8. F. Armstrong 89 



fail to infect any other plants, but even the infection areas already noted 

 did not spread, and the tissues lying around such areas gradually died 

 off and with them the parasite also. Numerous brown dead patches 

 in the tissue remained, but clearly the hosts were the victors in the 

 struggle. 



The experiment shows that, while heavy doses of nitrate do lead to 

 an intensification of rust attack on susceptible varieties, such treatment 

 of normally immune wheats fails actually to break down or destroy their 

 resistant powers. Under such treatment, however, the struggle which 

 goes on between invader and host is seriously prolonged and may lead 

 to an appreciable loss of leaf-tissue in the latter. 



(r) New combinations of parental characters. 



It was clear, however, that, in addition to such external factors as 

 those mentioned above, other causes were concerned in affecting the 

 plants" susceptibiUty. In the same season (1919), and under the same 

 external conditions, some of the homozygous susceptible cultures and 

 plants were more severely attacked than others. Similarly, among the 

 homozygous "immune" cultures (and plants) great differences existed 

 in regard to the degree of re.sistance shown. Two of these cultures (66/9 

 and 11/21), containing 77 plants, actually came through this most un- 

 favourable season, and in spite also of the heavy apphcation of nitrate, 

 matured off absolutely free from Yellow Bust infection. Two other cultures, 

 including 91 plants, survived the test almost equally well, only 7 plants 

 showing any sign of attack. These cultures did not escape by maturing 

 off earUer than the rest, for they were not ripe until the second week in 

 August. In the other pure "immune" cultures, however, while some 

 plants were rust-free or had traces only, the majority had a slight attack, 

 and a number were moderately rusted. 



The above facts are only exphcable on the assumption that the 

 differences are partly due — at least indirectly — to other inherited features. 

 It was observed that new combinations of such characters as ear shape, 

 chaff colour, bearded or beardless condition, etc., appeared to be without 

 effect upon a plant's susceptibihty. But this does not involve the as- 

 sumption that all combinations of other characters are hkewise \vithout 

 effect. In a paper recently published Hayes (8) points to some evidence 

 of Unkage between rust resistance and certain morphological characters 

 in the offspring resulting from crosses between Triticum vulgare and 

 varieties of T. durum and T. dicoccunt. It is, indeed, extremely probable 



