504 MR. O. THOMAS ON THE GENUS ECHINOPS, [JunC 14, 



None of the specimens show any trace of a lower m'*. 



In these three genera of Centetidce and, so far as I can make out 

 from the published figures, in Potamogale, we have a premolar 

 formula of (0 2 3 4; the p* (both milk and permanent) is abso- 



J 2 3 4 



lutely molariform, while p^ (again both milk and permanent) is 

 functionally a " carnassial " — characters all in contrast to those of 

 Solenodon, on whose close relationship to tbe Centetidce so much 

 has been written. 



In this animal there are, it is true, three premolars as in Oentetes 

 and the others, but of these the first does not change, a fact which 

 seems to show that it is homologous with the non-changing p' of 

 other Ferae, and not with the changing p^. If this be the case, and 

 the two posterior changing premolars be still looked upon as p^ and 

 p*, we get the formula P.j 1 3 4, while as a further difference 



1 3 4 



p^ is absolutely premolariform, and it is p* that is functionally 

 " carnassial " or semi-molariform, no premolar being truly molari- 

 form ; and finally there is neither that striking resemblance between 

 the milk and permanent teeth, nor the unusually early development 

 of the true molars, characteristic of the Centetidce. 



It appears to me therefore that not only was Dr. Dobson 

 perfectly justified in separating the Solenodontidce as a distinct 

 family from the Centetidce, with which Peters had placed them, but 

 I would go further and suggest that their main connecting link — 

 their common trituberculism — may be merely in each case a remnant 

 of a time when, as the American zoologists have shown, trituberculism 

 was a far more common character than at present, and that they 

 have really no very close relationship to each other whatever. The 

 proper solution of this problem will probably rest with the palae- 

 ontologists of North America, on which continent the ancestors of 

 Solenodon may be expected to have lived ; and in consideration of 

 its great general and geographical interest, I would specially com- 

 mend it to the notice of any of them who may have the oppor- 

 tunity of examining remains referable to early Insectivora. 



The following is a tabular arrangement of the formulae above 

 described, with the addition of that of Gymnura, put in for com- 

 parison with the rest. Numbers in italics represent teeth which, 

 although present, are minute and apparently functionless. No 

 definite suggestions as to the serial homologies of p^ or of the naolars 

 are intended to be conveyed by the type in which their respective 

 numbers are printed. 



I. C. P. M. 



Centetes . 



Ericulus. 



12 3 4 



1 2 3 



1 2 3 



1 2 3 



