256 Bulletin American Museinn of Natural History, [^'^ol. XX\'II, 



resemblance extends in far more diverse directions than might have been 

 expected, including such characters common to both groups as the following: 

 (1) Iridescence of the hair; (2) presence of a nasal shield; (3) general 

 configuration of the skull; (4) presence of a "third lower arm bone" (an 

 ossification in the tendon of the flexor digitorum profundus muscle — only 

 the distal end ossified in Notoryctes); (5) peculiar specialization of the 

 claws (different digits, however, are concerned in the tw'o cases); (6) great 

 reduction of the thyroid ("obturator") foramen; (7) insertion of the latis- 

 simus dorsi muscle; (8) form of the brain (extraordinarily similar). 



To this list of convergent characters may be added, somewhat doubt- 

 fully: (9) similarity in the dentition, especially in the molars. This character 

 Leche omits because "ein gemeinsamer Charakter sehr alten Datums sein 

 kann, und nicht der Konvergenz seine Entstehung zu verdanken braucht." 

 While very true as a general proposition, this dictum is very possibly not 

 applicable here. Bensley (1903, p. 120), accepting the view that in Chryso- 

 chloris the main internal cusp is equivalent to the combined para- and meta- 

 cone of Talpa, suggests that in Notoryctes also the main internal cusp is 

 homologous with the combined para- and metacone of Didelphids. There 

 are however reasons for doubting (pp. 238, 290) this assumed cusp homology 

 even in Chrysochloris. But at any rate, in view of the probability that Noto- 

 ryctes is a highly specialized ofi'shoot of the Polyprotodont stock, it seems 

 likely that the total lack of a talonid in the lower molars of Notoryctes is a 

 secondary condition and that the spacing of both upper and lower molars is 

 an indication that formerly the anteroposterior diameters of both ui)per and 

 lower molars were greater than at present, and that therefore the resemblances 

 between the molars of Chrysochloris and those of Notoryctes are in part due 

 to convergent or parallel modifications from somewhat similar beginnings. 



(10) A further resemblance lies in the fact that in Notoryctes the num- 

 ber of incisors are reduced to | while the total number of teeth is the same 

 as in Chrysochloris (Leche, 1907, p. 58). 



Several additional resemblances between Notoryctes and Chrysochloris 

 appear in the skeleton, e. g., humerus with entocondyle greatly produced; 

 olecranon very long and hook-like. Leche regards this as the most complete 

 case of convergence known among the higher animals. 



The following observations may now be added in order to reinforce 

 Leche's conclusion that there is at most only a very remote (subclass) genetic 

 connection between Notoryctes and Chrysochloris: 



(1) The skull of Chrysochloris presents only a superficial analogy with 

 that of Notoryctes: In Notoryctes the malar is rather stout and appears 

 to be of the normal Polyprotodont type; in Chrysochloris the malar is lack- 

 ing and the zygomatic arch is composed chiefly of the zygomatic process of 



