OF ARTS AND SCIENCES. 35 



color, depending upon their proximity to the spermogonia, and in 

 cases where many spermogonia have developed on a leaf the whole 

 organ takes on a yellowish color. 



From the observations described above, it seems impossible to 

 consider the cavity in which the spermogonium develops as origi- 

 nating from a single cell, being rather from the junction of several 

 cells. In other words, the spermogonium starts in the first place 

 as an outgrowth betioeen, and not in, the epidermal cells of the host, 

 soon however by pressure breaking through and absorbing the 

 confining walls, and making its way into the cavities of the sur- 

 rounding cells. As the absorption of these walls may take place 

 very early in the development of the spermogonium. before the mass 

 of hyphse has reached any considerable size, one might be easily 

 deceived into thinking that the spermogonium was forming in the 

 cavity of a single epidermal cell, which had been, perhaps, enlarged 

 somewhat by a morbid growth excited by the fungus. It is also 

 possible that one of the haustoria, already referred to, might have 

 been mistaken for the very earliest stage in the growth of a sper- 

 mogonium. The question might arise as to whether the young 

 stages herein described were not in reality tangential sections of 

 much older ones, but in the light of the evidence afforded both by 

 the sequence of stages observed and by carefully cut series of sec- 

 tions it is sufficient to say that such could not have been the case. 



As regards comparison with other spermogonia, a few words may 

 be of interest. The spermogonium of Cazoma nitens is not, it is 

 true, exactly comparable with the typical flask-shaped form usually 

 found in connection with other Uredinese, as it is much more super- 

 ficial, the growth starting immediately beneath the epidermis and 

 extending outwards towards the surface of the leaf, while in the 

 case of the spermogonium of jEcidium berberi<lis. far instance, the 

 spermogonium originates in the parenchyma of the leaf, and remains 

 very largely buried in it even to maturity. On the other hand; the 

 spermogonium of which the form found on Anemone isatype Is seen 

 to be even more superficial than that of Cteoni" nitens, the very 

 much flattened cone-shaped masses being located immediately below 

 the cuticula. The course of the development as well as the ultimate 

 result is simpler in Cceoma nitens than in JEcidium berberidis, aa 

 the hyphse merely remain in parallel rows, and have for their only 

 covering the cell walls of the host plant, while in the case of the 

 flask-shaped spermogonium there is a definite false peridium formed 

 by a more or less distinct matting together of the peripheral bypbje. 



