1892.] NUMERICAL VARIATION IN TKETH. Ill 



than the other, which Mr. Thomas tells me is abvKjrmally small for 

 the species. In the upper jaw of a normal skull there are two small 

 premolars (p' and p'^ of Thomas) and behind these four molars. 

 The molars increase in size from the first to the third, which is by 

 far the largest. Behind the third is the fourth molar, which is 

 much thinner than the others. On comparing the abnormal skull 

 with the normal one it is seen, firstly, that on the left side there are 

 seven teeth behind the canine, while on the right side there are 

 only six such teeth, as usual. On the right side, however, the last 

 molar has not the thin flattened form of the last molar of a normal 

 skull, but is a fair-sized thick tooth. In each lower jaw there are 

 seven back-teeth instead of six. In making a more detailed com- 

 parison, the first five teeth on each side are clearly alike, while from 

 its form the seventh on the left side might be thought to represent 

 the normal sixth, and this is the view originally proposed by Mr. 

 Thomas in his ' Catalogue of Marsupialia,' p. 2G5, note. The 

 difficulty in this view is tbat it offers no suggestion as to the nature 

 of the sixth tooth on the right side. In the light, however, of what 

 has been observed in other cases of extra molars, it seems likely that 

 on the right side m' has been raised from a small tooth to one of 

 fair size, while on the left side the process has gone further, m^ 

 being still larger and another tooth having been formed behind it. 

 Mr. Thomas, to whom I am greatly indebted for having first shown 

 me this specimen, allows me to say that he is prepared to accept the 

 view here suggested. 



This phenomenon, of the enlargement of the terminal member of 

 a series when it becomes the penultimate, is not by any means con- 

 fined to teeth, for the same is true in the case of ribs, digits, &c., 

 and it is possibly a regular property of the Variation of Series of 

 Midtiple Parts which are so graduated that the terminal member is 

 the smallest. This fact will be found of great importance in any 

 attempt to conceive the physical process of the formation of Mul- 

 tiple Parts, and, pending a full discussion of this and kindred pro- 

 cesses, it may be remarked that such a fact strikingly brings out the 

 truth that the whole Series of Multiple Parts is bound together into 

 one common whole, and that the addition of a member to the series 

 may be correlated with a change in the series itself, and may occur 

 in such a way that the general configuration of the whole series is 

 preserved. In this case the new member of the series seems, as it 

 were, to have been reckoned for before the division of the series into 

 parts. This is, of course, only one way in which numerical Variation 

 may take place ; for, as was described in the previous section, ad- 

 ditions to the series may be formed by the division of single members 

 of the series, and in this case the configuration and proportions of 

 the rest of the series remain normal. Examples of these two distinct 

 methods of numerical Variation occur among Series of Multiple 

 Parts of many kinds (digits, vertebrae, &c.). 



Re-constitution of Parts of the Series. 

 Some curious instances of what is almost a remodelling of parts 



