268 Ethel McLennan: 



cells of the carpellary stalk do not contain starch, and many of 

 them stained deeply with the gentian violet after counter-staining- 

 the sections with bismarck brown. These blue cells were always- 

 plentiful in this region; such cells have been invaded by the fungus, 

 and their contents probably used for its own nutrition. They 

 always stain deeply, and the hyphae in their vicinity do likewise. 

 If the cells do not stain too darkly, it is possible to observe fung.al 

 threads forming a network in the lumen of the cells, many of the- 

 threads being exceedingly fine. The position of these cells is shown 

 in Plate XX. The cells are not enlarged, and apparently only 

 differ from those around them by their different staining properties. 

 The cell walls in this area are pitted ; whether this is the normal' 

 condition or whether the pit has resulted from a secretion of the- 

 fungus is a debatal)le point. 



The lateral walls of the carpel are packed with small compound 

 starch grains, and in tliis region the hyphae only occur between 

 the oells. They run in all dii-cctions, but are, as far as I have- 

 observed, strictly inter-cellular in this position, at this stage. 

 However, Avhen the stigmatic region is reached they seem to get the- 

 upper hand, and a large number of cells become their prey. These 

 cells are also starch-containing, and when so intruded upon they 

 immediately react to the violet stain. Sometimes the whole of this 

 area will aj^peav a dense violet colour, for the great majority of 

 the cells in this part are attacked at this period. When the cell is 

 first invaded, the starch is seen to become swollen and disorganised, 

 and loses its ])Ower of reacting to the iodini' wash used in prepar- 

 ing the sections (Plate XXI., Figs. 2 and 5). The fine hyplial 

 threads wrap round the starch groups, and even enter between 

 each individual grain (Plate XXI. Fig. 5), apparentlv digesting 

 them. There is no doubt that these cells are suffering at the hands 

 of tlie fungus, and that their contents are being transferred to tliis 

 fungal system. Some of the cells show an entire absence of starch; 

 they appear to be practically empty, and somewhat collapsed. 

 These have l>cen invaded at an earlier stage, and yielded their con- 

 tents in a similar way. The stigmatic tracts present in the carpel 

 wall generally show hyphae in abundance; they extend right into- 

 the stigmas, and even here become intra-cellular, but do so proliably 

 only arcer fertilisation has taken place, wlien the function of the- 

 stigmas has been completed. 



Occasionally, the base of a staminal filament remained attaclied' 

 to the uvarv during sectioning, and hyphae were found to extend; 



