36 BULLETIN OF THE 



angle which is seen on the outer margin of the disk in stage A (Fig. 

 38, x), and which indicates a tendency on the part of the disk to grow 

 over the adjacent ectoderm, is represented in stage B by a very acute 

 angle (Fig. 39, x), while the disk itself forms an actual fold, which 

 covers a portion of the undifferentiated ectoderm. 



The separation of the retina and its ganglion is accomplished by the 

 production of a basement membrane. Tiiis is gradually developed in 

 the substance of the disk between the regions of the retina and the optic 

 ganglion. In order to distinguish the newly formed membrane from 

 the original basement membi'ane which bounds the under surface of the 

 disk, I shall speak of the former as the intercejiting mtmhrane. The in- 

 tercepting membrane (Fig. 39, mb. i cpl.) takes its origin from the base- 

 ment membrane on a line a little within the lateral edge of the disk. 

 From this position it extends at this stage as a delicate lamella for a 

 short distance through the tissue of the disk. The direction of its course 

 is approximately parallel to the outer surface of the disk, and it divides 

 the distal portion of the disk into two masses, one of which is superficial, 

 the other deep. The superficial mass is the first portion of the retina 

 to be differentiated, and, as I have previously stated, the growth in this 

 region is chiefly lateral. The deep mass is the beginning of the optic 

 ganglion. The intercepting membrane does not extend so far as to 

 separate the superficial portion of the disk from the deeper part in the 

 proximal half. This condition is what one might expect, since in the 

 pi'oximal region of the disk the superficial nuclei divide in such planes 

 as to thicken the disk, and a membrane which would separate the deep 

 and superficial parts would be a soui'ce of interference in the process of 

 thickening. 



The intercepting membrane is not an involution of the original base- 

 ment membrane, but is produced by the activity of the ectodermic cells 

 between which it lies. There is no i-eason for doubting, I believe, that 

 at this early stage (B) it is strictly an ectodermic secretion. At this 

 stage it is fan-shaped, the handle of the fan being formed by that part of 

 the membrane which is in contact with the original basement membrane. 

 The plane of the fan is parallel with the outer surface of the disk. In 

 sections which are either anterior or posterior, but parallel to that shown 

 in Figure 39, portions of the intercepting membrane are often visible, 

 and may be apparently luiconnected with the basement membrane. 

 This is due to the fact that the plane of section cuts the anterior or 

 posterior edge of the fan without including the handle. 



The chief difference between the intercepting membranes at stages 



