THE HYPOPHYSIS CEREBRI 457 



which constitute the pars infundibularis. In some forms, e.g. 

 Gadus morrhna (Herring (7 and 8), Anguilla (Stendell, 19), 

 Tilney (21)), the processes of the pars neuralis are situated 

 both anterior and posterior to the pars distalis, which forms a 

 solid mass of cells which, according to Herring (8), are chromo- 

 philic and form a wedge-shaped lobe between two chromo- 

 phobic portions. These lightly staining portions are penetrated 

 by the tissue of the pars neuralis and represent the pars infundi- 

 bularis. 



The primordium of the pars buccalis in Teleostomi is solid, 

 and in some cases at least it is bilobed. 



Dipnoi. — The descriptions of the hypophysis of Dipnoi are 

 somewhat scanty. Haller (6) criticises Burckhardt's (3) de- 

 scription on the ground that he confuses the saccus vasculosus 

 with the nervous lobe of the pituitary. 



Graham Kerr (10) gives a brief description of the develop- 

 ment of the hypophysis in Lepidosiren, from which it appears 

 that the hypophysial ingrowth is solid at first and later acquires 

 a lumen. He also states that the hypophysis " becomes here 

 as elsewhere closely united with the infundibulum, its dorsal 

 portion becoming penetrated by outgrowths of the latter." 1 



Amphibia. — The hypophysis of Amphibia has been de- 

 scribed by numerous workers, including Herring (8), Haller (6), 

 Tilney (21), and Stendell (19). 



In the Urodeles the pars neuralis is only slightly developed. 

 In Triton the ventral wall of the infundibular process remains 

 thin, whilst in Salamandra it is slightly thickened. In corre- 

 lation with this, the pars infundibularis is only slightly dif- 

 ferentiated from the pars distalis which forms a relatively 

 compact mass of cells. 



In Triton there are two cell strands which lie against the 

 tuber cinereum anterior to the pars distalis. From their posi- 

 tion and from comparison with the origin of similar structures 

 in the Frogs and in Reptiles, it seems probable that these cell 

 strands represent a pars tuberalis. 



In the Anura (fig. 3) a swollen, solid pars neuralis (P.N.) 

 is present, consisting, according to Herring (8) of neuroglia 



1 Since writing the above, I have had the opportunity of examining an embryo 

 (about 13 mm. long) of Ceratodus in the possession of Mr. D. M. S. Watson. 

 In this form the hypophysis is represented by a flattened pouch, which at this 

 stage at least is hollow. 



30 



