ONION SMUT FUNGUS. Ill 



selves as conidia, but does not state that he ever found them detached. 

 This resembles the process in U. cepvlne in (1) the production of the whorl 

 of branches, (2) the complete absence of sporidia and (3) the progressive 

 emptying of the cells. The main point of difference is in the elongated, 

 ultimatel}^ septate, promycelium in U. occulta which replaces the globose 

 vesicle of U. cepidae. ^ 



In U. Tritici Koern. the process is almost identical with that of U. 

 occulta according to McAlpine (11), but the promycelium is at times uni- 

 cellular, a condition which suggests that of U. cepidae. 



The germination process in U. Anemones (Pers.) Wint. has been studied 

 by Fischer von Waldheim (7), Plowright (14) and Brefeld (4). As de- 

 scribed and figured by Brefeld it is almost identical wdth the process which 

 the wTiter observed in U. cepidae except that the promycelium is not so 

 large. The whorl of 2 to 4 branches arises very close to the surface of the 

 spore on a very much reduced promj'celium, and they remain permanently 

 sterile. 



In U. Filipendidge Tul. (Brefeld (4) ) the whole process is identical wath 

 that of U. Anemones. 



Germination of the spores of U. Violae Sow. has been studied by Prillieux, 

 Dangeard, Brefeld (4) and others, being a favorite subject for study because 

 of the ease with which germination can be brought about in water. Each 

 fertile cell of the spore ball produces an elongated promycelium which 

 becomes septate just as in U. occidta. A whorl of three to eight diverging 

 branches is produced at the apex. Each verticillate branch grows out at 

 the distal end into a slender sterigma on which is borne a long cylindrical 

 conidium. In nutrient solution these primary conidia may produce sec- 

 ondar}'' or tertiary conidia. The process in this species differs from that 

 of U. cepidae (1) in the length of the promj^celium, and more especially (2) 

 in the development of conidia. 



In general, then, we may conclude that U. cepidae differs in its germina- 

 tion from the other species of Urocystis (except U. Violae) only in theshape 

 of the promycelium w4aich is here reduced to a nonseptate hemispherical 

 vesicle. All other details of development appear to be identical. 



Comparison with the Process as described by Thaxter. 

 As described and figured by Thaxter the spores germinate by a single 

 long irregularly branched tube on the tips and lateral branches of which 

 are borne small ellipsoidal to long ovoidal conidia. He does not mention 

 a globose prom,ycelium or whorl of branches such as the ^ATiter has always 

 observed. The marked differences in the process as observed by the wTiter 

 and as described by Thaxter are difficult to explain, unless they are due to 

 contamination in the cultures used by the latter. He states that he was 

 unable to obtain the material pure, and that "all the cultures swarmed 

 with bacteria." The presence of these same bacteria might produce a 

 difference in the development of the germination process. The writer in 

 attempting to secure germination by Thaxter's method also failed to keep 

 the spores free from bacteria and therefore changed to a different method. 



