74 Transactions British Mycological Society. 
3 and 4) be separated from those which were less severely 
attacked, the following totals are obtained: 
Bad to severely attacked... - 307 plants 
Moderately or slightly attacked... 832 ,, 
Rust-free ... oe ae i 4a 
While these figures do not show a very close approximation to 
the 1: 2: 1 Mendelian ratio, the figures expected being 390, 
780, and 390 respectively they are suggestive of it, and, as 
pointed out before, if the autumn-sown plants alone are taken 
into account, the correspondence is very close. 
As regards the F; generation, considerations of time and space 
precluded cultures being raised from all the F, plants, so a 
number of the F, plants were selected so as to include indi- 
viduals showing every degree of attack as well as a number 
which had remained rust-free. Altogether 198 plants were 
selected to provide the F, cultures. On account of a severe 
drought during May and June (1919) these F, cultures were 
top-dressed with nitrate of soda as it was feared that some of 
the plots would be ruined. One result of this application was 
that, almost without exception, the cultures were more severely 
rusted than their F, parents had been in the previous year. 
Furthermore, of the 69 cultures which gave evidence of segre- 
gation into badly-rusted and slightly-rusted individuals, very 
few plants remained actually rust-free. In addition, of the 17 
cultures raised from F, plants which had been free from rust 
in 1918, in only two cultures did every plant remain absolutely 
rust-free in 1919. Notwithstanding these results, Armstrong 
points out that a comparison of these particular cultures with 
the homozygous susceptible ones growing alongside showed that 
the relative difference in the extent of rust attack was as great 
as had existed between their respective F, parents in the 
previous season. Armstrong concludes that, taking all the ex- 
perimental evidence into account, and more especially the 
environmental conditions, these cultures were “genetically 
immune,” but that the greater or less degree of predisposition 
to attack was due to the interaction of other causes or factors. 
Of 12 cultures raised from F, plants showing only traces of at- 
tack, three proved highly resistant and were therefore probably 
homozygous “‘immune” cultures; the other nine cultures gave 
evidence of segregation intoslightly, moderately, and badlyrusted 
types, and were clearly the offspring of heterozygous plants. 
Sixty-three cultures were raised from F, plants which had 
been moderately attacked. Fourteen of these proved homo- 
zygous susceptible cultures, for every plant became badly 
rusted. The remaining cultures gave evidence that they were 
the offspring of F, heterozygotes. 
