888 DR. C. F. SONNTAG ON THE CLASSIFICATION 



devoid of iobate thickening in the Phascolarctidse. This, in my 

 opinion, is a character of importance, for it does not appear to 

 depend on modes of life. Pocock has also shown how the various 

 genera can be separated by the auricular characters. As regards 

 the mammae the Phascolarctidse have only two, but the Phalan- 

 geridfe have four. Pocock found two only in Trichostorus. The 

 tail is a well-developed organ in the Phalangeridse, and rudi- 

 mentary but muscular in the Phascolarctida3 ; probably the 

 reduction is the result of life habits; The Koala clings only by 

 its hands and feet to the branches, so the tail would be useless ; 

 and a tail would be in reality an impediment to the Wombat, 

 The Phalangeridse have vibrissse on the wrist and hind foot, 

 but no traces of these tactile hairs are present in the Phasco- 

 larctidpe. Moreover, the integuments of the hind feet differ. 

 In the Phalangeridse the skin is striated over the well- 

 marked pads, but it is granular on the reduced pads in the 

 Phascolarctidse. 



Myology/ : — In the Phascolarctidse there is no cleido-occipital, 

 the omo-trachelian is single, and the unbi-oken deltoid gives a slip 

 to the forearm. In the Phalangeridse, on the other hand, the 

 cleido-occipital is present, the omo-trachelian is double, and the 

 deltoid, which is in two parts, does not give a slip to the forearm. 

 Although these are prominent differences, they are of very minor 

 importance in classification. 



Circulatory System: — In the Phascolarctidse the pericardium 

 adheres to the diaphragm, the post-caval vein does not conceal the 

 abdominal aorta, and anterior jugular veins are absent. In the 

 Phalangeridee the pericardium is connected to the diaphragm by 

 two membranous sheets, and the post-caval vein conceals the 

 posterior part of the abdominal aorta. I found anterior jugular 

 veins in IVichosurus and Phalanger, but not in Pseudocliirus ; 

 perhaps they exist in the latter, for my specimens had been 

 preserved so long in spirit that the veins, if present, had shrunk 

 to minute proportions. 



Alimentary Canal: — Cheek-pouches are absent in the Phalan- 

 geridse, but present in the Phascolarctidse ; they are well-marked 

 in Phascolarctos, but rudimentary in Phascolomys. I believe, 

 however, that these diverticula are, in reality, adaptive in function. 

 The stomach is simple in both families, but possesses the so-called 

 gastric gland in the Phascolarctidse. That structure has been 

 regarded by many zoologists as of special value for showing the 

 close relationship between Phascolarctos and Phascolomys. Indeed , 

 Forbes (3) regarded it as the main test. But Johnstone (16) 

 showed that it is no special gland ; it is an evagination and folding 

 of the mucous membrane plus an extension of the normal gastric 

 glands. It is in reality a means of increasing the secreting area 

 of the stomach to fulfil the demands imposed by a bulky diet. 

 The Phalangers do not consume a bulky diet, so cheek-pouches 

 are absent, and there is no trace of the gland patch. Unless 

 there were other characters showing a close relationship between 



