881 



single layers of cells. The lateral epiblast beyond this first becomes 

 thicker at a point which represents the Zwischenstrang of His. At 

 the reentering angle between spinal cord and epiblast there is on each 

 side a little group of cells still slightly attached to the inner portion 

 of the single epiblastic layer, and lying on what is now the extreme 

 dorsal summit of each limb of the spinal cord. 



From this stage onwards, although, be it said, a sharp eye can 

 distinguish even earlier the limits of ganglion Anlage and spinal cord, 

 there is always a well marked sharply defined boundary between the 

 two structures. Further forwards the lips of the spinal cord nearly 

 meet in the middle line; they do not close, as long as the ganglionic 

 Anlage rests on them they cannot do so. With the meeting of the 

 neural folds, the ganglionic Anlagen on each side fuse together and 

 separate from the skin. They now form what has been incorrectly 

 described as the first stage on the ganglionic formation. In the Chick 

 and some other types it is comparatively easy at this stage to see 

 that although the Anlage rests on the neural tube it really has no 

 connection with it. Soon after this the ganglia begin to leave their 

 position above the spinal cord and to shift down towards the lateral 

 surface. As they do so the cells rapidly proliferate. But we are here 

 dealing with well established facts. Indeed, my share in the description 

 of the ganglionic development really ceases with the wandering of the 

 ganglionic Anlagen from their position between the neural lips to the 

 lateral surface of the spinal cord. 



From the cells of the spinal ganglion the sensory nerves are out- 

 growths, and here as in the case of the cranial nerves also, not 

 merely fibrous outgrowths. A member of the cells of the ganglion take 

 part in the formation of the developing nerve. 



The cranial ganglia of Elasmobranchii. 

 The difference in appearance between the cells forming the ganglia 

 and those forming the neural plate in Elasmobranchs is very slight, 

 and as the ganglionic Anlagen are very early separated from the skin, 

 indeed much earlier than in the Chick, one gets a deceptive appear- 

 ance as though the ganglia were outgrowths of the neural plate. In 

 fact, the ganglionic Anlagen get invaginated along with the neural 

 plate, and unless the specimens examined are very well preserved, and 

 treated in a particular manner 0, it often looks as though after the 



1) Haemotoxylin and alum carmine are bad reagents for showing 

 the early development of the ganglia. — The best results are obtainable 

 with, either borax carmine or picroborax carmine. 



