796 



SCIENCE 



[N. S. Vol. XXVI. No. 67S 



Witkowski (1896), are Cp = 0.3467 and 

 C^ = 0.2467 cubic-foot atmospkeres. 



' Harvey N. Davis 



Cambeidge, Mass. 



occurrence of three species of beaked 



whales of the genus mesoplodon on 



the atlantic coast of the 



united states 



To THE Editor of Science: The few speci- 

 mens of beaked whales of the genus Mesoplo- 

 don which have hitherto been obtained on the 

 Atlantic coast of the United States have been 

 tacitly assigned to a single species, M. hidens 

 (Sowerby). After a detailed study of the 

 material available, I am convinced that three 

 species are represented. These are : (1) M. 

 hidens, the species most commonly found in 

 the North Atlantic; (2) M. europwus, a species 

 known hitherto only from a single specimen 

 found floating in the English Channel about 

 seventy years ago; and (3) a species which is 

 apparently M. densirostris, of which only a 

 few specimens from the Indian Ocean are 

 known. 



The name M. europwus (Gervais) will prob- 

 ably have to be replaced by M. gervaisi 

 (Deslongchamps). The former specific name 

 was originally published in the first edition of 

 Gervais's " Zoologie et Paleontologie f ran- 

 gaises " (1846-52) under the genus Dioplodon. 

 I have not seen the first edition of this work, 

 but in the second edition (1859) the name is 

 a nomen nudum. The species was apparently 

 first described by Deslongchamps in 1866, who 

 renamed it gervaisi (Dioplodon gervaisi). 



An account of the American specimens of 

 this and other genera of ziphioid whales in 

 the National Museum is in preparation, and 

 will probably be published in a few months. 



F. W. True 



National Museum, 

 November 5, 1907 



SPECIAL ARTICLES 



SOME recent advances in south AFRICAN 



paleontology 

 During the past year a considerable number 

 of new fossil reptiles have been discovered in 



the Permian and Triassic deposits which are 

 included in the Karroo formation. Three new 

 Therocephalian genera have been discovered 

 in beds which are probably Middle Permian. 

 These are all more or less closely allied ta 

 others previously known, and do not reveal 

 much that is new in the general anatomy of 

 the group. In beds which are probably Lower 

 Triassic a new type of Therocephalian has 

 been discovered and named Arnognathus. 

 Unfortunately, it is only represented by a 

 dentary bone. The discovery is chiefly inter- 

 esting from the fact that no Therocephaliaa 

 has hitherto been got in beds more recent 

 than Upper Permian. It is rather a degen- 

 erate form than one leading on to the special- 

 ized Cynodonts. 



Very much more important than these is 

 the discovery in Lower Triassic beds of the 

 nearly complete skeleton of a small reptile 

 which must be placed in a new suborder. 

 Hitherto we have had no evidence throwing 

 any light on the origin of the Therocephalian 

 reptiles. Pareiasaurus and some other of the 

 Cotylosauroid forms seemed to suggest a pos- 

 sible origin among these early types, while 

 one or two points in Mammalian morphology 

 and embryology such as the mammal-like type 

 of the organ of Jacobson in Sphenodon 

 seemed rather to favor an affinity with the 

 Rhynchocephalians. The discovery of this 

 new fossil reptile, which is called Galechirus, 

 strongly favors the descent of the Thero- 

 cephalians from an early Ehynchocephaloid 

 ancestor. This new type is about the size of 

 a rat and the following are its most impor- 

 tant characters: the dentition is thecodont 

 and homodont, there being no enlarged ca- 

 nine; the lower jaw is very similar to that 

 of the Therocephalians, but there is no en- 

 larged coronoid process; the shoulder girdle 

 is typically Therocephalian, the precoracoid 

 being weU developed; the digital formula is 

 2, 3, 3, 3, 3; the pelvis is plate-like with the 

 ilium directed somewhat backwards ; and well- 

 developed abdominal ribs are present. Un- 

 fortunately, the temporal region and palate 

 are unknown. The afiinities are undoubtedly 

 mainly with the Therocephalians, but in none 



