Tlic rchitioiis of iiiif iiclioiidi i.i ••lini oilier r-yto|il;i.smic constituents etc. 491 



<^T(';it \';ii'i('1y of lliiids: iilcdlml, InniKilin. Allnuiiiirs usiiiic-liiclii-oiiiate, 

 Uciida's lliiid. cliidiiic-siiMiiiiJilc. Inriii.iliii-Zciikcr, < idhi's ai'S(fiiic»ns acid 

 roimaJiii and alcoliul mixlinv. I Icniiaiiirs lliiid. 2"j„ osiiiic aciil. Canioy's 

 ():3:1 fluid. Znikci's tiiiid. and many otlicrs. Tlin iiiftliod of stainin<r 

 makes little or no difference; for tlie canals generally apjx-ar as clear 

 unstained spaces or tubules against tlie coloured backgi-ound of the 

 cytoplasm: indeed. Von Berg-en (li)04. p. 017) states tliai itm- di tin- 

 cil aracteri sties of tlie canals of his type I is that they remain prac- 

 tically unstained. Three general types may be observed in adult spinal 

 ganglion cells of the pigeon. The first occurs most al)undantly and 

 is found in the large cells. It is distributed fairly equally throughuut 

 the cytoplasm (figs. 19, 22 and 28). The second t^^ie obtains in 

 the large and in the medium sized cells also, it is circumsci-ihcd and 

 is seen in an excentric position at one or other side of tlic nucleus 

 (figs. 20, 23 and 29). The third type, on the other hand, occurs only 

 in the smallest cells. It is circumnuclear in distribution and seldom 

 extends to the periphery of the cell (figs. 21, 24 and 30). The canals 

 are evidently more continuous than in the other types and form a 

 veritable network about the nucleus. 



Unless the canals are impregnated in some way, they are rather 

 difficult of interpretation, for they may be either exaggerated or reduced 

 in size by the fixing agent, and sometimes they are not even \isible. 

 There are two chief sources of error. Clefts are often produced in 

 cells by faulty sectioning. When they are only slight they occur 

 simply on the lower surface of the section, as it is cut. and in such 

 cases by careful focussing may be distinguished from the canals wliicli 

 are seen at all levels in the section. The parallel arrangement of 

 these clefts is also distinctive. In some cases they are actually con- 

 tinuous with the true canals. Cracks and fissures may also be caused 

 by rapid, repeated, dehydration and clearing. These do not show a 

 parallel arrangement. The canals cannot be confused with either the 

 mitochondria, the Xissl substance, or the neurofibrils: for they may 

 be demonstrated in the same cell in company with these structiu-es 

 by the acid fuchsin-toluidin blue method of Bensley (fig. 1\ They 

 may, however, be seen to best advantage in the cells with neuro- 



