Neio Fossil Reptiles. 35 



The following are some of the principal measurements of the 

 limb bones : — 



Humerus (sp. A) about 35 mm. 



(sp.B) „ 37 



Eadius (sp. A) ,, 25-5 



„ (sp.B) „ 26-5 



Ulna (sp. B) „ 29 



Metacarpal 1 (sp. A) ,, 5 



2(sp. A) „ 7-3 



3 (sp. A) „ 9-5 



4(sp. A) „ 11-2 



' „ 5(sp. A) „ 7-5 



Femur (sp. A) ,, 38-5 



Tibia (sp. A) „ 37 



In many respects this little animal is the most important find that 

 has been made in South Africa since the complete skeleton of 

 Oudenodon was discovered in 1900. Hitherto we have been quite in 

 the dark as to the origin of the Therocephalians. Pareiasaurus 

 seems to be in some respects intermediate between them and 

 the Cotylosaurians, but it is too specialised and too dissimilar 

 to throw much light on the origin of any of the mammal-like 

 reptiles. The Anomodonts and the Dinocephalians both show 

 affinities to the Therocephalians, and the Cynodonts to the Anomo- 

 donts, and the primitive Monotremes were probably not much 

 further removed from the Cynodonts than the various suborders of 

 the Therapsida from each other, but how the whole of the mammal- 

 like group arose has till now been quite obscure. The discovery that 

 Sphenodon has an organ of Jacobson of a very mammal-like nature 

 suggested the possibility that the point of connection between the 

 Diapsidan and Synapsidan phyla was perhaps not quite so remote 

 as the Cotylosaurians. The finding of this new fossil form 

 Galechirus also points in the same direction. 



• Galechirus is more closely allied to the Therocephalians than 

 to any other known reptile, but it is distinctly more primitive than 

 any of these reptiles yet known. The absence of an enlarged 

 canme and of any distinct coronoid process mark Galechirus sharply 

 off from the Therocephalians and suggest rather an affinity with the 

 Anomodonts. Hitherto I have supposed that the Anomodonts have 

 sprung from some Therocephalian ancestor, and that the coronoid 

 process of the latter had been lost in the former by change of habit, 

 just as the coronoid process has become lost in the Monotremes and 



