272 Journal of Agricultural Research voi. xix. no. 6 



Nine different varieties of wheat seed were sown, and the plants were 

 examined for rust every four or five days. Leafrust appeared on all 

 the plots on June 5, and stemrust appea-red from July 17 to July 29. 

 Johnson sums up the results as follows: Rusted durum, Fife, and 

 Bluestem kernels produced plants showing no earlier or more severe 

 development of rust than adjacent plants from clean, uninfected seed. 



Experiment 2. — On April 12, 191 6, rusted kernels of wheat were 

 sown in separate flats in the greenhouse. About 25 kernels were used 

 from each of the following varieties: Allora (C. I. No. 1698), Kubanka 

 (C. I. No. 1440), and Marquis (C. I. No. 3641). These flats were trans- 

 ferred to the pathological garden May 1 1 , and were at that time in the 

 fifth or sixth leaf. They were headed about June 22, and stemrust did 

 not develop upon them until July 13, when a few leaves of the Marquis 

 wheat, which still remained green, bore sori of Puccinia graminis. It will 

 be noted by reference to Table IV that this was about the date upon which 

 stemrust developed upon wheat in the field plots and was indeed about 

 the date when stemrust appeared upon all the wheat in the vicinity. The 

 season was very backward, and rust did not make its appearance 

 nearly so early as usual. 



HISTOLOGY OK SEEDS AND SEEDLINGS 



Histology of seed. — The general appearance of the exterior of wheat 

 kernels infected with stemrust has been previously described. In order 

 to examine the interior of these kernels two methods were found to be 

 fairly satisfactory: One, in which the grains of wheat were boiled in 

 water and then sectioned on the freezing microtome ; the other, a modifi- 

 cation of the glycerin method described by Howard (jj). This latter 

 method was found to be satisfactory, and good sections of mature wheat 

 kernels were obtained. After sectioning, Pianeze stain was used with 

 good results. 



When sections of infected kernels were examined with a microscope 

 it was found that not all the sori appeared upon the surface. In some 

 instances the entire hilar region of the kernel was found to be filled with 

 sori, of which from i to 12 were found in a single kernel. These sori 

 often were found facing inward against the aleurone layer which was 

 very much distorted by the pressure (PI. 42). Other sori were found, 

 nearly spherical in form, entirely embedded in the pericarp tissue. There 

 seemed to be no regular arrangement, although the sori were often ar- 

 ranged in a circle around the hilum. This is what would be expected, 

 for many of them undoubtedly were coimected with Infection on the 

 rachilla before the kernel was broken away from the point of attachment. 

 Plate 43 is a longitudinal section through the hilum of an infected kernel 

 and shows the hilum nearly cut off by a large sorus, which probably was 

 formed from several sori that had become confluent. Plate 44 is a cross 



