378 JOHN BEARD, 



Torpedo. It is seen that, while the process just mentioned is going 

 on, there are a few cells among the most dorsally situated in the 

 cord, as also some of those among the wandering cells of the future 

 spinal ganglia, which are taking on ganglionic characters (Figs. 1, 2, 

 3 and 5). They are growing large, the nucleus is increasing much 

 in size, and one or more highly refractile nucleoli are present. Owing 

 to this phenomenon the roof of the cord, as seen in transverse section, 

 looks as though it had undergone an upheaval. The formation of 

 ganglion-cells as just briefly related begins somewhat in front of 

 the pronephros; from this point it gradually extends backwards until 

 it covers the region of some 25—26 somites. In other words it begins 

 at about the 6th trunk segment, and is completed somewhere about 

 the 31 St. At no period are any of these cells formed much in front 

 of this region ^). 



As a general rule, by the time all the gill-clefts are formed, and 

 the embryo has attained a length of 9— 10 mm, the cord in this region 

 has a tesselated roof of gangliou-cells (Figs. 73, 97). The position of 

 these closely-packed cells is at the extreme dorsal limit of the cord, 

 as reference to many of the figures on plate 22 will prove. The 

 number encountered in any single transverse section varies somewhat, 

 usually there are two or three at least, often five, six or even eight. 

 As the cells are much more numerous in Raja than in Scyllium, 

 Lepidosteus or Salmo., each section must of course contain more of 

 them than are met with in sections of the latter. Good horizontal 

 sections of embryos in which most of these cells have become ganglionic 

 reveal a beautiful mosaic arrangement of them, such as depicted 

 (here for only very small portions of the cord) in Figs. 73, 90, 97 

 and 99. The number of these cells in the region of one somite 

 averages about twenty, and the total of those on the cord cannot be 

 less than 500 — 600. That is, when no account is taken of those that 

 have wandered outwards, and also of such as occur in regions posterior 

 to the 31st trunk-somite. 



These cells are often termed giant ganglion-cells l)ut the term is 

 a misnomer, for their size is not specially great, being little if anything 

 greater than that ot the spinal ganglion-cells ^) of newly hatched 

 embryos '*). Most, if not all of these central ganglion-cells are multi- 



1) This statement holds good for embryos of all the forms as yet 

 examined by me (see note p. 379). 



2) As also previously noted by Rohon in Salmo. 



3) Tor this reason it seems also inappropriate to use the term 

 macro-ganglion-cells except in a descriptive sense. , 



