14 USTILAGO AVENAE AND USTILAGO LEVIS 



insufficient to explain the epiphytotic attacks often occurring. If this is 

 true, the 26.4 per cent obtained by Roesch offers no better explanation. 



This, it seems to the writer, throws some doubt on the possibility that the 

 mycelium and the gemmae in the glumes play an important role. 



THE CORNELL INVESTIGATIONS 



BLOSSOM INFECTION AND SEEDLING INVASION 



Before proceeding with a detailed discussion of the investigations car- 

 ried on at Cornell University, it may be well to indicate the principal 

 problem about which these researches center. This may be presented in 

 brief by contrasting the conclusions of Zade and his colleagues with that 

 of the writer on the point at issue. 



On microscopic examination of blossoms which were artificially inocu- 

 lated with spores of Vstilago avenue at pollination time, it was observed 

 that the majority of the spores had germinated. No sporidia were pro- 

 duced, probably due to unfavorable temperature as suggested by the in- 

 vestigations of Arland (1924: 80), Bartholomew and Jones (192:5: 575), 

 and Jones (1923: 590). By frequent examination for several weeks 

 after germination of the spores, it was found that mycelium is developed 

 which penetrates and becomes established in the glumes (Plate II, 3), 

 just as the German workers found. On the other hand, mycelium was 

 also found abundantly on, in, and under the epidermis of the caryopsis 

 (Plate I, 2). The presence of abundant mycelium in the pericarp is re- 

 ported also by Diehl, as previously pointed out. Zade was unable to find 

 such mycelium in any instance, and Arland noticed it only in a few 

 cases. Roesch does not make particular mention of it. All of these in- 

 vestigators assume that the mycelium and the gemmae in the glumes and 

 in the anther and stigma remains are the chief sources of inoculum for 

 invasion of the developing seedling. In the opinion of the writer, how- 

 i ver, it is the mycelium developed on, in, eind un , r the epidermis of the 

 caryopsis that is accountable for most of the seedling invasion. The data 

 on which his conclusions as to the character of blossom infection and 

 subsequent seedling invasion are based, are presented in the following 

 pages. 



Positional relation of pathogene structures to the oat flower and grain 

 Investigations by the writer were begun in June, 1924. It was rec- 

 ognized at once that accurate interpretation of the data would depend 

 to a very great extent upon the ability to work with material (oats for 

 seeding) the exact history of which was definitely known. Since no 

 such seed was available, several preliminary studies were made with a 

 supply of oats which had been harvested the previous year from a held 

 known to have had considerable smut in it. 



