[42 BOTANICAL GAZETTE. [ June, 



tween the epidermis and the hypodermal layer. This is the 

 beginning of the stroma* (hymenium, sporenlager, spore-bed, 

 etc.), a mass of fused fungal hyphae from which the spores 

 arise. 



In section, the mature stroma gives the appearance of a 

 regular tissue displaying many large cell-like openings often 

 much larger than the diameter of the ordinary hypha, due 

 to the interstices between uniting filaments and the solution 

 of some of the uniting walls (figs. 2 b and 11 c). 



The fungal filaments not only have the power of uniting 

 themselves with the cell walls of the host by fusion, but prob- 

 ably by the secretion of an unorganized ferment. ' can pene- 

 trate, pass through or wholly dissolve them. Hence it is that 

 the hypodermal host tissues are always to be found closely 

 united (fused) with the stroma and are not distorted, though 

 the pustule, since its formation, has expanded greatly. This 

 quality of fusion is most noticeable in the so-called "para 

 physes" which accompany the teleutospores of P. rubigo-vera/ 



The various descriptions of this species invariably refer to 

 ''dark brown paraphyses" intermingled with or surround- 

 ing the spores. Burrill also records the same with regard to 

 P. coronata. The bodies are easily found in both species by 

 scraping off the pustules and examining the debris. But 

 their varying tissue-like dimensions, dissimilarity to ordinary 

 paraphyses (compare fig. 6c with fig. 7), and the further fact 

 that they are always not only connected at the base with 

 the stroma but also with the epidermis at the top, led me to 

 believe that they could not be paraphyses in the best accepted 

 sense, but were merely vertical extensions of the basal stroma. 

 Furthermore, in all cases in which a microchemical reagent 

 affected these bodies, the action was the same as that for the 

 stroma. 



Carefully prepared serial, vertical and tangential sections 

 of the host leaves, passing through the pustules, confirmed 

 the truth of this supposition. 



Soon after the coalescence of the filaments which start 

 the new sorus, the young spores appear, developing centrif- 

 ugally, while the hyphae which form the spore -bed eat away 

 the host tissues, spread rapidly, and finally enclose the spores 

 on the sides (fig. 2 b and/), thus, when fused t< o 

 with the epidermis at the top, cutting off what may°be termed 



.-1 



4 For the use of the word see Sorauer, Pflanzenkrankheiten ed. 2, vol ii, p. 212. 

 Also Plowright, M British Uredinea; and Ustilagineje," p. ! 



■ Vines, Physiology of Plan p. 191. , 



-accardo, Bylloge Fungorum, vol. vii., part 2, p. 625. 



