CYNODONT REPTILES GOMPnOGNATHUS AND CYNOGNATHUS. 203 



condylar prolong.ation is damaged beneath and at its posterior end, 

 whose transverse section shows a triangular shape, the articular 

 surface of the condyle cannot be certainly established. On the 

 left side (text-fig. 4) only a part of the prominent backwai-d condylar 

 prolongation is preserved, so that on the outer side of the left 

 jaw the hind end of the articular bones seems to go farther back 

 than the hind end of the dentary, while on the right side the 

 reverse seems to take place. But the hind end of the articular 

 bones on the right side is evidently damaged, so that it is not 

 impossible that the dentary reached As far back as the articular 

 bones, and that a double articulation really did take place in 

 Gomphog7iathus*. 



The possibility of the double articulation in Gomphognaihus 

 follows also from the fact of two different articular surfaces 

 preserved in our specimen on the left side (comp. text-fig. 8). 

 A comparison with the corresponding region in the type skull of 

 Diademodon described by Watson (comp. fig. 3 in Watson 1911) 

 shows (comp. also fig. 8 in Watson 1911 representing the quad- 

 ra'te in Goinphognathus jr,olyp]iagus) that only the triangular 

 hollow on the front face of the squamosal, in which the missing 

 quadrate was received, is preserved in our specimen on the left 

 side together with the two notches which received the two pro- 

 cesses of the quadi-ate (these notches being filled with matrix). 

 As there is in our specimen on the left of these notches and 

 in a more forward position another flat surface on the squa- 

 mosal, which lies in the same direction as the hind end of 

 the dentary (comp. text-figs. 1 and 8), it is quite possible that 

 into this flat surface was received the hind end of the condylar 

 prolongation of the dentary. Indeed, as the articular end in 

 our specimen did not reach the squa,mosal on the external side 

 of the dentary as in Gomphognathus kannevieyeri (comp. fig. 1 in 

 Seeley (1) 1895, p. 5, and especially fig. 2 in Broom, 1904, 

 pi. xxxv.), or as in Cynognathus (comp. fig. 8 in Seeley (2) 1895, 

 p. 81, and fig. 1 in Broom, 1904, pi. xxxv.), so the flat surface in 

 question could be reached in our specimen only by the hind end 

 of the dentary (supposing that it was reached by the lower jaw 

 at all). 



The third point is a corolloTy of the second. As the posterior 

 edge of the left dentary of GompJwgnathus, extending from the 

 angle up to the condylar prolongation, is almost undamaged 

 (comp. text-fig. 4), whilst the corresponding posterior edge of the 

 right dentary is not inconsiderably damaged, the difference in the 

 extent of the covering of the articular bones by the dentary from 

 outside in Gomphognathus and Cynognathus is strikingly shown 



* The length of the two dcntaries of Gomjiho/fnafhus, as they are preserved, is 

 almost the same (149 mm.), when measured from the hind point of their symphj'sis. 

 But a certain asymmetry of them is not iniprohahle, hrcause when we compare the 

 length of the upper edge of the left dentary wiih the corres])onding line in the 

 coronoid process of the right deutar3-, we tind the first to he 63 and the second 

 66 mm; 



