26 FREDERICK TILNEY AND LUTHER F. WARREN 



large evagination, generally membranous, and in several forms 

 having marked prolongations. Thus in A mia there are two such 

 prolongations, the more dorsal of which extends as far back as 

 the midbrain, while in Polypterus a prolongation of the dorsal 

 sac arches over the midbrain and extends as far caudad as the 

 cerebellum. No anterior intercalated part is present in the 

 ganoid, but a well-marked habenular commissure is present 

 immediately cephalad of the epiphyseal complex. This latter 

 consists of a single evagination from the roof-plate. The anterior 

 epiphyseal element is absent in the ganoid so that the pineal 

 organ alone is encountered in this region. Immediately follow- 

 ing the latter structure is a short pars intercalaris posterior and 

 then a large posterior commissure. 



The pineal region in ganoids differs from that in selachians 

 mainly in the presence of a large and glandular paraphysis; also 

 in the existence of an unusually large and extensive dorsal sac, 

 prolongations of which are apt to extend far beyond the usual 

 limits of this structure, even arching over the midbrain and 

 reaching the cerebellum. 



Of the early works upon ganoids, Salensky 341 in 1881 first 

 gave a description of the development of the pineal region in 

 Acipenser. Accounts of the ontogenesis in this same form were 

 later given by Owsiannikow ('90) 297 and Kupffer ('93). 223 Bal- 

 four and Parker ('82) 12 gave a description of the development of 

 this region in Lepidosteus. Hill ('94) 18 contends that there are 

 two epiphyseal outgrowths from the roof of the interbrain in 

 Amia calva. The more anterior of these two outgrowths or 

 vesicles, Hill thinks, is homologous with the parietal eye of 

 Lacertilia, and he further maintains that it is extremely prob- 

 able that the two vesicles in their primitive position were side by 

 side, thus indicating the existence of two organs which in the 

 primitive form, like the lateral eyes, were arranged as a pair for 

 some definite function. Eycleshymer and Davis ('97) 113 con- 

 firmed the findings of Hill and added the further important 

 observations that in the late embryonic state nerve fibers could 

 be seen connecting the commissure habenularis with the para- 

 pineal as well as the pineal organ. 



