226 Bulletin American Museum of Natural History. [Vol. XXVII, 



the early mammals the skull had already advanced far beyoiul that of the 

 Cynodonts in the backward prolongation of the hard palate, in the reduction 

 of the quadrate and articulare and possibly in the transformation of these 

 bones into the incus and malleus respectively (p. 140). 



After the branching off of the Monotremes the Marsupio-Placental 

 remnant still retained certain reptilian features in the skull, l)ut owing to 

 the assumption of semi-arboreal hal)its it lost the Monotreme characters of 

 the shoulder girdle, clear traces of which (Broom) are still retained in 

 embryo Diprotodonts (p. 157). Improved climbing powers resulted from 

 the atrophy of the episternum and procoracoid, the reduction of the coracoid, 

 the pulling outward of the glenoid away from the middle line and the de- 

 velopment of the prespinous fosste and its muscles. 



At this point the divergence into Marsupials and Placentals may have 

 begun. The forerunners of the Marsupials, we may imagine, went on im- 

 proving the arboreal characters, finally actpiiring the rudiments of syndactyly 

 in the pes, which, from its occurrence (Bensley) in Marmosa pusilla (p. 217), 

 may be suspected to be a primary character in the Polyprotodontia as well 

 as in the Diprotodontia (p. 215). The Pro-Placentals however avoided this 

 particidar feature of arboreal specialization, as well as the lengthening of 

 the fourth digit of the pes, merely retaining a partly divergent poUex and 

 hallux. The change from oviparity to viviparity was meanwhile proceeding 

 in the manner outlined above (p. 14S). The pro-Marsupials laid stress on 

 the extra-uterine, or marsupial adaptations, so that the young, through the 

 retention and emphasis of certain embryonic features, such as the intra- 

 narial epiglottis, ac({uired larval characters. Peculiar relations between the 

 maternal teat and the mouth of the offspring, joined to the desirability of 

 precocious independence in gaining food, may have conditioned the early 

 specialization of the last upper and lower milk molar and the suppression 

 (except in certain Sparassodonts, p. 207) of the anterior milk molars (Leche, 

 quoted by Weber, 1904, p. 341). The Pro Placentals on the contrary, 

 laying stress on the intra-uterine ada])tations, as well as on the improvement 

 of lactation, lengthened the period during which the young are dependent 

 on the mother, with the result that a more leisurely and full development 

 of both milk and permanent dentitions followed. 



The ]Marsupials also accjuired certain peculiar specializations which 

 were avoided both by INIonotremes and by Placentals. The allantois, which, 

 as in Perameles, may have begun to unite with the uterine wall was crowded 

 out by the yolk sack, which also assumed this function. 



The hard palate, which possibly was prolonged backward at a very early 

 period began secondarily to acquire fenestrse, or to enlarge its fenestras, through 

 the absorption of tissue (p. 220). The pterygoid fossa of the mandible 



