260 Annals of the South African 



plementary ossicle," and part of the preopercular. The hyomandi- 

 bular measures 20 mm. in length, and is constricted in the middle. 

 It supports a very small quadrate, and there appears to be a distinct 

 symplectic. The preopercular lies along the upper side of the 

 maxilla. The squamosal lies above the opercular, the hyomandi- 

 bular, and the preopercular. Behind it is a narrow supratemporal 

 dividing it from the post-temporal. The sensory canal is continued 

 from the lateral line over the siipraclavicle and can be seen crossing 

 the squamosal. The maxilla and the dentary are well developed 

 and have two series of teeth, an outer row of small irregularly 

 arranged teeth, and an inner row of large well-spaced laniaries. 



Axial Skeleton of Trunk. The axial skeleton of the posterior half 

 of the larger specimen is well preserved. There is clear evidence of 

 the persistence of the notochord. In the middle region of the body 

 a series of short neural arches are well displayed. The height of the 

 arches is about equal to the diameter of the notochord. In the 

 posterior region the arches are much obscured by the scales. Above 

 each arch is a slender spine about twice the length of the arch. The 

 head of each arch is dilated and the lower end of each spine 

 similarly expanded to articulate with it. The middle part of each 

 spine is slender, but the ends seem to be expanded, but they are not 

 well preserved. The spines can be traced back into the tail. In the 

 region of the dorsal tin a row of short interspinous bones are dis- 

 played, of which eight are clearly seen. Above these and articulating 

 with them is a second row of interspinous bones arranged along the 

 base of the dorsal fin. Of these there are probably 20, and they are 

 all somewhat hour-glass shaped, about three or four times as long 

 as broad. 



In the region below the anterior part of the dorsal fin are seen a 

 few haemal arches. Like the neural arches they are short, and like 

 them articulate with long haemal spines. The haemal arches cannot 

 be traced beyond the plane of the back part of the dorsal fin, but the 

 spines can easily be traced into the tail. In the region of the 

 posterior part of the anal fin they are rather feeble, but become 

 much more strongly developed where they are required to give 

 support to the lower lobe of the caudal fin. The anal fin is 

 supported by a single series of long interspinous bones, many of 

 which are curved to fit the ends of the haemal spines. 



The condition of the neural and haemal arches, spinous and inter- 

 spinous bones will be seen to be essentially similar to those described 

 by Traquair in Pygopterus and Nematoptychius. He says : " I can 

 find no clear evidence of the presence of two sets of interspinous 



