262 



Journal of Agricultural Research 



Vol. XIX, No. 6 



because the kernels were not developed far enough to survive the inva- 

 sion of the parasite, and the infection was so great that none of the kernels 

 developed. The glumes and rachis at the base of the spikelet in each 

 case were covered with sori :o days after inoculation and the inner 

 surfaces of the glumes were filled with urediniospores. 



These results appear to confirm Johnson's {15) observations regarding 

 the effect of rust infection upon floret sterility in wheat. He found 

 floret sterility increased 20.03 per cent when wheat heads were sprayed 

 with a water suspension of a mixture of urediniospores of Puccinia grami- 

 nis and P. iriticina. His conclusions were that when the rust attacks the 

 ovary early enough it prevents its development, and other semiparasitic 

 fungi complete the process of destruction, while if it attacks the embryo 

 after it is fertilized and has begun to enlarge, a rusted kernel results. 

 Table I shows the outcome of a second set of inoculations. Kubanka 

 wheat (C. I. No. 1440) was used for these experiments. The glumes were 

 opened, and a very few spores were placed at the base of the inside of 

 the glumes with a fine platinum needle. The heads were then sprayed 

 with distilled water and inclosed in a test tube as before. Every spikelet 

 in each head, with the exception of the smallest ones at the tip, was thus 

 inoculated. 



Table I. — Results of artificial inoculation of wheat ovaries at different stages of develop- 

 ment 



It will be noted from Table I that in no case was kernel infection 

 obtained when inoculations were made while the ovary was very small. 

 On the other hand, when the inoculations were delayed until the kernels 

 had attained about two-thirds of their normal size at maturity, the ker- 

 nels were able to continue development, and a high percentage of rusted 

 ones resulted. It would seem, therefore, that the amount of kernel 

 infection each year does not depend alone upon the amount of rust 

 occurring upon the heads of the wheat but also upon the time when this 

 infection takes place and whether the kernels are at the right stage of 

 development to become infected. The weather conditions where the 

 kernels are at the right stage of development are also a very important 

 factor. 



