230 Bulletin American Museum of Natural History. [Yol. XX^'II, 



From the foregoing it appears probable that the Diprotodontia have 

 retained certain primitive mammahan characters, which have been lost in 

 the Polyprotodontia, while on the other hand the Didelphids have remained 

 primitive in the dentition, skull, and foot-structure. The problem of the 

 genetic relations of the Diprotodontia and the Polyprotodontia is compli- 

 cated to a certain extent by the existence of the Ci^nolestoids; but the opinion 

 may be expressed that probably the resemblances of certain Csenolestoids to 

 the Multituberculates (p. 214) is an instance of convergence between related 

 suborders, and that the same is true, but to a less extent, of the resemblances 

 of other Csenolestoids to the Diprotodont phalangers. 



The foregoing views of the relations of the suborders of INIarsupials to 

 each other and to the stem of the Placentals are expressed in the accompany- 

 ing tables (p. 229). 



The consideration of the more detailed evidence of the common origin 

 of the Marsupials and Placentals may be deferred until after the review of the 

 Insectivores and Creodonts (p. 307). 



XI. Diagnoses of the divisions Prototheria, Didelphia, 



MONODELPHIA. 



SUBCLASS PROTOTHERIA. Reptilian angular bone absent (at least in propria 

 fortna) . Oviparous ; no secondary teats. Shoulder girdle and humerus much as 

 in the mammal-like reptiles: prespinous fossa of the scapula absent or rudi- 

 mentary, lower part of the anterior border of the scapula corresponding to the 

 spina scapulse and acromion of higher mammals; large coracoids and pro- 

 coracoids overlapping in the mid- ventral line; large T-shaped interclavicle ; 

 glenoid facet for humerus very near the sternum. Habits primarily semifos- 

 sorial. Epipubic bones present. No corpus callosunj. 

 SUBCLASS THERIA ' Parker and Haswell. 

 (= Eutheria Gill, 1872, non Huxley.) 



Viviparous, with teats; coracoid and procoracoid reduced, not touching 

 sternvmi; interclavicle absent or vestigial; glenoid facet for humerus 

 widely separated from sternum; prespinous fossa present (much re- 

 duced in Cetacea). Habits primarily semi-arboreal. 

 1. Infraclass Didelphia de Blainville. 



Epipubic bones; no corpus callosvun (Elliot Smith, 1894); angle inflected 

 (save in Tarsipcs) ; typically with only the posterior milk molar 

 replaced by a successor; p§ absent; molars typically 4-. 



1 It is unfortunate that the term "Eutheria" has come to stand for two very distinct con- 

 cepts: first, for Marsupiaha + Placentalia (Gill, Osborn, Beddard); secondly, for the Placen- 

 talia alone (Huxley). Although Gill's usage of the term has the priority (p. 92), it will probably 

 be impossible at this late date to eradicate Huxley's usage of the term, which is very widely 

 disseminated through the literature. Rather than contribute to this regrettable confusion, 

 and recognizing that the law of priority has not generally been applied rigidly in the case of 

 larger taxonomic groups, the writer has employed the term "Theria," as used by Parker and 

 Haswell, as the equivalent of "Eutheria" Gill. 



