334 ANNALS NEW YORK ACADEMY OF SCIENCES 



with tlio more cninlal ones in owing its oi'igin to dclainination from the- 

 raedullarv fokl. 



In the quintal anlagc the sulcns is conspicnous and its horizontal por- 

 tion is associated w ilh a ridge-like projection (Plates XXV and XXIX) 

 in the angle hetwecn neural plate and somatic ectoderm, which yet 

 reveals its closer afhnity to the neural plate by blending with it at both 

 of its extremities in the four-somite embryo. It then appears subtended 

 by a longitudinal furrow, and differs more in size than in any essential 

 character from the more caudal ganglionic crest. 



Finally, the optic vesicle, a pure evagination, presents at an early 

 stage mucli resemblance to the quintal anlage (Plates XXIV and 

 XXVIII), from -which it differs chiefly in size and in the more ventral 

 position of its sulcus. 



We see in these anlages a series of structures, passing by gradations 

 from the delaminated ganglionic crest through the acoustico-facialis and 

 quintus to the optic vesicle, which is formed by evagination alone. It 

 seems to us, therefore, that the primitive neural plate in the cat gives 

 rise both to the neural tube and to the ganglionic crest, the latter being- 

 a derived and secondary element and not a coordinate intermediate be- 

 tween the medullary plate and the somatic ectoderm. In the forebrain 

 elements equivalent to the ganglionic crest are retained in the wall of 

 the vesicles and constitute the dorsal region of the neural plates, for if 

 the crest secondarily separates from the neural tube, in regions where 

 such separation fails to occur, it is more probable that the crest is in- 

 cluded in the l)rain than that it has been absolutely suppressed. .4c- 

 rordinf/Ji/. Hip analysis of the prosencephalon is not to he a'iempted in 

 terms of ilie hasal and alar plates alone, as Jias been rustornari/ since His, 

 hut must include a dorsal or ectal slrij) equivalent to the ganglionic crest 

 along its conrexl'i/, and tJiis miisl Imhule at least as wurli of the brain 

 wall as lies ectal to the optic sidcii^. Tins ganglionic element (the 

 primitive optic vesicle) at four somites forms the cephalic extremity of 

 the neural fold, and arches ventrad to the floor. If our argument is- 

 correct, the optic vesicle aud the ectoptic structures, whetlier above, in 

 front or below the optic region, must be considered of ganglionic e(puva- 

 lency, a conclusion which entails a revision of His's analysis of the brain. 



The question of the substitution of mesectodemi for a neurogenic 

 ganglionic crest in the prootic region in mammals can liardly, we believe^ 

 receive an affirmative ans-\ver in view of the conditions observed in the 

 cat, for we find quite generally the separation betAveen ectoderm and 

 neural tube clean cut, and the space between these structures unoccupied 

 by cells. The mesectodemi of the ichthyopsid and sauropsid emljryo,. 



