3o6 A TEXTBOOK OF THEORETICAL BOTANY 



specialized Monocotyledon. It must not be assumed by this that all Rusts 

 alternate between Dicotyledons and Monocotyledons, though this is fre- 

 quently the case. There are many which occur only on Dicotyledons, while 

 others occur on Ferns. It must be noted, however, that the two host plants 

 normally grow near each other. 



There are many examples of Rusts in which the life-cycle has been 

 reduced by the dropping out of a variable number of the spore types. These 

 are known as short-cycle rusts. The Hollyhock Rust, for example, only 

 produces teleutospores and basidiospores, while the Antirrhinum Rust gives 

 rise to uredospores as well. In all cases of short-cycle Rusts teleutospores 

 are developed, from which the essential basidia can be formed. The cyto- 

 logical behaviour of these forms has not been fully studied. 



It frequently happens that a heteroecious Rust may give rise to a short- 

 cycle Rust when one of the two essential hosts is lacking. In South Africa 

 and Australia, in the absence of the Barberry, a short-cycle form of Piiccinia 

 graminis is found in which only uredospores, teleutospores and basidiospores 

 are produced, thus cutting out the part of the cycle involving the Barberry. 



On the other hand, in America, and in particular in Minnesota, Dakota 

 and Manitoba where the heteroecious form of P. graminis occurs, the Rust 

 may cause a 50 per cent, reduction in the wheat crop, and rigorous methods 

 are taken to eradicate the Barberry which is about as common there as the 

 Bramble in this country. 



A further complexity in the life-history of the Rusts is introduced by the 

 existence of biologic races. Such races are morphologically indistinguish- 

 able from one another, but it is found that the spores of one race are usually 

 incapable of infecting the host plant of another. Three important biologic 

 races of P. graminis have been recognized, though a number of other less 

 clearly defined races have been described. 



Piiccinia graminis tritici usually attacks Wheat only, though it may cause 

 a feeble infection in Barley and Rye. P. graminis avenae attacks Oats, Dactylis 

 glomerata and some other grasses, but not Wheat, while P. graminis secalis 

 is found on Barley, Rye, Agropyron repens and some other grasses, but not 

 on Wheat or Oats. It should be realized, however, that the series of biologic 

 races observed in one country need not necessarily behave in the same 

 way in another, and a biologic race which does not attack Wheat in 

 America may be found to be a serious disease of Wheat in Australia or 

 in this country. 



Resistance to Wheat Rust has been studied extensively. It has been 

 shown that susceptibility or resistance is dependent on one or more INIendelian 

 factors (see Volume IV). By careful breeding and selection it has been 

 found possible to produce a series of Wheats immune or largely resistant to 

 Black Rust, though the process is a long one, for it does not of necessity 

 follow that an immune variety is at the same time a good grower or a high- 

 yielding strain. Work along these lines, however, would seem likely in the 

 long run to produce more positive results than the wholesale destruction of 

 Barberry bushes. 



