24 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 76 



of which, in Agorophius at least (the skulls of the two other animals 

 are imperfect in this region), is equal to about one-third or one-fourth 

 the greatest width of the skull. (The entire braincase of Xenorophus 

 is lost, but the essential agreement with the two other genera is shown 

 by the structure of the orbits and the interorbital region.) In this 

 character there is a close analogy with the condition present among 

 the Cetothcriidcc. The maxillary in two of the known specimens is 

 imperfect. It is sufficiently well preserved, however, to show that in 

 Agorophius and Xenorophus it had already overridden the frontal in 

 very much the same manner as that which is normally seen in por- 

 poises, and to such an extent that it overlaps the anterior edge of the 

 temporal fossa, while in Archcvo del phis it has apparently not pushed 

 backward quite to the level of the posterior border of the orbit. In 

 ArchcEodelphis the maxillary is described and figured as forming 

 part of the anterior orbital rim, a generalized mammalian feature not 

 known in any other odontocete. In Xenorophus the maxillary is 

 excluded from the orbital rim by the jugal. Which of these two 

 conditions existed in Agorophius the type specimen was too much 

 damaged to show. Other primitive features known to occur in 

 Archccodelphis and Xenorophus but not demonstrable in the figures 

 of Agorophius are connected with the narial passage. This pas- 

 sage slopes backward essentially as in the baleen whales ; the proxi- 

 mal ethmoid region is roofed by the nasals and the median portion 

 of the frontals ; the palatines form no part of the anterior wall 

 of the narial passage, thus agreeing with the structure present in 

 normal mammals, baleen whales, and physeterine toothed whales. 

 While it appears unlikely that a family existing so recently as the 

 upper Eocene or lower Oligocene could have been genetically an- 

 cestral to any considerable number of cetacean types, two of these 

 early odontocetes seem clearly to represent morphological stages of 

 development through which the ancestors of some of the modern 

 toothed cetacea might have passed. I think it can be appreciated 

 after comparing the skull of Agorophius (pi. 5, fig. 2) with that of 

 recent Dclphinus (pi. 5, fig. 4) that a structure like that which is seen 

 in the living delphinoids might not improbably have been developed 

 from the one present in the fossil by a process which consisted 

 primarily in a forward movement of the occipital region until the 

 supraoccipital came in contact with the upper margin of the frontal, 

 and the anterior extremity of the articular portion of the squamosal 

 arrived at a point beneath the tip of the postorbital process. Similarly 

 the relationship of the maxillary to the orbit and frontal in Archceo- 



