O5 
were large, as was also each conidium. Toward the dry end the conidia- 
clusters and the conidia both became smaller. From graphs of conidial 
length (Fig. F) it is obvious that the mean length of the conidium is much 
me ow ao ® 
Fic. 2.—Showing at left, glass slip in test-tube: O0=water level, 
1, 2, 3, ete., position of wheat shoots. In other test-tube, combustion 
boat supported on a bit of glass tubing, colonies growing on the agar 
in the boat. 
