12 DISEASES OF CULTIVATED PLANTS 



in their parasitism to a single kind of plant, or even to a 

 particular variety of a plant, through those fungi that are 

 restricted to a group of closely allied plants ; ending with 

 those whose parasitism is yet in a rudimentary or incipient 

 stage, and consequently exhibit but little discrimination in 

 the selection of a host, but attack many different kinds of 

 plants when conditions are favourable. The last-named 

 category of parasites can also live as saprophytes when 

 favourable conditions for parasitism are not forthcoming. 



Now chemotropism is the force that enables a parasitic 

 fungus to gain an entrance into its host-plant. In the most 

 highly differentiated parasites the germ-tubes of the germin- 

 ating spores have become so sharply specialised that they 

 respond to, or are only attracted by, some special chemotactic 

 substance or substances met with only in the cells of the one 

 particular kind of plant they have adapted themselves to. 

 Of course the spores of the most highly specialised parasites 

 germinate on the surface of any or every kind of plant on 

 which they happen to alight, but if the chemotactic substance 

 attractive to the fungus is not present, the germ-tube of the 

 fungus is not attracted into the tissues of the plant, and 

 infection does not follow. To this category belong several 

 kinds of ' rust ' fungi attacking cereals and other grasses. 



A second batch of parasites, including those causing the 

 hollyhock and the potato-diseases respectively, are somewhat 

 less specialised than the members of the preceding group, 

 and respond to the chemotactic substance characteristic of 

 a certain family of plants. 



Even in the very worst cases of an epidemic it is not 

 unusual to find that certain plants of the kind attacked 

 remain perfectly free from disease. On investigating such 

 individuals I discovered that their freedom from disease 

 was due to the absence of the special chemotactic substance 

 that enabled the fungus to infect the plant ; in other words, 

 such plants were immune to that particular disease. 



This discovery should prove to be of value in any attempt 

 to produce strains of plants immune to their most destructive 

 parasite. Until quite recently attempts in this direction 

 have been based on the production of a more resistent 

 cuticle, formation of bloom, or some other structural 

 character that would prevent the entrance of the germ- 

 tubes of the fungus into the tissues of the host-plant. 

 Marshall Ward has paid special attention to this subject, and 



