SOME CARNIVOROUS THERAPSIDS. 1031 
this region in the Hndothiodon-zone Gorgonopsid Scymnognathus 
are obviously prophetic of those of Cynognathids, it will follow 
that the Bauride have had a quite different ancestry since that 
time, which is long anterior to the development of a ‘‘Cynodont ” 
structure—7. e, a secondary palate in any form. 
This conclusion, if true, will necessitate the splitting up of the 
Order ‘* Cynodontia,” and the recognition that it includes at least 
two distinct branches which were separate when they indepen- 
dently acquired a secondary palate. 
The detailed structure of the palate of any Gorgonopsid is 
not known beyond the possibility of doubt. 
All Gorgonopsids which I have seen agree in the following 
features :— 
lst. There is a long region in front of the basisphenoid in 
which the two pterygoids are in contact, apparently with a 
median parasphenoid, but perhaps for part of the distance with 
each other, and in which their lateral margins are parallel ; 
that the whole forms a narrow bar behind the powerful descending 
processes of the pterygoids. 
2nd. If an interpterygoid vacuity is present it is very small, 
and lies just between the descending flanges of the pterygoids. 
3rd. There is a deep groove down the middle of the posterior 
part of the palate. 
Ath. There are no suborbital vacuities. 
5th. There are very large posterior nares, which are separated 
by a bar lying considerably above the level of the lower edges of 
the maxille, which tend to be approximated. 
No skull I have seen shows all the sutures on the palate satis- 
factorily. The skull of Arctops shows many of them, but most 
unfortunately, owing to its having been split longitudinally, 
those in the middle are not clearly visible posteriorly. 
The type-skull of Goryonops shows many sutures in a rather 
indefinite manner. 
As Dr. Broom has pointed out. the bar which divides the pos- 
terior nares in Gorgonops and many othe: forms is actually single. 
It has to be considered whether, like the single *‘ vomex” of the 
Anomodonts, it 1s really composed of a pair of fused prevomers, 
or whether it is a parasphenoid. Its complete resemblance in 
shape, even to the grooves on its upper surface for Jacobson’s 
organs, to the undoubted prevomers of many Therocephalia, 
suggests that it is to be interpreted in the same way; and in the 
type-skull of Arctops it actually seems to be double when seen in 
section on the anterior end of the specimen, about one centimetre 
in front of the posterior end of the posterior nares. In this 
specimen, in place of the single dorsal ridge which occurs in 
Scymnognathus, there are two recs than a millimetre apart, and 
the groove between them seems to be continued by a suture into 
the palate. If this observation is correct, and the condition of 
the specimen in this region is not good enough for certainty, then 
there can be no doubt Siipats the bar between the palato-nares in 
