I04 Journal of Agricultural Research voi.xvi. No. 3 



of wheat and other species of Triticum, and also found a form which 

 developed normally on Kanred, P. 1066, and P. 1068. It was clear from 

 these results that the new forms could therefore do something which 

 neither of the two other forms could do. 



But none of the known forms could infect White Spring emmer (Minne- 

 sota 1 165), durum (Mindum, CI 5296), and several other varieties, mostly 

 durums. It was perfectly evident from the work with P. graminis 

 tritici-compacti ^ and the two other new forms that the biologic specializa- 

 tion within the genus Triticum could only be determined by testing 

 many species and varieties and that there was a strong probability 

 that forms of rust would be found which were capable of attacking 

 varieties resistant to all known forms of stemrust. This is exactly what 

 has been found. A form was found which infected White Spring emmer 

 and Mindum normally. The work has continued until about a dozen 

 forms have been found up to the present time (Oct. i, 1918). 



About 25 varieties and strains of Triticutn aestivum, T. durum, T. 

 compactwm, T. dicoccum, and T. rnonococcum are being used as diflfer- 

 ential hosts, and no variety so far tried is resistant to all of the rust 

 forms except Khapli (CI 4103), an emmer originally imported from 

 India. Some of the forms are very virulent on many varieties, while 

 others are weak and can attack only a few varieties successfully. Some 

 forms differ from each other only in their action on one or two varieties; 

 but these differences are definite and consistent. No attempt has yet 

 been made to name the recently discovered forms. 



The factors governing the distribution of the forms are not at all 

 clear. Material has been collected from 27 States and most sections 

 of the country are represented. Two distinct forms have often been 

 isolated from the same lot of material, and at least four have been found 

 in Minnesota. 



The fact that there are so many biologic forms of stemrust on wheat 

 seems to be of profound significance in at least two ways. It is an 

 additional reason for eradicating the rust-susceptible varieties of bar- 

 berry (Berberis spp.), and it is of the greatest importance in the work 

 of breeding wheats for rust resistance. 



Many of the virulent forms seem to occur in the Northern States, 

 where everyone will now concede that the barberry is of tremendous 

 importance in the persistence of stemrust from year to year. Eradicate 

 the common barberry and there is reason to believe that these forms 

 may gradually die out entirely, or at least be reduced to a condition 

 bordering on impotence. The fact also that in the South and on the 

 Pacific coast, where barberry does not rust commonly, the forms of rust 

 seem to be more uniform than in other regions certainly lends some 

 color to the view that the barberry may have some effect on their devel- 

 opment. A hypothesis that forms may have originated by hybridiza- 



1 Leach, J. B. op. cit. 



