78 PROCEEDINGS OF THE ACADEMY OF [1883. 



of characters whose significance must be reexamined. Thus it is 

 impossible to characterize the Creodonta as lacking a trochlear 

 groove of the astragalus, in view of the form of that element in 

 Mesonyx and Mioclaenus, where the groove is more or less dis- 

 tinct. It is impossible to distinguish the Insectivora from the 

 Creodonta by the deficiency of canine and large development of 

 incisor teeth. In Rhynchocyon the canines are large, and the 

 superior incisors wanting, while in Centetes the arrangement of 

 these teeth is precisely as in the Creodonta. As to the large 

 Achsenodon and other Arctocyonidae^ I find no characters whatever 

 to distinguish them from the generally small Mesodonta. 



In view of these inconsistencies, I have reexamined the subject, 

 and find the following definitions to be more nearly coincident 

 with the natural boundaries of the divisions of this large order. 

 The importance of the character of the tritubercular superior 

 molar has recently impressed me (see Proceedings of the 

 Academy, 1883, p. 56), as it had previously done Prof. Gill. 

 This zoologist has already distinguished two divisions of the 

 Insectivora (without the Galeopithecidae), by the forms of the 

 superior molar teeth. The first possesses quadritubercular molars 

 above, the second tritubercular. That these types represent 

 important stages in the development of the molar dentition I have 

 no doubt. These characters far outweigh in importance those 

 expressing the forms of the skull, matters of proportion only, 

 with which a few systematists unnecessarily overload their diag- 

 noses. Such characters are of little more than specific value, and 

 serve to obscure the mind of the inquirer for a true analysis. 

 They may be used empirically, it is true, to determine relation- 

 ships when the diagnostic parts are wanting. 



I propose to transfer the Insectivora with tritubercular superior 

 molars to the Creodonta, in spite of the fact that some of them 

 (Mythomys, Solenodon, Chrysochloris) have but weakly developed 

 canine teeth, and Chrysochloris has large incisors. As an extreme 

 form, Esthonyx will follow, standing next the Tillodonta. It will 

 then be necessary to transfer the Arctocyonidae and all the 

 Mesodonta to the Insectivora, where they will find affinity with the 

 Tupeeidae. These have well-developed canines and small incisors, 

 as in the extinct groups named. The Chiromyidse must be dis- 

 tinguished from all of the other suborders, on. account of its 

 rodent-like incisors, combined with its lemur-like feet. 



