'mildew' of wheat 733 



In certain districts the parasite undoubtedly uses the barberry 

 as a second host, nevertheless the latter is not necessary for the 

 continuous propagation of the fungus since in Australia the 

 secidiospore stage is not met with. Even in this country the 

 absence of barberry bushes from certain localities does not 

 appear to diminish the prevalence of ' rust ' from such districts. 

 Moreover, where the barberry is common our experience leads 

 us to conclude that the secidium-stage of the parasite is much 

 rarer than it is generally assumed to be and in no sort of pro- 

 portion to ' rust ' attack. As previously stated, it is still an open 

 question whether the basidiosporeS are able to infect a wheat 

 plant and give rise to ' rust.' 



Eriksson obtained ' rusted ' plants in six to eight weeks from 

 cereal grains which were sown in sterilised soil and carefully 

 protected from any possible outside infection. He therefore 

 concludes that the fungus is transmitted from the 'rusted' 

 parent to the grain as a 'plasm,' which lives a latent life in 

 the cells of the embryo and young plant until just before the 

 eruption of ' rust ' spots, at which time it develops an ordinary 

 mycelium. 



This view, if fully established, would account for much of the 

 extensive early appearance of 'rust' in cereal crops which is 

 otherwise difficult to explain. 



The fungus Puccinia graminis Pers., not only occurs upon 

 wheat but upon other cereals and grasses as well, and is divided 

 by Eriksson and Henning into six different ' specialised forms ' 

 or 'biological varieties.' All have their aecidiospores on Berberis 

 vulgaris, or on species of Mahonia ; their uredospores and teleuto- 

 spores are met with on the cereals and grasses as follows : — 



Var. i. tritici on wheat only. 

 ,, ii. secalis on rye, barley, and couch grass (Agropyrum 



repens). 

 „ iii. avenm on oat, meadow foxtail, tall oat-grass and cocks 

 foot. 



