298 



SCIENCE. 



[N. S. Vol. XXI. No. 530.. 



The author further stated that he phieed 

 the Como beds as equivalent to the 

 Wealden, and as representintr either the 

 upper or middle part of the Lower Cre- 

 taceous. 



Dr. E. C. Case read on 'Characters of the 

 Chelydosauria.' 



The Pemiian genus Diadcctes was shown 

 to be a member of the order Chelydosauria 

 of Cope. This order was founded on the 

 genera Otocoflus and Conodectes and was 

 considered as ancestral to the turtles. 



The discovery that Diadectcs is a mem- 

 ber of this order permits a more complete 

 description of its characters from espe- 

 cially perfect specimens collected by the 

 author of the paper. 



The members of the Chelydosauria 

 (Diadectidje and Otoccelid^e) differ from 

 the Cotylosauria (Pareiasauridas, Parioti- 

 chidte, Elginida3( ?) ) by the following char- 

 acters ; and in the same characters approach 

 the Testudinata. 



1. There are three pairs of openings 

 through the roof of the skull instead of 

 two; the extra pair being the openings of 

 the meatus auditus externus. 



2. The palate is very degenerate and the 

 transverse bone is lost or nearly so. 



3. The form of the quadrate and its rela- 

 tions to the surrounding bones are directly 

 comparable to those of the turtles. 



4. There are no teeth on the pteiygoids 

 and palatines. 



5. There is no anterior process (para- 

 sphonoid) on the basisphenoid bone and 

 correlated with this. 



6. There are no prevomers but a single, 

 anteriorly placed vomer (parasphenoid?). 



7. The internal carotid arteries do not 

 penetrate the lower surface of the basi- 

 sphenoid, as in the h'li ij ncoccpJudla. but en- 

 ter from the side as in many turtles. 



8. The presence of paired descending 

 plates from the under side of the parietal 



and the possible absence of the epiptery- 

 goid. 



9. The presence of dorsal plates over- 

 Ijing the dorsal ribs and the presence of 

 eighteen presacral vertebras (the last com- 

 mon to Cotylosaui'ia ) . 



There is no trace of a beginning plastron. 

 It is concluded that the characters of the 

 skull establish the validity of the order 

 Chelydosauria and indicate the direct 

 origin of the ti;rtles from the Cotylosauria. 

 On Eocene Insectivora and on Pantolestes 

 in parlicxdar: W. D. MATTHEVi\ 

 The genus Pantolestes Cope, variously 

 referred to the Primates, Creodonta and 

 Artiodaetyla, turns out to be a fossorial in- 

 sectivore of an archaic and peculiar type. 

 A well-preserved skull and jaws and a 

 large part of the skeleton were obtained by 

 the American ]\Iuseum Expedition in the 

 Bridger Basin last summer. The teeth 

 resemble those of the most primitive 

 creodonts, the skull is most like that of the 

 Centetida^, and iinlike the creodont or con- 

 dylarth skulls; the limbs and feet are 

 specialized for digging, more than in the 

 badger but less than in the moles; the tail 

 is long and very massive, the postsacral 

 vertebrfe being larger than the presacrals. 

 Palceosinopa Matthew is the Lower Eocene 

 ancestor of Pantolestes ; Pentacodon Scott, 

 of the Basal Eocene, appears to be a related 

 genus. The addition of the Pantolestidae 

 and of some undescribed genera recently 

 discovered, and also, if Wortman's asser- 

 tion is supported by evidence, of the 

 Hyopsodontida?, to the list of Eocene In- 

 sectivora, greatly increases the importance 

 of that order among the IMammalia of the 

 early Tertiary. The present distribution 

 of the Insectivora indicates that they must 

 once have been an abundant and varied 

 group ; ])ut fossil insectivores have hitherto 

 been quite rare and for the most part 

 nearly related to the three common living 

 families, the hedgehogs, moles and shrews. 



