History of a transient nervous apparatus in certain Ichthyopsida. 343 



system are taking on ganglionic characters aud a few short nerve- 

 processes of some of the cells are encountered. A little posterior to 

 the pronephric region there are a few transient nerves with a ventral 

 extension as low down as the segmental duct, the total number of 

 these is, however, very limited ; and only some half a dozen were noted. 



But the transient system as a whole is only in course of devel- 

 opment, and there is as yet a marked absence of fully developed 

 transient ganglion-cells on the cord. In the present embryo the 

 migration of the future spinal ganglion-cells from the lips of the cord 

 is not completed in that region in which the transient system first 

 develops. 



In embryo No. 551 (10,25 mm) five gill-pouches or clefts were 

 externally visible. Of these the spiracular pouch appears to be about 

 to rupture, and the first and second branchials are actually open. 

 Hi ere are about 97 somites behind the gill-region. The auditory in- 

 vagination is still widely open. The segmental duct terminates in 

 the epiblast some distance in front of the cloacal region. 



The most anterior transient nerve (on one side only) lies some 

 20 sections anterior to the commencement of the pronephros. Only 

 14 nerves were counted as yet developed in this embryo and of these 

 only four seemed capable of resolution into pairs. 



The ganglionic system and the nerves are only in process of 

 development, and the number of cases of mitosis in which future 

 members of the transient system are concerned affords striking evidence 

 of this. 



All the same it can be noted that there is a practical cessation 

 of transient ganglion-cells at about the point where the embryo is 

 leaving the yolk-sac. 



Two figures from the present embryo are intended to show more 

 especially the way in which the transient ganglion-cells strive towards 

 the epiblast. In Fig, 76 a the single ganglion- cell has spun its process 

 {n.p) into direct touch with the epiblast. In the other figure (Fig. 76) 

 several ganglion-cells have processes reaching to the skin. The latter 

 figure also illustrates the intertwining or anastomosis of processes of 

 centrally lying cells. It proves, as is in fact frequently enough mani- 

 fested, that the centrally lying cells are in contact by means of a 

 meshwork of nerve-processes. 



Embryo No. 563 measures about 10,5 mm. The spiracular pouch 

 is still closed, as is also the second branchial. Behind this there are 



