224 Mr. D. M. S. Watson on the 



of which is a deep groove, bounded laterally by a distinct 

 ridge. Immediately external to this on the frontal is a 

 smaller groove and ridge, outside which is a pocket in the 

 prefrontal, along the bottom of which is a ridge. On the 

 side of the nasal cavity outside this pocket are three other 

 ridges, making six in all. 



This series of ridges and grooves is identical in all features 

 with those on the lower surface of mammalian frontals and 

 nasals which are associated with the attachment of ethmo- 

 turbinals. They seem to me to afford conclusive evidence 

 that Nythosaurus and Diademodon possessed a series of 

 unossiKed ethmo-turbinals on the same plan as such a mammal 

 as Dasyurus. 



1 have not yet found similar evidence of the presence of a 

 maxillo-turbinal, perhaps only on account of inadequate 

 material. 



Another feature extremely clearly shown by these two 

 ironstone casts is the course of the naso-lachrymal duct. 

 This opens into the orbit by two foramina in the 

 lachrymal bone, leading into canals, which soon unite and 

 then travel forward still in the lachrymal bone until they 

 open into the nasal cavity on the inner side of the maxilla. 

 These relations are so very similar to those existing in 

 Ptromeles between the duct and the lachrymal bone as to 

 leave no doubt that the bone in Cynodonts is homologous 

 with that of the mammal. It is, however, the lower of the 

 two bones usually called lachrymal and prefrontal, and as 

 the Cynodonts are certainly more nearly allied to the 

 mammals than the lizards, it leaves no doubt that Gaupp 

 was not justified in homologizing the reptilian prefrontal 

 with the mammalian lachrymal. Meek, from the conditions 

 in the Crocodile, has already controverted this view, Gaupp 

 having no doubt been misled by the great reduction of the 

 true lachrymal in the types studied by him. 



The occurrence of a large deep fossa subarcuata is of great 

 importance, because it shows that the brain in its region 

 and at least to its level filled the brain-cavity, as does that of 

 a mammal. If the brain had not filled the cavity there 

 would be no necessity for providing a special diverticulum 

 for the flocculus. 



We are thus justified in regarding the lower part of the 

 cerebellar region of the cast of the cranial cavity as a rela- 

 tive^ perfect copy of the brain. 



The under surface of the medulla is slightly convex in all 

 directions, and is terminated in front by the pituitary, which 



