400 JOHN BEARD, 



No doubt such a mode of disposiog of the problem has its ad- 

 vantages — it is a simple solution, but withal a delusive one. 



As long ago as 1889 it was noticed that in late stages of Scyl- 

 lium there are ganglion-cells in the spinal cord which remind one stri- 

 kingly of the giant ganglion-cells of Amphioxus. The description of these 

 in Scyllium may be deferred until its transient nervous apparatus 

 should have been considered. At the present juncture it may be 

 stated that similar cells have also been found in older Raja embryos. 

 Two such cells have been figured from embryos of Raja batis in 

 Figs. 33 and 57. The first of these is from the anterior trunk-region 

 of an embryo of 12,5 cm, the second from the tail of an embryo of 

 15,5 cm. In the former embryo the formation of the posterior fissure 

 of the cord was about completed, and in the section drawn there was 

 a large ganglion-cell with two processes extended across the fissure a 

 little above the central canal. The same section also contained two 

 degenerating ganglion-cells of the transient system lying at the summit 

 of the fissure. 



In the figure (Fig. 57) from the other embryo the ordinary 

 ganglion-cells of the cord — especially the motor cells — are seen 

 to be well developed. The posterior fissure, as evidenced by the 

 natural break in the upper part of the wall of the canal, is barely 

 complete. Stretching across the posterior fissure there is a large ganglion- 

 cell with five poles, of which one is obviously an axis-cylinder process. 



It would have been quite possible to have found other such cells 

 "reminiscent" of Amphioxus, and even, after an immense amount of 

 labour, to have determined their number in the embryo; but these 

 cells make their appearance at a late period of the development when 

 the embryo is of very large size. The lateness of their appearance 

 must be emphasized; they are entirely absent in embryos of 45 mm 

 and under, of which complete series were made; and only come upon 

 the scene when the transient nervous apparatus is in course of de- 

 generation. Therefore, the idea of a correspondence of this latter to 

 the few giant ganglion-cells of Amphioxus may be dismissed as futile. 

 If any cells in the cord of Elasmobranch embryos be the homologues 

 of the giant ganglion-cells of Amphioxus, it must be the few deeper 

 lying, "scattered" and isolated ganglion-cells, like the two described above. 



Some other than the Amphioxus one would therefore appear to 

 be requisite as the solution of the problem of the nature of the 

 transient nervous apparatus of Raja and other fornis. 



At the same time the giant ganglion- cells of Amphioxus and the 



