41 S Bulletin A?ncrican Museum of Nntvral History. [Vol. XX\"'IT, 



The teeth of Squalodon are morphologically intermediate between those 

 of the Odontocetes and those of the Zeuglodonts. So far as the form of the 

 teeth is concerned, analogy with the PhocidtB would indicate that the cylin- 

 drical teeth of the Odontocetes are a secondary adaptation to marine pis- 

 civorous habits. The great variability in the characters of the teeth among 

 the Odontocetes and among the Pinnipeds indicates that in both groups 

 rapid evolutionary changes in the dentition have been in progress at a 

 relatively recent date. In the ancestors of the Odontocetes probably the 

 form of the teeth may have been rather diverse: some types may have been 

 derived from the SqimJodon type, others more directly from the Protocetus 



type- 



We may, in short, readily agree with Dr. True that the known Zeuglo- 

 dontia differ from the known Odontoceti in many characters, notably in 

 the greater length of the humerus and in the retention of a distinct radio- 

 ulnar trochlea, but if the Cetacea have been derived from land mammals 

 these very characters must be looked for in their immediate, less completely 

 aquatic ancestors. And before the Zeuglodontia as an order are proved to be 

 morphologically not ancestral to the Cetacea it would be necessary to cite 

 evidence to show that the Cetacea had been derived from some very different 

 type of mammals, say for example the Sirenia. 



Assuming then that the Zeuglodontia are offshoots of the ancestral 

 Cetacean stock, the next cjuestion is, from what orders of terrestrial quad- 

 rupeds are the Zeuglodonts in turn derived ? The discovery of the remark- 

 ably primitive species Protocetus atavus led Professor Fraas (1904) to infer 

 that the ancestors of Protocetus must have been Creodonts. To this view 

 Dr. Matthew, in conversation with the writer, at once took exception, on the 

 ground that the characters of the Protocetus skull might be derived as readily 

 from the Eocene Insectivore type represented by Pantolestes as from the 

 Creodont type. Among the facts which tend to support Dr. Matthew's 

 view the following may be mentioned: 



(1) The elongate rostrum of Protocetus is more readily derived from the 

 Insectivore type, with the incisors arranged in an antero-posterior series, 

 than from the Creodont type. The elongate rostrum and antero-posterior 

 arrangement of the incisors are exceedingly characteristic of the Cetacea. 

 If the latter had been derived from the Creodont type having a transverse 

 incisor series it is likely that Protocetus would have showed some traces of 

 this arrangement, and that the end of the snoiit would have been analogous 

 to that in the Gavial. 



(2) The sharply triangular outline of the skull in palatal view, together 

 with the relatively weak malar bones is more consistent with derivation from 

 an Insectivore type than from a Creodont type. The very large attachments 



