BUTLER AND KULKAKNI. 237 



(PI. II, Fig. 2) are sent freely into the cells of the pallisade and 

 spongy parenchyma and sometimes into those of the epidermis. 

 They are simple, or occasionally branched, finger-shaped processes, 

 not differing from those of Sclerospora or Pythium yalmivomm in 

 any essential. In the petiole the hyphge are still inter-cellular, 

 being found especially in the air-spaces between the cells, and the 

 haustoria are numerous and very distinct. Neither in the leaf nor 

 in the petiole are the bundles ever penetrated. In the conn, 

 on the other hand, not alone are hyphse found in the storage cells, 

 the mycelium being rather intra- than inter-cellular, but the bundles 

 are commonly penetrated (PL II, Fig. 5). Haustoria-like branches 

 extend from [the inter-cellular hyphse into the storage cells but 

 they are often not clearly distinguishable from the ordinary 

 branches of the mycelium, which themselves may extend from 

 cell to cell without appearing to experience the slightest difficulty 

 in penetrating the cell walls. This difference in habit according 

 to the tissue attacked is of considerable interest and appears to 

 suggest that it is not so much the structure of the wall, as the 

 nature of the substances occurring within the cells, that determines 

 whether penetration will take place or not. 



In assimilating cells the chloroplasts commence to lose colour 

 soon after infection takes place. Their sharpness of outline then 

 gradually disappears and the individual plastids run together into 

 a grumous mass in which, for a while, the starch granules are still 

 visible. Then the cells turn brown, the phloem elements sharing 

 in this change although not penetrated by the parasite. In the 

 corm, as soon as rotting is well established, and in the centre of the 

 leaf spots, starch is entirely absent from infected cells. 



No trace of reproductive organs has been detected in any part 

 of the internal mycelium. Sporangia are produced externally on 

 short stalks, arising from the epidermis at the margins of the spots 

 on both surfaces of the leaf blade and on the petiole. They come 

 out in successive crops as the centrifugal spread of the fungus con- 

 tinues. They have not been observed occurring naturally on the 

 surface of rotting conns but have been obtained by removing with 



