NO. 8 INTERRELATIONSHIPS OF THE CETACEA WINGE 33 



difference is only slight, but it finally increases so as to become very 

 noticeable. At the same time that the anterior facial part of the skull 

 is shortened, because the mouth is no longer used as a pair of forceps 

 but as a " clap-trap," it becomes flatter and broader, while its upper 

 side is more pressed upon by the facial cushion. The cushion becomes 

 larger, especially widening itself out anteriorly and pushing into the 

 originally slender " beak." The intermaxillaries, in their anterior 

 portion particularly, together lose their structure as an upstanding 

 roof-ridge, and finally become quite flat, each of the bones widened 

 out. 



The genera of the section Legenorhynchi depart so slightly from 

 the more primitive members of the section Delphini, such as Pro- 

 delphinus, that there would scarcely be any reason to set them apart 

 in a special group were it not evident that they represent the begin- 

 ning of new series of forms. 



Doubtless Tursiops occupies the lowest position. The anterior 

 facial part of the skull is indeed broader than in Prodelphinus, but 

 it has, however, not lost its form as a roof-ridge, and it has still a 

 considerable length. 



Near Tursiops probably belongs Tursio [Lissodelphis^, which 

 also has the fore-face rather long, though more flattened. Another 

 difference is that it lacks the dorsal fin, either because it has lost it or 

 has never acquired it. 



Lagcnorhynchus (to which should probably be joined Cephalo- 

 rhynchus and Sagmatias, and perhaps " Feresa") has gone a step 

 further than Tursiops and Tursio in the direction of shortening and 

 flattening the rostrum. 



Among the Delphinids in which the process of shortening and 

 flattening the rostrum has been more perfected the members of the 

 section Globicipites are contrasted with those of the section Phocsenae 

 by reason of their greater primitiveness. In them the crowns of the 

 teeth have retained their primitive conical form, while in the Pho- 

 caenans the crowns have become entirely peculiar. 



Orca l^Orcinus'] is the one among the Globicipites which has 

 retained most of the ordinary dolphin type in the structure of the 

 rostrum, particularly as regards the narrowness of the intermaxillary. 

 The rather short, rounded-off form of the hand might appear to be 

 primitive also, but various circumstances strongly indicate that it has 

 arisen through the shortening of an ordinary, pointed, porpoise 

 flipper : the number of phalanges in the second finger is rather large ; 

 the finger is merely more strongly arcuate than usual. In the trans- 

 formation of the dentition to a conspicuously powerful biting imple- 



