442 Bulletin American Museum of Natund History. [WA. XX\'II, 



mals, bears digits IV and V. The trapezium (1) is also much larger than 

 the trapezoid (2), the second metacarpal is thrust up slightly between meta- 

 carpal 1 and metacarpals as in primitive mammals (Osborn). The relations 

 of the radiale, intermedium and centrale 1 j)recisely foreshadow the coalesced 

 scapho-lunar-centrale of the embryonic Echidna (Fig. 28, no. 8). Centrale 



I occupies the position which it does in the manus of Eocene Creodonts, 

 Pantolamhda, recent Insectivores, Rodents, etc. By the absorption or dis- 

 appearance of the minute centrale 2, the lunar (intermedium) would be 

 brought into contact with the unciform (4) and thus the alternating and 

 falsehj called ''displaced" type of carpus is foreshadoived even in the Triassic. 



The manus of the Cynodontia (reptiles which as shown above resemble 

 the carnivorous Marsupials in many characters) is unfortunately not well 

 known. Broom has figured the lower j)art of the carpus with the meta- 

 carpals of Mlnrosuchus and Seeley (1895, pi. i, fig. 7) figures the imperfect 

 manus of Microgomphodon (Fig. 7). The "thrust up" position of digit 



II is again noticeable. Otherwise there are no very important differences 

 from the type represented in the supposed Therocephalian Theriodesmus. 

 No trace of the "pmehallux" appears. 



The above described manus of Permian and Triassic Therapsids might 

 be expected to confirm or disprove the following ingenious suggestions, the 

 arguments for which have been summarized by Matthew (1904, pp. 814- 

 815): first that the trapezoid of mammals, which is unusually large in cer- 

 tain Eocene Creodonts, may be the serial homologue not of its fellow the 

 trapezium, but of the metacarpals; secondly that the so called metacarpal of 

 the first digit is really a modified proximal phalanx. This would explain first, 

 the occurrence of only two phalanges on the first digit, secondly, the excep- 

 tional relations of the metacarpal and j)halangeal epiphyses of digit I, 

 thirdly, the anomolous arrangement of the muscles (Matthew). It cannot 

 be said, however, that distal carpale 1 (the homologue of the trapezium) 

 in either Eri/ops, Galeckirus, Opisthoctcriodon, .Elurosnchus, Theriodesmus 

 or Microgomphodon, any more than in the embryonic Echidna and Didel- 

 pliis, is of such a form as to give definite su])port to the hypothesis that it is 

 the homologue of metacarpal II, although in Oudenodon it does suggest 

 such a comparison. Broom (1906) has explained the anomolous relations 

 of the epiphyses in digit I on mechanical (adai)tive) grounds. 



Monotremes. — The manus of the embryonic Echidna (Fig. 28, no. 8), as 

 figured by Emery shows the following very interesting features : 



(1) The distal end of the radius and ulna are now in contact, whereas 

 in all the Therapsida they are separated. This condition was foreshadowed 

 to some extent in Oudenodon (Fig. 28, no. 2) in which the radius had begun to 

 overlap the intermedium and in which (as in Echidna) the ulna was also in 



