682 DISCONTINUOUS DISTKIUUTION OF AIAMMALS. 



but also the soft parts are available, it seems rather questionable 

 whether such a aistmction can be upheld. As examples, I may 

 mention the discovery of a placenta in Perameles ; the non- 

 inflected lower jaw of Tarsipcs; the absence, or, at least, vesti- 

 geal nature, of the epipubes in T/iylacinus; epipubic-like bones 

 m Oanis; palatal vacuities in Insectivora ; and marsupial vestiges 

 in many Euthena, especially in Lemur catta. 



Such anomalies have induced the belief that marsupials have 

 evolved from placental mammals. But although this opinion 

 has crystallised, it is still believed that the marsupials of Aus- 

 tralia and of America represent a single monophyletic group ; 

 and there are several theories to account for their present distri- 

 bution. The obvious explanation is naturally the one which 

 has received most support, viz., that Australia and South America 

 were formerly both united to the Antarctic continent. A strong 

 support of this theory was the belief that the separation of mar- 

 supials into Poly- and DiJ protbdontia was a natural one. and 

 that both groups occurred in, both regions. There are here two 

 hypotheses, neither of which is conclusively proved : the 

 American Polyprotodont opossums approximate towards the 

 Australian Diprotodont phalangers in general appearance 

 rather than to any Australian group of Polyprotodonts. 

 It may be said that this general resemblance is but skin- 

 deep, but there has been no change of habit ; the climbing habit 

 is retained in both, so that the approximation in form cannot 

 be adaptive, but must have some phyletic significance. It is said 

 (Mr. Thomas) that the American opossum is " very closely 

 allied to the Dasyuridse, from which, were it not for its isolated 

 geographical position, it w^ould be very doubtfully separable." 

 It seems to me, however, that the prehensile tail, so common in 

 living archaic forms, usually said to have retained Creodont 

 characters, distinguishes the more archaic opossums from the 

 less archaic Dasyuridge. Moreover, the American Ccenolestes 

 was at first referred to the Diprotodontia, but it is now referred 

 to the Polyprotodoiitia owing to mutual similarities between it, 

 the Australian Pouched Mice and the American opossums. The 

 above indicates that the subdivision of the marsupials is a merely 

 arbitrary one, based on Australian forms, and that the American 

 forms can only with difficulty be squeezed into the arrangement. 

 This first support for the Antarctic theory is therefore far from 

 being convincing. f'~*~ 



A second series of evidence is 'the occurrence among Ameri- 

 can fossils of forms recalling definitely Thylacinns and Dasyurits. 

 Beddard, however, mentions that these so-called marsupials 

 show no lack of ossification of the palate, and that their inclu- 

 sion amongst marsupials is oi doubtful validity. 



Thirdly, if is said that there' are' reasons to believe "that 

 the Diprotodonts are not to be recognised with certainty amongst 

 the earlier European marsupials, so that migration could not 

 have occurred by the Northern route. It may no doubt be un- 

 certain whether the Plaeiaulacidie were Diprotodonts, but if we 

 take this group (the Diprotodonts) as merely a , type adapted 



