SAPROPHYTISM AND SYMBIOSIS 767 



ing stage probably is that of obligate parasitism, involving the loss of a 

 capacity for saprophytism. Doubtless there are degrees of parasitism 

 even among the obligate parasites, since some species (as Pythium De 

 Baryanum) soon cause the death of the host and hence their own death, 

 The rusts, which rarely kill their host plants, appear on this account to 

 exhibit a higher stage of parasitic evolution. Another possible stage, 

 still further removed from autophytism, is illustrated in the univorous 

 habit; it has been noted that univores may arise from plurivores through 

 progressive variability, and that a return to a plurivorous condition is 

 equally possible. 



Probably the most complicated situation of all is afforded by heteroecism, and 

 no species has been so much discussed in this connection as the wheat rust. This 

 species may once have had all stages both on the barberry and on wheat; if the 

 aecidial stage developed poorly on the wheat and the uredo stage poorly on the 

 barberry, each ultimately may have become eliminated on those hosts, resulting 

 in the present heteroecious state. Another conception regarding the wheat rust is 

 that the barberry was the original host, and that by mutation or otherwise the uredo 

 and teleuto stages came to develop on various grasses; this view is favored by the 

 greater generalization of the aecidial stage. A possible mode of evolution of physi- 

 ological species is suggested by Puccinia Hieracii, a rust that infects a great many 

 species of Hieracium. It appears that each infected species of the latter is char- 

 acterized by a physiological species of the former. There is evidence that the 

 evolution of the species of Hieracium is very recent, and it is believed that the 

 physiological species of the rust have developed with the species of the host. 



Chemotropism is thought to have some connection with the origin 

 of parasitism. All fungi are prochemotropic with reference to certain 

 substances, and apochemotropic with reference to others. For example, 

 most saprophytes and facultative parasites grow toward saccharose. 

 The notably plurivorous parasite, Botrytis cinerea, though prochemo- 

 tropic with reference to saccharose, does not penetrate apples, because 

 it is apochemotropic with reference to malic acid. The fact that para- 

 sitic hyphae grow toward decoctions of their host plants suggests the 

 likelihood of chemotropic relations. Penicillium, a representative 

 saprophyte, grows in living tissues if the latter are injected with sub- 

 stances to which Penicilliiim reacts prochemotropically. Trichothe- 

 cium, another saprophyte, when grown for twelve or fifteen generations 

 on begonias that have been injected with sugar, is able thenceforth to 

 continue as a parasite without such injection and to bear fruit in the 

 usual manner. Similarly, certain parasites may be made to infect new 

 hosts, if the latter are thus injected. 



