154 PROTOPLASM 



cylinder, such as Nansen has already discovered and figured 

 very well in Homarus and other Crustacea, for which reason 

 I need only refer to the figures given by him. 



In the literature there are two works, as far as I am aware, 

 in which results have been attained that approach very closely to 

 mine, namely, the investigations of Nansen (1887) and Joseph 

 (1888). I therefore think I ought to point out particularly that 

 I only subsequently came to know of these works, and thus 

 attained the same results in an unbiassed manner. Both 

 investigators observed the reticulation quite plainly in iine 

 transverse sections of the nerve fibres : Nansen in numerous 

 invertebrate animals, Amphioxus and Myxine ; Joseph, on the other 

 hand, in the medullated fibres of a number of vertebrata. The 

 latter author, moreover, got one step farther than Nansen, since 

 he also found, in longitudinal sections through axis-cylinders pre- 

 served in osmic, the fibrillse frequently connected by cross threads. 

 Nansen has never observed anything of the kind, but on the 

 contrary has set up a view concerning the structure of the axis- 

 cylinder which is incompatible with the existence of such cross 

 threads. In agreement with Leydig's views (1885) as to the true 

 nervous significance of the hyaloplasm, i.e. the clear intervening 

 matrix of the framework of ganglion cells and protoplasm gener- 

 ally, Nansen arrived at the idea that the axis-cylinder consists 

 of numerous fine tubules about 1 /x. in diameter, which are 

 formed of a "viscous," clear, structureless substance. Each 

 primitive tubule is supposed to be surrounded by a thin 

 sheath of spongioplasm. Now since the primitive tubules 

 would be closely compressed in the formation of an axis-cylinder 

 or of a corresponding nerve fibre, their spongioplasmic sheaths 

 are united into a framework, since the spongioplasm is not to be 

 regarded as "a quite firm and non-adherent substance." The 

 spongioplasmic framework that had arisen in this manner would 

 appear, therefore, in cross section as the reticulation described. 

 I can agree with Nansen to this extent, that I consider it proved, 

 from the comparison of longitudinal and transverse sections, that 

 we are not dealing with isolated fibrils in the axis-cylinder, but 

 with a honeycomb-like structure, the edges of which appear in 

 longitudinal section as the supposed fibrillse. On the other hand, 

 the longitudinal sections show perfectly clearly that the 

 cavities of this honeycomb certainly cannot be filled by con- 

 tinuous tubules, for they are divided by cross threads into 

 numerous chambers arranged in a row, one behind the other. 



Although it cannot of course be proved with absolute cer- 

 tainty by simple observation that these cross threads are parti- 



