THE AMERICAN NATURALIST. [Vol. XXXVIII. 



3 that is known of the Mesozoic MammaUa fits in 

 ypothesis of their arboreal habitat but adds Httle 

 ■ncc in its favor. Practically nothing is known of 

 al structure ; they are all of small or minute size, 



<it insectivorous or granivorous type. They have 

 c(l to inonotrcmcs, marsupials and insectivores, in 



ow \cr\- insul'ticient evidence, but their ordinal 

 s have little to do with the question of their habi- 

 (1 not be considered here. Their minute size, and 



in strata of fresh or brackish water origin, with 

 ibious and aquatic reptiles and with abundance of 

 , suggest that the deposits in which they occur 

 •wn in extensive forest-clad river deltas and coastal 

 I that the minute mammalia represent the arboreal 



t well acquainted with the embryologic evidence 

 t hear upon the hypothesis advocated in this paper, 

 igations into mammalian ontogeny have had other 

 1 \iew. The deviation of placentals from mar- 

 icc \ ersa, and the ultimate origin of the mammalia 

 i)ian or reptilian ancestors, do not directly affect the 

 the habitat of the Mesozoic ancestors of the Ter- 

 ials. But undoubtedly important evidence on this 



he obtained from the ontogeny of their modern 

 , although not, perhaps, of the same force as the 



e\itlcnce from palaeontology. 



