776 



SCIENCE 



[N. S. Vol. XLI. No. lOfiS 



these being relatively rapid. No guaran- 

 tee of future disease resistance has been 

 found in either the hybrids or in the wild 

 species. 



Not only do closely related species show 

 a difference in susceptibility to disease, 

 but varieties and races of the same species 

 behave differently in this respect. An ex- 

 ample of this is Triticum dicoccum, one 

 variety of which, as already stated, is re- 

 sistant to rust and the other non-resistant. 

 Additional examples are T. vulgare, a few 

 varieties of which are resistant; certain 

 varieties of potatoes with reference to 

 Phytophthora infestans; Finns sylvestris 

 with reference to Lophodermium pinastri; 

 and other cultivated plants. This differ- 

 ence in disease-resistance between races of 

 the same species is of far greater impor- 

 tance than the difference between two spe- 

 cies, because generally there is greater sim- 

 ilarity between the cultural value of the 

 two races. 



The occurrence of healthy plants among 

 diseased ones is not absolute proof of the 

 resistance of such plants, and therefore to 

 make sure of the immunity of any special 

 strain careful experiments are necessary. 

 It is not enough to raise a number of plants 

 of an apparently resistant strain in a cer- 

 tain place. The question of resistance 

 should be investigated from the beginning 

 on the broadest basis. One of the princi- 

 pal things necessary is to expose the re- 

 sistant plants to the fungus causing the 

 disease to which they appear to be resist- 

 ant. In the case of fungi which live in 

 the soil, such, for instance, as the fungus 

 causing stinking smut, the first requisite is 

 to determine whether they are present and, 

 if not present, to introduce them, while in 

 the case of fungi spread by the wind, such 

 as those causing rust and mildew, the in- 

 fection should be induced either naturally 

 or artificially. 



The presence of the fungus, however, is 

 only one factor in the experiment. The 

 second factor is the disposition of the host 

 plant, that is, its internal qualities, which 

 makes infection possible. The third fac- 

 tor is the coincidence of the infection 

 period with the susceptible condition of 

 the host. When all of these factors are 

 present the possibility of infection is cer- 

 tain, and only under such circumstances 

 will the results be reliable. 



Fluctuation in the prevalence of fun- 

 gous plant diseases is due to the presence 

 or absence of proper conditions for the de- 

 velopment of the fungi causing them. 

 For instance, loose smut appears to a very 

 serious extent in certain summers, and 

 naturally it would be expected to be still 

 more prevalent the succeeding summer. 

 The fact is, however, that although spores 

 in sufficient quantity to infect all the flow- 

 ers in the field were scattered, the disease 

 may be much less serious, the reason be- 

 ing that the plant was not in the proper 

 stage a sufficient length of time to receive 

 the infection, or in other words the 

 weather conditions caused too rapid wither- 

 ing of the flowers to permit infection. 



The effect of different conditions on the 

 relation of host and parasite makes it nec- 

 essary that investigations to determine the 

 resistance of strains shall be carried on 

 not only for a number of years, but also in 

 different localities. Even under such cir- 

 cumstances the outcome may be uncertain. 

 In many cases immune forms when culti- 

 vated prove to be only partly immune. 



The best opportunity for finding im- 

 mune strains is afforded by diseases which 

 are of regular occurrence. In such cases 

 it is possible to find with a degree of cer- 

 tainty forms which are immune in a cer- 

 tain locality, but while such experiments 

 may give results of practical value, the 



