3 io A BOOK OF WHALES 



existing whales have no lot or part in forming the top 

 of the skull, meet in these ancient whale-like creatures 

 to form a sagittal crest upon the vertex. The 

 cervical vertebrae, as in the ancient Platanistidae 

 and in a few only of other existing Cetacea, are 

 separate ; they are, moreover, not compressed antero- 

 posteriorly as are those of recent whales, but are not 

 different in length from the succeeding vertebrae. 

 The scapula is not typically Cetaceous, since it has 

 but a small coracoid process and a large acromion. 

 The ribs are double-headed, like the anterior series 

 of the toothed whales. The sternum too is con- 

 structed upon the plan that characterises the Odonto- 

 ceti, being composed of several pieces. 



If the Archaeoceti are the most primitive of whales, 

 it must be among them that the clue to the relation- 

 ship of the whales will be found. This is a topic, 

 however, about which more has been written than 

 ascertained. The only view that demands a notice 

 here (for we cannot, of course, accept any Ichthyo- 

 saurian descent for these animals) is the opinion held 

 by one or two persons that the Zeuglodonts are most 

 nearly related to seals. 



The facts upon which these comparisons have been 

 based are principally the characters of the teeth, the 

 long neck "like that of a seal in proportions" -the 

 scapula without the typical whale-like form. All these 

 points are just so far seal-like as they are generalised 

 characters. All mammals except the Cetacea, and to 

 a less extent the Sirenia, have moderately long cer- 

 vical vertebrae. Included in this series are, of course, 



