May 19,1923 
Statistical Study of Biologic Forms 
553 
forms of P . graminis and the size of the urediniospores and aeciospores of 
the same forms. As in the case of the last two spore types, so in the 
teliospores, P. graminis tritici and P. graminis agrostis constitute the 
extremes with respect to spore size. Figures 5 and 6 show very clearly 
that the teliospores of the wheat rust are larger in every way than those of 
any of the other forms. The secalis and avenae forms are practically the 
same in length but differ noticeably in width. On the other hand, the rusts 
of rye and of redtop vary insignificantly in spore width. In length the 
agrostis form approaches that of phleipratensis, but is appreciably differ¬ 
ent from it in width. The constants given in the third section of Tables 
III and IV and the differences in the means presented in Table VIII show 
when these similarities and differences are real and significant and when 
they are only apparent and inconclusive. 
M/C/SOA/S 
Fig. 6.—Differences in widths of teliospores of biologic forms of Puccinia graminis 
grown on congenial host plants and under tmiform cultural conditions. 
P. GRAMINIS TRITICI 
There is a difference of at least 5 microns between the mode for length 
of the teliospores of this biologic form and the mode for the longitudinal 
measurement of any other biologic form. A similar difference is found 
in the means for length. The greatest difference in the width of the telio- 
spore between this and any other form is about one micron for either mode 
or means. The means for this biologic form, as given in section 3 of 
Tables III and IV, are 51.80±o.49X 16.67±o.i2/x. 
P. GRAMINIS SECALIS 
The difference between this biologic form and the one described above 
can readily be seen by comparing their means, which for the secalis form 
are 47.35 ±0.45X 14.77 ±o.i2jLt. The difference in the means of the rye- 
rust and wheat-rust forms is 6.64 times its probable error for length and 
11.23 times greater than its probable error for width. Consequently, 
the difference must be considered as truly significant. 
