494 E- V. Cowdry, 



ratus in animal cells. He found (p. 192) that a mixture of formalin, 

 bichromate and sublimate was very well adapted for the demonstration 

 of this system in both plant and animal cells. Spinal ganglia were 

 fixed in this fluid and the distribution of the clear canals studied. 

 The three general types of configuration, already mentioned, could be 

 distinguished (figures 28, 29 and 30). The diameter of the canals 

 varies considerably as is illustrated in figures 28 and 42, which have 

 been drawn from neighbouring cells in the same section of the same 

 spinal ganglion. Occasionally the canals coil upon themselves forming 

 veritable spirals and nets, two or more of which may be seen appa- 

 rently disconnected in the same cell. It will at once be observed 

 that cells such as these (fig, 40) present a very striking resemblance 

 to certain figures published by Nelis, 1900, plate XXVII, fig. 15. 

 After fixation in formalin-bichromate-sublimate the Nissl substance 

 and the mitochondria can be specifically stained in the same cell with 

 the canals. 



Kingsbury's (1911) modification of the Weigert hematoxylin 

 method was applied to spinal ganglion cells and it gave a far clearer 

 and more precise stain of the canalicular apparatus than that obtained 

 with the iron hematoxylin. All of the three types of canalicular 

 formation seen in specimens prepared by the above methods were 

 represented (figures 31, 32 and 33). The uneven outlines of the net- 

 work, its rough appearance, and the occurrence of little ringlets and 

 crescents is sufficient evidence of its identity with the internal net of 

 Kopsch. This hematoxylin- stained system is morphologically independent 

 of the Nissl substance and of the mitochondria, which, as in the 

 cases cited above, may be specifically stained in the same cell 

 with it. 



The evidence in favour of the identity of the structures demon- 

 strated by these methods is as follows. In Kopsch preparations all 

 grades may be observed in a single section between cells possessing 

 only clear canals (fig. 37), clear canals and a few blackened meshes 

 of reticulum (fig. 38), and those containing only the black reticulum 

 (fig. 39). Furthermore, the blackened reticulum occurs in three general 

 types (figs. 25, 26 and 27) like the clear canals. It seems evident, 



