184 R- R- BENSLEY. 



various epithi-lia ,ind from other sources. The existence of tin- 

 canals has also been confirmed by a large number of observers 

 including Kolster Coo), Fragnito foo), Lugaro ('oo), Donaggio 

 ( ), Pugnat ('oi), Sjovall ('01), Smirnow ('01), von Bergen ('04) 

 and others. 



F"or comparison with the results of the Golgi and Kopsch 

 techniques an enumeration of the different types of cells in which 

 a canalicular apparatus has been found may be of interest. 

 Holmgren demonstrated the canals in the following cells: gland 

 cells of the pancreas and parotid, intestinal and gastric epithe- 

 lium, epithelium of the epididymus, biliary duct epithelium, 

 uterine epithelium, thyroid epithelium, liver cells, epithelium 

 of the suprarenal gland. Retzius ('oi) described similar canals 

 in the giant cells of the bone-marrow, which, like Holmgren, he 

 considered to be in direct connection with pericellular spaces. 



It is to be noted that many of the objects studied by Holm- 

 gren coincide with those studied by Golgi and his pupils, and 

 with those investigated by von Bergen, and that where this is the 

 case, the canalicular structures described by Holmgren correspond 

 closely in tln-ir location and in their configuration with those 

 demonstrated by the other methods. Whatever conclusion we 

 may reach with regard to the relation between the canalicular 

 apparatus of Holmgren and the reticular apparatus in nerve 

 cells, few who have studied the actual preparations made ac- 

 cording to these different techniques in respect to epithelial 

 cells and cartilage cells will deny their substantial identity. Ii 

 is true that there are differences in the appearances obtained, 

 but, in the opinion of many, these are sufficiently accounted for 

 by the differences in the thickness of the sections studied in the 

 different methods, and so, in the completeness of the apparatus 

 which is brought to expression in a single preparation. 



In his later papers dealing with these structures Holmgren 

 has abandoned his original opinion that the canals are lymphatic 

 in nature and constructed an entirely ne\v hypothesis as to their 

 nature. This hypothesis is based on the confirmation by him 

 of the interesting observations of Nansen ('86) and Rhode ('91, 

 '93. '95). that the nerve cells of certain Crustacea (NaiiM-n, '86), 

 and those of certain Gastropoda and Hirudinea i Rhode, loc. 



