444 Bulletin American Museum of Natural History. [Yo\. XX\'II, 



donts, in Pantolamhda and in Euprotogonia, but also in all the more primi- 

 tive existing Inseetivores and Rodents. The ancestral Placental probably 

 had another character of the greatest importance: namely, the lunar in 

 the front view rested on the centrale, on the magnum and perhaps partly on 

 the unciform. 



The contact between the lunar and unciform may have been established 

 as far back as the Triassic, as already noted. In varying degrees it persists 

 in the majority of the Marsupials, Edentates, Rodents, Inseetivores, Creo- 

 donts, Fissipedes, Tillodonts, Primates, Condylarths (Euprotogonia), Ani- 

 blypods (Pantolamhda), Notoungulata, Perissodactyla, Artiodactyla, Sirenia 

 and Cetacea. The principal orders in which the lunar-unciform contact is 

 o-reatly reduced or absent are the Tubulidentata, Hyracoidea, Embrithopoda 

 and Proboscidea and it is very probable that its reduction in these orders is 

 secondary (see below). The lunar-unciform contact and the existence of a 

 separate centrale produces the alternating or interlocking arrangement, as 

 fii-st noted by Matthew (1897, pp. 299, 308). The idea that this interlock- 

 ing arrangement in the Unguiculates is largely secondaiy arose from the 

 unwarranted extension to the carj.ais of Unguiculates of Cope's famous but 

 untenable theory of the evolution of the carpus of Ungulates (cf. pp. 449, 451). 



The truly primitive nature of the so called displaced carpus of the Creo- 

 donts was first pointed out by Matthew (1897, pp. 308-309) who also 

 advanced the theory (1904, pp. 811-814) that the primitive Placentals had 

 been derived from pentadactyl forms in which the pollex and hallux were 

 divergent and more or less grasping in function. The evidence for the 

 primitive nature of the interlocking carpus' in the various orders of Placentals 

 may be reviewed briefly. 



Insectivora. It has already been shown that the Insectivora as an order 

 are exceedingly primitive in many ways. The lunar-unciform contact and 

 a free centrale are preserved in the majority of the forms. x\mong the 

 Zalambdodonta, Centetcs and Ericulus show these characters, but in the 

 otherwise primitive Microgale (figured by Leche, 1907, p. 80) the scaphoid, 

 lunar and centrale coalesce and the internal inferior angle of the cuneiform 

 is thrust in between the hmar and the unciform. In Oryzoryctes this tend- 

 encv is greatly emphasized so that the scapho-lunar-centrale is widely 

 separated from the unciform. In all these Zalambdodonts the trapezium 

 is relativelv large, as it is in Creodonts, and the pollex is more or less diver- 

 gent. The carpus of Zalambdodonts is reviewed more in detail above 

 (pp. 241-251). In the Eocene and Oligocene Leptictidse the manus is not 

 known. In Erinaceus and Gymniira the lunar-unciform contact is preserved. 

 In Talpa, perhaps owing to the necessity for o compact and strong carpus, 

 the interlocking features are emphasized in the oblique facets, and the 



