984 
Journal of Agricultural Research 
Vol. XXIV, No. 12 
ACCURACY OF THE RESULTS 
In estimating the accuracy of the types of infection that appeared on 
the hybrids the behavior of the rust on pure varieties was taken as a 
criterion of the average amount of variation that occurred. In Table I 
the results of inoculations made with Form I on February 10 and 20 
are given for a number of varieties, including Mindum and Pentad. 
In several cases, as those of Little Club and of Khapli, the results from 
the two periods are practically identical. In the majority of cases, how¬ 
ever, the rust development from the inoculations of February 10 was 
weaker than from those inoculations made February 20. For example, 
einkom had type 3 infections from the first inoculation but type 4 
from the later inoculation. Similarly, with Kota, the February 10 
inoculations produced type 3 infections whereas type 4 resulted from 
the February 20 inoculations. 
TablS I .—A comparison of the types of infection produced hy biologic Form i on several 
varieties of wheat at two different periods 
The appearance of a type 4 infection on i out of ii Kanred plants 
inoculated February 20 and similarly on i out of 31 Mindum plants, 
indicated accidental infection from some other form of rust, or impurity 
in the seed due to mechanical mixture or to natural crossing. The 
