342 ANNALS NEW YORK AfWDiniY OF SCIENCES 



recedes I'l'din the iiiargin, and the ieleiH-e[)liah)ii beeomes continuous witli 

 the infundibular region. There has thus been fomied troni the periphery 

 of tlie vesicle a series of derivatives which constitute an ectoptic arcade 

 and in each case tlie ])rocess has been the same, a remodeling of the 

 primitive optic vesicle, so that its central region idtimately constitutes 

 the definitive vesicle, wliile its ])eriphery becomes allotted to the anlages 

 of the other elements of the prosencephalon wliidi ai-c. tbei-efore, ectoptic 

 in their arrangement and cannot be reduced to a linear scries of neuro- 

 meres referred lo the longitudinal axis of the neural tube. We have 

 now summarized the ontogeny of the forebrain as it a))pears in our series 

 of embryos of the cat. It is hardly necessary to add that we are not 

 offering these conclusions as an explanation of tlie ])hylogeny of the 

 mammalian brain. 



Mesencephalon 



Closure of the neural tube and its separation from the ectoderm are 

 accomplished first in the midbrain. As elsewhere, the resulting roof is 

 concave. The concavity is present in the whole lengtli of the neuraxis 

 but its degree increases cephalad and is most conspicuous in the fore- 

 brain. In the mesencephalon it is marked and persists to the stage of 

 sixteen somites ; it is associated with a sagittal ridge which depends into 

 the., lumen entally. In section the ridge is often constricted at its base 

 of attachment and occasionally a fragment of it is found separate as a 

 small group of cells within the neural canal. In the embryo of eight 

 somites the mesencephalon (15) is closed in half of its extent and forms 

 the highest region of the neural tube (Plate XXYII, Fig. 3). At nine 

 somites the closure is complete, the midbrain has lengthened and its 

 height has markedly diminished (Plate XXXI, Fig. 2, 15). A com- 

 parison of these two embryos throws some light upon the nature of these 

 changes. The distance between the quintal anlage and Sessel's pocket 

 is the same in both models, Imt the interval between the optic vesicle 

 and the quintal ganglion has markedly increased, the axis of the optic 

 vesicle has altered and the forebrain has come to project strongly ventrad 

 at right angles to the rest of the neuraxis. These facts taken in con- 

 junction with the diminished height of the midbrain cannot be ade- 

 quately interpreted on the principle of unequal growth alone. It must 

 be taken as the expression of a remodeling of the whole regioji, in par- 

 ticular a lengthening of the dorsal portion of the midbrain without cor- 

 responding increase of its ventral parts, in consequence of which the 

 optic vesicle has not only been displaced but rotated tlirough 90° and 

 the neuraxis has been bent ventrad in the region of the forebrain. There 



