8 Robinson, Puccinia malvaceanun and its host plant. 



the other hand it must be noted that the cells of the pith 

 directly interior to the affected patch contain distinctly 

 less starch than in the unaffected parts, although the 

 mycelium had not actually penetrated to the pith. In the 

 petiole the starch is similarly distributed in relation to 

 the individual vascular bundles, and a corresponding dis- 

 appearance of starch from the starch sheath and cells 

 surrounding the inner side of the bundle was observed in 

 affected areas. 



This disappearance of starch from cells of the affected 

 area may be due to the direct solution of the starch by 

 the haustoria of the fungus. Such direct attack, however, 

 seems improbable since the starch also disappears from 

 the cells of the pith which are not actually reached by the 

 mycelium. It is conceivable therefore that the fungus is 

 indirectly responsible for the disappearance of the starch 

 by tapping the carbohydrate stream directly in the phloem 

 and thus preventing the accumulation of new reserves. 



The effect of the fungus on the collenchyma is of 

 some interest. Here the cellulose thickenings on the cell 

 walls gradually disappear on the aggregation of the 

 mycelium between the cells. Text-Jig. 6a shows the normal 

 collenchyma before any mycelium had entered the inter- 

 cellular spaces, and the cellulose thickenings are repre- 

 sented by shading. In the case shown in Text-Jig: 6b the 

 mycelium, though not represented in this or the following 

 figure, occupied the expanded intercellular spaces of a 

 corresponding area of collenchyma. The cells showed 

 considerable enlargement, but at this early stage the areas 

 of thickening on the cell walls had undergone ver}- little 

 solution. At a later stage the cells were more widely 

 separated by the accumulated hyph;e and the thickenings 

 on the walls had been largely dissolved {Text-Jig. 6c). 

 In connection with this it may be mentioned that in both 



