CANAIJCUIAR APPARATUS OF ANIMAL CEI.I -. I S,~ 



affinit\- for the dye. It is certain, moreover, that the networks 

 apparently composed of solid fibers denvm-trated in the pancrca- 

 epit helium l>y the resorcin-fuchsin method, are not solid, for. 

 in preparation- made by this method, every cell in the section 

 \\ill -ho\\ Mich a deeply stained network while section- of the 

 same pancreas fixed in Kopsch's formol-bichromate solution 

 and -i. lined with iron-huematoxylin, will show in every cell a 



i) of c.m.il- \itti unstiined contents. 



AccordingK . u must either reject Holmgren's hypothesi- 

 or .i--ume ili.n then .ire two sorts of these nets, those of cartii 

 c.-ll- .nid epithelial cell- and leucocytes being different from th >-c 



of Del \e cell-. 



The -tatemeiu of Legeiidre foS-'og) that these structure- arc 

 either wholl\ .tb-ein or are the result of pathological chai 

 i- not to be -eriou-ly considered, in view of the fact that thi> 

 author in .itu-aiptin^ to explain the positive observations in 

 tlii- regard ot so m.my experienced investigators, is compelled 

 io re-ort to ihe \\holl\ unwarranted assumption that their re- 

 sults ha\e been due to the selection of unhealthy animal-, or 

 to the 1 1 \.nioi i ot ii-Mie- .ttier several days of inanition, or ^e\eral 

 hour- .liter de.tth. 



Main observers, includiiu Retxius ('oo), are inclined to 

 |.elie\e that the\ repiv-eiit an int racelluUir system of nutritive or 

 drain. t^e canals ha\ing direct relations with the lymphatic 

 system. I'lu- extracellular communications are, however, denied 



\>\ < ml^i and hi- jmpiU. who have never seen them in their 

 prep.n.iti..ii- nude b\ the chrome-silver impregnation method-. 

 nor by the -il\er reduction method. They are equally denied 

 by \on Rer^en. \\ho. however, admits the existence of canal- in 

 no \\IM ..nine, led \vith tlu-se, which he regards as artefact-. 

 \\hich do open on the -urface of the cell. 



\ "ii I'ti^'ii 114 who studied these structures by all three 

 method- but, in particular, by the osmic acid method of Kop-ch, 

 agrees \\ith < iol^i that the structures are for the most part 

 net u i 'i k- 1 1| liber- compx >-ed of a -ul stance which reduce- i-mic 

 .icid. but e\|>lain> the discontinuous elements found b\- him in 

 main cell- studied, b\- the assumption that the>" represent dif- 

 lerent stages in tlu- formation or destruction of the apparatu- 



