28 THE RUSTS OF GRAINS IN THE UNITED STATES. 



P. graminis hordei (stem rust of barley) on barley, wheat, and rye. 



P. graminis secalis (stem rust of rye) on rye and barley. 



P. graminis avenae (stem rust of oats) on oats. 



P. rubigo-vera tritici (leaf rust of wheat) on wheat. 



P. simplex (leaf rust of barley) on barley. 



P. rubigo-vera secalis (leaf rust of rye) on rye. 



P. coronata (leaf rust of oats) on oats. 



(7) Two biologic forms may inhabit the same cereal or cereals (for instance, wheat 



and barley rusts on wheat and barley) without being identical. 



(8) By gradual variation and adaptation to varying conditions a certain rust spe- 



cies, widely distributed, may form a number of strains or types, differing in 

 physiological reactions. 



(9) The host plants exercise a strong influence, not only on the physiological and 



biological relationships, but in some cases even on the morphology of the 

 uredospores. 



In regard to the relationships of the cereal rust forms to the numer- 

 ous grass rusts of the United States there is much to be done. A 

 beginning has been made, and experiments have been performed 

 confirming Carleton's results (30, pp. 55, 61-63) in regard to the infec- 

 tion of Hordeum juhatum with the stem rusts of wheat and barley 

 and orchard grass with the stem rust of oats. That Agropyron repens 

 also acts as host for the stem rust of wheat has been proved. The 

 relationship of Puccinia phlei-pratemis to other rusts has been inves- 

 tigated and a summary of results published (59, p. 791). The 

 importance of this phase of the biologic forms of cereal rusts is very 

 great and demands early attention. The most extensive results 

 obtained up to the present time are those of Carleton with the 

 American and Eriksson with the European rusts. 1 



THE ^CLDIAL STAGE OF RUSTS. 



HISTORY OF BARBERRIES IX RELATIOX TO RUST. 



Up to 1864-65, when De Bary demonstrated the heteroecism of 

 Puccinia graminis Pers., rust life histories were very incompletely 



i During the course of preparation of this bulletin several important papers have appeared throwing 

 further light on biologic forms of rust. J. C Arthur ("Cultures of Uredineae in 1909/' Mycologia, vol. 2. 

 no. 5, 1910, pp. 227, 228) cites experiments of his own showing that Puccinia poculiformis (Jacq.) Wettst. 

 (P. graminis Pers.) has been grown on Tritkum ruJgare from gecidiospores derived from inoculations on 

 Berberis vulgaris with teleutospores from Agropyron repens, A. tenerum, A. pseudorepens, Agrostis alba, 

 Cinna arundinacea, Elymus canandensis, and Sitanion longifolium, respectively. He concludes that 

 " although in the uredinial stage this rust shows racial strains that inhibit the ready transfer from one species 

 of host to another * * * yet in the serial stage racial strains play no part, and the barberry acts as a 

 bridging host between each and every other gramineous host." 



Jaczewski, on the other hand, in a recent article (Zeitsehrift fur Pflanzenkrankheiten, vol. 20, no. 6, 

 1910, pp. 356, 357) cites comprehensive inoculation experiments to show that the stem rusts of grains and 

 grasses in Russia as a rule are not interchangeable even with the barberry as a bridging host, but retain 

 distinct physiological specialization in the secidial as well as in the uredo stage. He also shows that the 

 aecidia produced from inoculations with the teleutospores from the stem rusts of wheat and barley, respec- 

 tively, behave differently when used for inoculation on the same series of grains and grasses, and he believes 

 it easily possible that the stem rust on barley is a distinct physiological species, a conclusion independently 

 derived in another way by the writers of this bulletin (pp. 17-20 and 25-27). Although it is evident from 

 the experiments cited by Arthur that the barberry may act as a bridging host for rusts between some 

 gramineous hosts, in the light of the work of Jaczewski and others it seems that further experimentation 

 on a large number of rusts is necessary before the sweeping statement that " in the aecial stage racial strains 

 play no part " can be generally accepted. 

 216 



