1889.] BOTANICAL GAZETTE. I4I 



are least subjected to irregularites of resistance. In every 

 case the pressure is due to enlargement of the spore during 

 growth, while the adjacent spores and the surrounding tis- 

 sues, simply through resistance, constitute the moulds which 

 shape the irregularities. 



These irregularities in the spores are of two distinct na- 

 tures, those which arise from an actual moulding of the spore 

 form due to turgidity of growth and inequalities of external 

 resistance and those which are due to an innate molecular 

 condition of the spore wall which permits of inequalities of 

 extension. The latter mode of formation is of an hereditary 

 nature, due to peculiarities of molecular structure effected by 

 the protoplasm of the species in which it occurs ; the first is 

 an accident of development. 



Herein lies what I take to be the chief structural differ- 

 ence between the species P. coronata and P. rubigo-vera. 

 While the digitate processes upon the spores of the first are 

 normal to the particular parts of the spore membrane, the 

 irregularities in the contour of the spores of the latter are 

 accidental, depending for the particular forms upon the 

 moulding of the young spores and a subsequent thickening 



of the cell walls. 



That the points upon the spores of P. coronata have no 

 constant regularity of form, number or position, does not 

 invalidate this idea. Certain portions of the spore membrane 

 are possessed of greater powers of extension, perhaps by im- 

 bibition, and they expand in the direction in which there is 

 least resistance. These points are always to be seen in ver- 

 tical sections of the sorus extending into depressions in the 

 epidermis and into interstices between the apices of the spores. 

 The position of the points with reference to the spores, as 

 seen from above, is shown in tig. 5. 



The Stroma.— In those species in which the teleutospores 

 are truly sub-epidermal during their whole development, they 

 are not "formed upon the same spore bed as the uredospoies. 

 which necessarily rupture the epidermis, but are aggregated 

 in a new spot under the uninjured epidermis, it is in the 

 completion of this new sorus that the fungus displays a high 

 degree of parasitism in that, when mature, the mat ot tungal 

 tissue which surrounds the spores becomes essentially a pait 

 of the host. The fungal hyph* ramify in the host tissues, 

 principally by way of the intercellular spaces (figs. 2, 3 and 4). 

 When a fruiting spot is formed two or more hyphal branches 

 coalesce, as seen in tig. 3, in an intercellular space lying be- 



