336 ANNALS NEW YORK XCMn'.MY OF S('lf:\Ci:S 



in the cat, the latter arising from tlie neiiraxis liv delamination coiu- 

 hined with the evagination constitutes the ci-anial ganglia, and at the 

 extreme cej^halic pole, failing, we helieve, to se])arate remains incor- 

 porated as the primitive optic vesicle. 



Ci.osi i;i': (»i' I'll i: X kti; \i, Tiiu: 



At the extremities of the axis the (■lc\;ition ot the neural folds may 

 he resolved into two acts. Cephalad lln' median or hasal portion is first 

 elevated, while laterad the plate has still a horizontal direction. This 

 condition is present in the embr\'os of two and iliree somites (Plate 

 XXIII, Fig. 2). The condition seems to he the effect of modeling of 

 the ectoderm upon the ])araxial aceumnlation of mesoderm. At four 

 somites (Plate XX \'] I, Fig. 1) the fold is erect in its whole extent and 

 rises well above the mesoderm, the somatic ectoderm being closely applied 

 to the nenral plate in its dorsal half. Candad the process differs; the 

 lateral part of the plate is first elevated at some distance from the median 

 line (Plate XXVIT, Fig. 1, and Plate XXVII, Fig. 2) and forms a low 

 w^all for the broad rhomboid fossa. When the tube closes here, its diam- 

 eter is much less than the width of the fossa would lead one to ex])ect. 

 This is suggestive of the possibility that the lengthening of the tube is 

 not due alone to axial growth, but may be assisted by a rearrangement 

 of the material of the neural plate in the sense of a shift towards the 

 median line so that the plate is extended caudad as it narrows. The 

 elevation of its lateral margins is associated with the moderate entypy 

 of the blastoderm and the early completion of the amnion at its caudal 

 end. 



The neural folds lirst meet in the region of the future mesencephalon, 

 but their closure is not simply progressive from this point in both 

 directions. On the contrary, it is incident simultaneously at several 

 points which may he rather widely separated. In the eight-somite 

 embryo, in addition to the closure of the midbrain, which extends from 

 the optic anlage to the quintal ganglion, there is a second closure be- 

 tween the quintal and acoustico-facial anlages: and again, after an inter- 

 val, at a third ]inint the folds seem on the verge of meeting (Plate 

 XX\'II, Fig. 2). There is also sonu' fusion cephalad at the ventral 

 margin of the neuropore. This is of some theoretical importance and 

 diminishes the significance of the neuropore as a morphologic landmark. 

 The gaping of the tube in the I'egion of the ojitic and (|uiiual anlages 

 suggests that such structures in some way delay closure, but as the 

 neurosomatic junction is now sharply defined, it is difficult to believe 



