116 



i n characters of m^ when a m* is present. Why the upper jaw should 

 exemplify Bateson's rule and the lower follow another line is difficult 

 to explain. It would be interesting to find other specimens in which 

 the upper nv^ are modified. It may be added in this connection that 

 the illustration given by Hayes of the lower jaw shows no abnormality 

 in the form of the third molar. 



In cases of supernumerary teeth it is always of moment to dis- 

 cover if there is any abnormality in the form and size of the premolars 

 and molars. For one would expect that the presence of an extra tooth 

 n the series would produce some modification of the normal constituents. 

 Anyone who has carefully observed the cheek-teeth of the horse must 

 have been struck by the constant diiferences, clearly evident to the 

 eye, in the size and shape of the wearing surfaces of the different 

 teeth. But, where an accurate comparison is desired, it is not safe 

 to trust to the eye alone. Indeed the eye is notoriously misleading. 

 The only reliable method by which the size of a tooth can be deter- 

 mined is by careful measurement. Further, it is necessary, when 

 making comparisons, to eliminate any possibility of error which may 

 arise from the differences of stature of the animals from which the 

 structures to be compared have been derived. This can be done by 

 reducing the measurements to indices. 



In order to compare the size of the teeth in the abnormal 

 specimens with that of the normal cheek-teeth of the horse, I took 

 ten mandibles of adult animals and subjected them to the method of 

 measurement which has been employed by Salensky (7); that is, the 

 length and breadth of their wearing surfaces were taken by means of 

 graduated calipers. From these measurements a "length-breadth", or, 

 as it will be more convenient to call it, a "tooth index" was computed, 

 by assuming the length of the tooth to be 100, and calculating what 

 proportion of 100 was capable of expressing the breadth. The following 

 formula will more clearly explain the method: 



Breadth of tooth X 10Ö murin' 



; — ji Q = Tooth Index M. 



Length of tooth 



It would take up too much space to give all the measurements 

 of the normal teeth, and the indices calculated from them. It will 

 suffice to give the results in a general way. That the indices varied 

 very considerably was not more than was to be expected; and, there- 



1) It is better to use the term "tooth index", rather than "dental 

 index", as the latter expression has come to have a definite meaning in 

 Anthropology, different from that here indicated. 



