MOODIE: TURTLES AND PLESIOSAURS. 323 



cone occurs within the perichondrium. Neither turtles nor 

 plesiosaurs have epiphyses, and there is, hence, no basis for 

 relationship between the two groups so far as these struc- 

 tures are concerned. 



The relations and shape of the pectoral girdle have also 

 been used as a basis for relationship of the groups under dis- 

 cussion. It has been claimed that the biradiate character of 

 the scapula in the two groups is due to the fusion of the pro- 

 coracoid with the true scapula. The fact that the fusion was 

 alike in the two groups is, therefore, taken as a basis of rela- 

 tionship. In neither group, however, has this fusion taken 

 place. I have recently studied the development of the skeleton 

 in the Chelonia and hope to present a full discussion elsewhere. 



Fig. 2. 



The question which interests us here is the condition of the 

 scapula in the early chelonian embryos. In turtles, as has been 

 stated, the scapula is a biradiate element. If it were a com- 

 posite element there should appear, in the early embryonic 

 stages, at or near the center of the shaft of each prong, a 

 center of ossification, or at least a center of calcification in the 

 very early stages. Specimens have been cleared by the 

 Schultze method, recently set forth by Mall, Hill and myself, 

 and in these specimens it has been possible to ascertain the 

 exact condition of affairs. 



The method of ossification in the embryos of reptiles is very 

 peculiar. The bony matter is laid down in long wavy threads, 

 which extend in a cylindrical shaft around the central part of 

 the bone, being densest near the middle of the shaft. The bony 

 fibers in the scapula of Trionyx, as shown in figure 2, do not 

 depart from this plan of formation. 



The fibers are not found to be densest near the center of the 



