the Caslacanth Fish. 331 



frontals and with the undetermined bone described above. 

 Just behind its articulation it is excavated by two deep 

 channels, which plunge down into the bone. 



The postorbital and other cheek-plates are sufficiently ex- 

 plained by fig. 5. 



There i?:, however, in one of my specimens and in B. M. N. H. 

 no. 49834 a small bone covered with ''denticles'' like a scale 

 lying below the quadrato-jugal and in front of the lower end 

 of tlie operculum. 



Loioer Jaw. — The lower jaw hus been already well described 

 by iSmitli Woodward. The long straight prearticular extends 

 forward from the hinder end to the symphysis, where it meets 

 its fellow as a massive rounded bone on the dorsal surface of 

 the jaw. 



The articubir and the postglenoid ossicles have already 

 been described by Smith Woodward, although in my speci- 

 men they are not visibly se|)arated by sutures from the 

 angular. The dentary, splenial = in£radentary, and coronoid 

 are also well known. 



I find, however, that the dentaries do not quite meet at 

 the symphysis ; they are separated by the prearticulars for a 

 space of about 8 mm. in a large skull. 



The upper surface of the prearticular and dentary for a 

 distance of about 1 cm. from the symphysis is covered by a 

 series of three small bones built up of fused tooth-bases. The 

 anterior of these, at au}^ rate, supports a single larger tusk on 

 its posterior surface. 



The opercular apparatus and shoulder-girdle are so well 

 known as to require no further description. 



Sh'ipe oj Head. — Specimens of Macropoma are in general 

 only very slightly distorted. The height of the skull is fixed 

 very definitely by the pterygoidal apparatus, which extends 

 from the quadrate to the skull-roof. The width of the head 

 can be determined with considerable accuracy, because the 

 gulars obviously, as in Polt/pterus, fill up the wiiole space 

 between the rami of the lower jaw. In transverse section 

 each gular is bent round very nearly for a quadrant of a 

 circle ; when placed together in natural position the ventral 

 surface is horizontal, and in section there is a smooth tran- 

 sition into the vertical lateral surface of the angular — in 

 fact, the section between points at about half the height of the 

 angulars is rather accurately semicircular. 



This fixes the position of the lower jaws, and that of the 

 pterygoids and sides of the face follows directly. 



The anterior end of Macropoma so reconstructed is very 



