooth. 



In Fig.s. 13, 14, Plate XXXV IT., the obtuse angle, formed 

 where the cut face (a) meets the edge of the bone (b), coincides 

 with the angle (c) of the cutting ridge of the upper tooth 



