May 19,1923 
Statistical Study of Biologic Forms 
565 
spores. Plate II, F, G, shows the response of the aecial infection of the 
redtop rust to two species of Berberis, one susceptible, B, vulgaris (F), 
and one resistant, B. hrevipaniculata (G). The effect of these hosts on 
the size of the aeciospores is graphically illustrated in figures 9 and 10. 
Urediniospores and teliospores were also affected in a similar manner. 
This effect, however, is not permanent; it lasts only as long as the 
unfavorable circumstances prevail. This is especially true in the case 
of urediniospores. As soon as the rust is transferred to a congenial 
host plant and favorable cultural conditions are reestablished it again 
develops normally and the spores attain their normal dimensions. 
The ill effect of excessive heat, insufficient light, and deficient soil 
moisture was quite definitely proved for urediniospores. This would 
lead to the assumption that aeciospores and teliospores are probably 
likewise affected by unfavorable cultural conditions. 
Host plants belonging to different species and varieties are not in 
themselves sufficient to cause a change in the normal development of the 
rust fungus or in the size and shape of the spores unless they vary mark- 
Fig. 14.—Parallelism between the means for lengths and widths of the different spores of biologic forms 
of Puccinia graminis. 
edly in the degree of susceptibility to a biologic form. Plate II, D, E, 
shows an equal amount of aecial infection of P. graminis secalis on B. 
vulgaris and B. sieholdii. The uniformity in spore size of both cases is 
illustrated by graphs in figures 7 and 8. 
Besides differences in spore size, deviations in the shape of the spores 
of the various biologic forms were also observed. These were described 
by Stakman and Levine (14, p. 48) as especially pronounced in the ure¬ 
diniospores. Somewhat less pronounced differences in shape were found 
by the writer in the aeciospores and teliospores, particularly the latter. 
Unfortunately the necessary statistical data were not available for cal¬ 
culating the ratios of length to width in the different types of spores, 
and hence constants could not be computed, nor could differences in 
mean of diameter be established. This is all the more regrettable because 
of the full realization of the importance of expressing quantitatively the 
differences occurring not only in the size but also in the shape of spores. 
A comprehensive method for this purpose has been described by Rosen¬ 
baum (jo,/?.-250). 
