August 9, 1901.] 



SCIENCE. 



195 



the bodies or centra of two vei'tebrse, in 

 such a way as to suggest that its proper 

 place was between them ; that is, inter- 

 central in position. This condition has led 

 Cope and others to the view that a verte- 

 bral element, the intercentrum, once ex- 

 isted here as it does in many lower verte- 

 brates, and that the rib formerly articulated 

 with this. By the disappearance of this 

 intercentrum the mammalian relations have 

 been brought about. In some of the therio- 

 morphous reptiles the capitular head is also 

 intercentral. 



In many mammals there is a small hole, 

 with the disproportionately large name of 

 entepicondylar foramen, in the inner lower 

 end of the humerus, the bone of the upper 

 arm. This opening is for the passage of the 

 brachial artery and the median nerve. In 

 the lower vertebrates such a foramen is 

 unknown, except in the theriomorphs. 



Again, in the mammals the lower jaw ar- 

 ticulates directly with the cranium by 

 means of a shallow pit, the glenoid fossa, 

 on the ventral surface of the squamosal 

 portion of the temporal bone, no other ele- 

 ment intervening between the two. In 

 most of the lower vertebrates the lower jaw 

 does not articulate direct with the cranium, 

 but a movable bone, the quadrate, is in- 

 serted between them, and forms a suspensor 



Fig. a. Skull of a Theriomorphous Reptile {Gor- 

 donia). Showing the quadrate bone {q) firmly united 

 to the enlarged squamosal (s). After Newton. 



of the lower jaw. In the theriomorphs and 

 crocodiles the quadrate is fixed and immov- 

 able, and is held in position by a second 



bone, the squamosal, which, together with 

 the quadrate, takes part in the formation 

 of the articulation of the lower jaw. Now, 

 say the advocates of the reptilian ancestry 

 of the mammals, if the quadrate were to 

 become completely fused with the squa- 

 mosal, the result would give a condition 

 from which the mammalian articulation 

 could readily be derived. In support of 

 this view, they cite the case of a human 

 skull, described by Albrecht, in which a 

 separate bone, which he interpreted as the 

 quadrate, appeared in this very region. 



Further evidence, which is regarded as 

 , pointing in this same direction of a therio- 

 morphous ancestry, is furnished by the 

 pelvis, while the imperfectly known tarsal 

 bones of these reptiles are doubtfully inter- 

 preted as supporting the same view. These 

 features, however, are of secondary im- 

 portance in comparison with those already 

 enumerated, for the peculiarities of the 

 mammalian pelvis and tarsus are as readily 

 derived from the amphibia as from the 

 theriomorphs. The foregoing enumerates 

 the chief osteological evidence for the rep- 

 tilian ancestry of the mammals. The only 

 other evidence recalled at present which 

 points in the same direction is the character 

 of the segmentation of the monotreme egg, 

 already alluded to. 



Within more recent years there has been 

 a tendency upon the part of some zoologists 

 to return to the support of an amphibian 

 parentage of the mammals. Klaatsch, 

 Maurer, Hubrecht and Beddard have 

 pointed out features of the soft parts of 

 mammals, which are more easily inter- 

 preted by this assumption ; but, of course, 

 this cannot be conclusive, for we can know 

 nothing of these structures in the therio- 

 morphs. Hence this evidence, which will 

 be summarized later, can only be regarded 

 as cumulative and not of first importance. 



The osteological facts which have already 

 been enumerated need analysis, for it is 



