SAPROPHYTISM AND SYMBIOSIS 775 



the nature of an antitoxin. Certain plants appear to be immune from parasitic 

 attack, possibly through the secretion of toxins, or of substances which induce apo- 

 chemotropic reactions in haustoria, or through the absence of substances which 

 induce prochemotropic reactions (e.g. glucose, in the case of Cuscuta). When the 

 pear is attacked by the mistletoe, the infected branches soon die, whereupon the 

 mistletoe dies; thus the pear generally is free from this parasite. Some trees are 

 immune to parasitic attack through the impenetrability of the cork layer, and 

 sometimes infected regions of host plants are isolated by cork formation. 



The origin of parasitism in seed plants. Whatever may be said of 

 the fungi, the parasitic seed plants obviously form a series of disconnected 

 groups, each of which is more or less clearly related to some autophytic 

 group, near whose level the parasitic group in question probably orig- 

 inated. The Euphrasieae appear to be in a state of active evolution, 

 and they exhibit all stages of gradation to autophytic Scrophulariaceae, 

 a family to which they clearly still belong. The parasitic dodders are 

 so close to the autophytic morning glories that usually they are regarded 

 as belonging to a common family. Even the Orobanchaceae, while 

 commonly accorded separate family rank, obviously are close to Gloxinia. 

 Rafflesia and Balanophora are more remote from any known autophytic 

 stock, and yet they are believed to be not very far distant from the Aris- 

 tolochiaceae. In no case is there a long genetic series of heterotrophic 

 forms, as in the fungi. 



While thallophytic parasites probably have passed through an intermediate 

 saprophytic stage, this may not have been the case in seed plants. An argument 

 for such a saprophytic stage is furnished by facultative forms like Lathraea and 

 Melampyrum, and also by the probable capacity of many ordinary autophytes for 

 partial saprophytism. Yet such parasitism as that of the mistletoe, where only 

 water and salts appear to be taken from the host, seems on the whole simpler than 

 the saprophytic absorption of organic foods. Practically nothing is known concern- 

 ing the exact causes underlying the development of parasitism in seed plants, and 

 even the various stages cannot as yet be regarded as certainly known. 



The intergradations among the Euphrasieae are very suggestive of 

 possible stages in parasitism, and the following may be hazarded as 

 a possible series in the development of such a root parasite as Orobanche. 

 The roots of plants of different species frequently come in contact in 

 the soil, and it may be supposed that the cells, either through mechanical 

 causes or through chemotropic reactions, may come into sufficiently 

 close contact to permit of osmotic interchange. Water or salts or both 

 would then pass from regions of high to those of low pressure, the plant 

 in which the pressure is low being the incipient parasite, while the other 



