f. J. F. SHAW. 1^ 



peculiarly liable to infection by the fungus. On the outer, 

 dorsal, surface of" the leaf sheath the epidermal cell wails are much 

 thinner in the areas between the vascular bundles than they are 

 immediately behind them, and, at the edges of the leaf, they are 

 extremely delicate. Moreover, immediately behind the vasculai' 

 bundles, not only the epidermal, but two or more hypodermal 

 layers have thickened walls, which give a cellulose reaction with 

 ehlorzinc iodide (see text figure). On the ventral surface of the 

 leaf sheath the epidermis is uniformly thin, and does not show any 

 such differentiation. Thus, on the outer surface of the leaf 

 sheath, there are areas which offer a poor resistance to any 

 parasitic attack. Once the fungus has gained an entry at one of 

 these spots, progress to the more delicate and deeper seated 

 tissues is easy. 



The hypha? are both inter-and intra-cellulai-. It is b}- no 

 means unusual to find that a hypha which has penetrated a cell 

 wall possesses a pronounced thickening in the portion on that 

 side of the cell wall fiom which penetration has taken place (PI. 

 II, Fig. 11). This swelling niaj^ be taken as evidence of the 

 increase in chemical activity, probably in the direction of 

 the secretion of enzymes, which precedes the solution and 

 penetration of a cell wall. 



A distinction must be noted here, between the behaviour of 

 the infected plants used in the inoculation experiments, and the 

 course of the disease in the field. A successful inoculation killed 

 tlie infected [)lant completel}^ (see PI. I, Fig. 3) ; but, in the 

 Held, the result of an attack seemed rather to be a gradual 

 weakening of the host, culminating in the failure to produce good 

 seed. Further experiments on a field scale are necessary to 

 elucidate this point. It ma}^ perhaps be the case that rice plants 

 grown under the sterile conditions of our experiments were less 

 capable of resistance to the disease than under normal cii- 

 cumstances 



Conclusion. 



The above account differs in several important respects from 

 that of Cattaneo. In particular the Italian author describes 



