Tooth-marks of Tliylacoleo. 105 



surface, showing the area exposed beneath covered with small 

 incisions. Though the difference in the nature of the material 

 renders the experiments of little value in certain respects, yet 

 the result of treating a rounded surface made of modelling clay 

 with the large premolars of Thylacoleo led to results of some 

 interest so far as the shape of the cuts was concerned. In the 

 case of the bone, the nature of the material introduces a facTor 

 that has doubtless much influence in deciding the ultimate form 

 of the curve. The force used may, for example, produce first 

 a cut of a certain curved form, and, beyond this and continuing 

 the terminal line of the cut, it may form a fracture of varying 

 length. This, we believe, is well shown in Fig. 7, Plate XXXVIL, 

 a narrow pointed fragment, one margin of which is convex, the 

 other concave. The origin of the convex fracture is seen at its 

 base, where there is unmistakable evidence of teeth having been 

 forced into the bone from both sides. The concave side, on the 

 contrary, bears no indication of the origin of the fracture. It 

 may be due wholly to a fracture, or, what is more likely, the 

 part of the broad end may be due to an incision which has been 

 continued as a fracture to the pointed end. 



Those forms selected for reference are, however, we think, 

 free from this doubt, a strong point in their favour being the 

 remarkable similarity in the curves of the three most striking 

 specimens (Figs 4, 5, 6, Plate XXXVIL). Apart from those fig- 

 ured, it must be remembered that numbers of specimens down to 

 some of the smallest in the collection, when carefully examined, 

 reveal the fact that concave cuts have been made upon them. 

 These are so constant and so frequent that we are safe in 

 assuming that we are dealing with cuts and not Avith fractures. 

 Not only are the examples numerous, but the specimens have 

 been gathered from different places extending over a distance 

 of one quarter of a mile along the bed of the old swamp. 



It is very difficult to understand, even with the aid of such 

 experiments as we were able to make, how certain of the cuts 

 were produced. Other results obtained, however, were so close 

 in their resemblance to the actual bone specimens that it seems 

 certain that we are dealing with something more than a mere 

 coincidence. By taking an impression of the cutting edge of 

 both lower and upper teeth and comparing them with the 

 curves of the specimens, they are seen to be very similar. For 



