330 THE BASIDIOMYCETES 



and the uredinial and telial ones on wheat, that the stages on 

 barberry could infect wheat but could not infect barberry and 

 vice versa, and, as a consequence, that both hosts were required 

 for the polymorphic rust to complete its cycle of development. 

 In spite of these experimentally established results, the heter- 

 oecism of rusts was not immediately accepted among mycologists, 

 at least not without temeritv^ 0rsted [Arthur et al. (1929)] 

 during the summer of 1865 showed the connection of Gymno- 

 sporangiinn sabijiae with Roestelia cmicellata. By 1880 it had 

 been established that 16 rusts are heteroecious, and during the 

 next 10-year period the list was extended to include about 100 

 species. Klebahn (1904) lists 153 species, distributed among the 

 genera as follows: Chrysomyxa 2, Coleosporium 14, Cronartium 

 3, Gymnosporangium 13, Melampsora 21, Melampsorella 1, 

 Melampsoridum 1, Puccinia 85, Pucciniastrum 3, and Uromyces 

 10. The list compiled by Dietel (1918) contains 264 species, and 

 since 1918 many additional ones have been demonstrated to be 

 heteroecious. 



The origin of the heteroecious habit among rusts has long been 

 a fertile field for speculation, and, as on all such topics, accord 

 is lacking. Acquaintance with this problem, with the views of 

 those who have contributed to its solution, and with the difficul- 

 ties involved can be gained from the accounts by Olive (1911), 

 Grove (1913), Orton (1927), and Jackson (1931). In general 

 two diametrically opposed views are held: (1) that the ancestral 

 forms were autoecious but able to live plurivorously; that is, they 

 were at least able to pass their entire developmental cycle on 

 either of two hosts; or (2) that ancestral rusts were heteroecious, 

 as well as heterothallic and polymorphic. Fischer is among those 

 who assume that heteroecious species once lived on either host 

 autoeciously, and that host specialization is a recent development. 

 Such a species as Fiiccinia graminella, with both aecia and telia 

 on Stipa, might be regarded as evidence of this point of view. 

 This species is not plurivorous, however, and there is no evidence 

 that it ever was. Other evidence for this point of view is that 

 the host for the monocaryotic phase is the more primitive and 

 that the stimulus imparted by the dicaryotic condition results in 

 invigoration. It would therefore follow that aeciospores pos- 

 sess greater vigor, and hence greater capacity for infecting, than 

 do other spore forms. The fact that aeciospores of some species 



