History of a transient nervous apparatus in certain Ichthyopsida. 377 



realize what the investigator might still wish that it had been, when 

 he has ended his task, and surveys his small conquest over Nature. 



Greneral summary of the deyelopment of the apparatus, its 

 probable functions and fate. 



The development and degeneration of an apparatus of transient 

 ganglion-cells and nerves have now been followed through an extensive 

 series of stages, beginning with Raja embryos in which there were as 

 yet no ganglion- cells, and ending with newly hatched specimens in 

 which the cells had degenerated almost to nothingness, and had 

 practically ceased to exist. 



The mode , of description hitherto adopted may at times have 

 proved tedious to the reader. Its advantages are, however, not to 

 be despised. Its merits to the observer are manifold; among other 

 things it obliges him to deal systematically with his material, and 

 he thereby gains a knowledge of its deficiencies. Its utility to the 

 reader, on the other hand, is to be recognised in the insight it affords 

 him into the author's data, as well as in the more or less adequate 

 picture it yields of the actual state of affairs at any given stage. 



But disadvantages accrue, in that such a seriatim mode of de- 

 scription forbids any, or only the barest, comparisons, and withal 

 excludes any putting together of the general results attained. Left 

 in such a state, the story is only half told, and the reader finds 

 himself in possession of a series of facts, which he is under the neces- 

 sity of summarising for himself. 



Thus there need be no apology for the following pages. 



1. Comparative development of the apparatus in Raja batis, 



a) Development of the ganglion-cells. When one exa- 

 mines sections of early embryos (6—8 mm) of this form — embryos 

 with recently closed-in medullary folds, with gill-pouches in course 

 of formation, and in which barely half ^) (i. e. about 60) of the trunk- 

 somites have been segmented off from the mesoblast, it is noticed 

 that the wandering out of the spinal ganglionic foundations from the 

 lips of the cord in the pronephric region is not of so simple a cha- 

 racter as usually described, or as it appears to be in such a form as 



1) As many as 140 somites have been counted in later embryos 

 of Raja batis. 



