History of a transient nervous apparatus in certain Ichthyopsida. 387 



category, among these Figs. 16, 39, and 42, and, with a less degree 

 of certainty, Fig. 74. In these instances ganglion-cells are applied 

 end to end, and there are nerve-fibres passing from the free poles of 

 the cells. 



Apart from these chain-like arrangements of ganglion-cells and 

 nerves, there are very many striking pictures of ganglion- cells in other 

 positions with regard to nerve fibres. These are mainly instances in 

 which one or more ganglion-cells are applied to nerve-fibres at various 

 points along the course of the latter. All the examples of this in the 

 plates are briefly described in the "description of plates" and are also 

 mentioned in greater detail in some or other of the preceding pages, 

 and, therefore, any further notice may be confined to one or two of 

 the figures on plate 23. 



Figs. 29, 30, 32 and 37 are of interest in this connection. In 

 all these instances the course of the sub-epiblastic nerve is quite 

 obvious. In Figs. 29 and 32 there are ganglion-cells applied to such 

 a nerve in its course under the epiblast, while in Figs. 30 and 37 

 they lie in close contact to a nerve, surrounding it on all sides, near 

 its point of origin. 



There are indeed all sorts of bewildering variations of this to be 

 met with in different embryos. The cells may be applied to the nerve 

 at any part of its course, even, as in Fig. 58, as low down as the 

 level of the lateralis nerve. 



The curious nature of Fig. 37 has been commented upon in the 

 preceding pages. On the one side there is a thick coating of ganglion- 

 cells on the nerve, whilst on the opposite side of the body the corre- 

 sponding nerve has "nuclei" only applied to it. 



Figures like Figs. 31, 35, 43, 46 etc. may be more fitly referred to 

 in connection with another question, viz, that of the degradation, not 

 degeneration, of ganglion-cells. 



f) Physiological nature, i. e. functions of the trans- 

 ient system of ganglion-cells and nerves. It is not 

 intended to discuss the physiological rôle, which the apparatus may 

 be supposed to play in the embryo, at any length. 



In contradiction with the view formerly adopted by myself in the 

 preliminary paper it now seems evident that the transient system 

 must be regarded as largely, if not entirely, sensory in nature. A 

 couple of years ago my conclusion was that it was both motor and 

 sensory in function. One reason for this belief was, that it was difficult 

 to conceive of a nervous mechanism, except perhaps a degraded one, 



Zool. Jahrb. IX. Abth. f. Morph. Ofi 



