CLASSIFICATION OF THE THERIQDONTIA. 75 



Case in Dvmetrodon, occurs in all the more primitive Anomodonts 



and is still of uncertain function. In all Pelycosaurs the septo- 

 maxilla is a small bone resting on the maxilla and premaxilla at 

 the back of the nostril, being touched by the nasal and forming 

 with the other bones the foramen, which I propose to call the 

 septomaxillary foramen. 



In Dimetrodon the lower edge of the septomaxilla, where it 

 rests on the maxilla, is turned inwards and forms a partial floor to 

 the nasal passage. The anterior border of the septomaxilla is 

 provided with a. process which partially divides the external 

 nostril into a lower and an upp^r part. 



It is probable that the process on the anterior edge of the 

 septomaxilla is associated with the original turbinal, simply 

 forming the anterior end of that ridge. In the living lizards and 

 snakes, the septomaxilla lies inside the nostril as it does in the 

 Captorhinids, the collateral ancestors of the Pelycosanrs ; it has 

 in them a characteristic and uniform situation, in that it is 

 ossified in the membrane dividing the main nasal cavit) 7 from 

 Jacobson's organ, running nearly horizontally from the maxilla 

 to the cartilaginous nasal septum. 



In Anomodonts I have only heard one possible suggestion for 

 the function of the septomaxillary foramen, that it served as an 

 outflow for the ductus naso-lachrymalis, the liquid poured out 

 from it serving to keep the muzzle wet as in Artiodactyls. This 

 view is in harmony with the known position of the duct in early 

 amphibia and reptiles. It, however, does not afford any satis- 

 factory explanation of the great size of the foramen in Gorgonops. 



The very peculiar conditions in the Deinocephalian Mormo- 

 saurus (Proc. Zool. Soc. 1914, p. 757, figs. 1 & 2) suggest another 

 explanation. In this animal the ordinary external nostril, which 

 in early reptiles always lies between the premaxilla, the septo- 

 maxilla, and nasal, appears to be represented by a minute foramen 

 between the septomaxilla and the nasal. The large opening 

 which is the functional nostril seems to be really a septomaxillary 

 foramen, as it lies between the premaxilla, the maxilla, and 

 septomaxilla, a situation which is never occupied by the ordinary 

 nostril in early reptiles, but agrees with that of the septo- 

 maxillary foramen in the earlier Theriodonts. Thus this foramen 

 must be of the nature of a nostril. The characteristic position 

 of the septomaxilla in Squamates suggests that the foramen leads 

 into Jacobson's organ, and it will follow that that organ was 

 the functional olfactory organ of Mormosaurus ; the shallowness 

 and small size of the upper part of the nasal cavity, which 

 disi inguishes Mormosaurus from such Deinocephalia as Moscho'ps 

 with a normal septomaxillary foramen and nostril, can thus be 

 accounted for. 



The advances over Varanosauriis which are shown in the nose 

 of Gorgonops are: — (1) the direction of the nostril forwards 

 instead of laterally, a change which renders the appreciation of 

 odours coming from the direction in which the animal is pro- 



