Guilliermond - Atkinson — 198 — 



Cytoplasm 



Relationships between the Golgi apparatus and the chondrio- 

 somes and plastids:- The above results have been a subject of much 

 debate in animal cytology, and various authors, among others 

 BowEN and Gatenby, Duboscq and Grasse have not been able to 

 confirm the observations of A. CoRTi and Parat. It must be noticed 

 that all these authors abandoned silver methods and used only 

 osmic methods. There was therefore the question as to whether 

 osmic methods produce the same results as the silver methods. 

 Work which we have done using these methods on plant cells has 

 shown us that the osmic methods are much less specific than the 

 silver methods. If the impregnations have lasted only a week, 

 there is a blackening of the vacuolar system only when it encloses 



Fig. 135. — Vicia Faba. Cells in the seed before 

 maturation. Silver impregnation of da Fano. 1-6, 

 parenchymatous cells of the cotyledon; 1-4, Golgi net- 

 work. 5-6, deprived of oxygen, the Golgi network 

 breaks up into vacuoles containing silver-impregnated 

 precipitates. 7, epidermal cells of the integument of * 



the seed; Golgi network. (After Sanchez). 



tannins, which instantaneously reduce osmic acid. Otherwise, it 

 is the chondriome which is impregnated and it may be well pre- 

 served but is often vesiculated. If the impregnation is prolonged 

 to two weeks, there is a more profound alteration of the plastids 

 and chondriosomes, which become large vesicles and sometimes 

 even anastomose into a fine network, much like the network of 

 Golgi, but now the vacuolar system may also be impregnated. These 

 impregnations are then very irregular and it is not rare to ob- 

 serve, side by side in the same section, cells in which the chondri- 

 ome alone is blackened, sometimes well-preserved, sometimes 

 strongly vesiculated, and other cells in which the chondriome and 

 the vacuolar system are both blackened, still other cells in which 

 only the vacuolar system is affected (Fig. 133, 11-18). These re- 

 sults demonstrate therefore that osmic methods constitute a 

 dangerous technique, a constant source of gravest error. 



Without having the pretension of entering here into a field 

 which is not our own, we will confine ourselves to saying that 



