RUST-RESISTANT VARIETIES OF WHEAT 47 



Certainly, however, there can be no question of a certain antagon- 

 ism between host and parasite when one observes phenomena such as 

 are illustrated in Plate VIII, 6. This is in remarkably sharp contrast 

 with the apparent congeniality exhibited in such cases as are shown 

 at 6, 7, and 8 in Plate VII. Antagonism would seem to be explicable 

 at present only by the toxin or enzyme theory. The recent work of 

 Bolley (1908 and 1909), Pole Evans (1909 and 1911), McAlpine 

 (1910), Freeman and Johnson (1911), and BifTen (1912) indicates 

 clearly that immunity and resistance are concepts which, from the 

 very nature of their variability and sometimes apparent capriciousness, 

 must be cautiously discussed. At least one substance, commonly 

 found in plants, has been found by Cook and Taubenhaus (1911, pp. 

 40 and 43) to be toxic to certain rusts of the genus Uromyces. 



Whatever the immediate instruments governing congeniality or 

 antagonism, the fundamental facts brought out quite clearly in the 

 results described in the present investigation have a bearing on the 

 practical and theoretical questions involved in the problem of pre- 

 venting cereal rusts and the breeding of rust-resisting varieties. Ex- 

 ternal morphology as pointed out by Ward (1902-1) for brome rusts, 

 Salmon (1905-2) for mildews, and Biffen (1907) for yellow rust is 

 also of very slight importance in the immunity of cereal varieties to 

 stem rust. No observed facts in intimate histology, moreover, give 

 any clue to resistance. In the absence of biochemical information 

 concerning the activities of invading hyphae and invaded host tissues, 

 actual performance alone can be depended upon as a safe criterion 

 in the development of resistant forms or immune varieties. This is 

 all the more true since Pole Evans (1911) has found that a hybrid 

 wheat produced by crossing rust-immune and rust-susceptible wheats 

 may rust quite badly and be capable of causing infection of the immune 

 parent and a more severe attack of rust on the susceptible parent 

 variety than rust from that variety itself will cause. The production 

 of flecks and dead areas on an inoculated plant is a character of 

 possible use in indicating at least a semi-immunity. 



There does not seem to be any obvious constant correlation be- 

 tween immunity and other observable characters, as for instance 

 drought-resistance. Although the immune varieties of wheat used 

 in this investigation are drought-resistant, it is also a well-estab- 

 lished fact that other drought-resistant wheats such as Kubanka 1516 

 are very susceptible. Moreover, the antagonism exhibited by Minne- 

 sota No. 163 wheat toward Puccinia graminis avenae and by rye 

 toward Puccinia graminis hordei does not differ fundamentally from 

 that exhibited by Khapli toward Puccinia graminis tritici'and in the 

 first two cases a correlation with drought-resistance is out of the 

 question. 



