Sept. 1, 1920 Genetics of Rust Resistance 527 



METHODS OF STUDYING INHERITANCI? OF RUST RESISTANCE 



In the studies here recorde'd crosses were made between Marquis and 

 resistant durum and emmer wheats. Precautions were taken to protect 

 the emasculated heads from foreign pollen. The Fj plants were grown in 

 individually spaced plots, and seed from each Fj plant was grown sepa- 

 rately. The F2 families of crosses between the same parent varieties 

 gave similar results and were considered as a single cross. No effort 

 was made with either the Fj or Fj plants to protect them from natural 

 crossing, and an error was thus introduced which will be discussed later. 



The correlation between resistance in the Fj and the Fj generations 

 gave unusual results which indicated that some uncontrolled factor was 

 causing complications. For example, the Fj cross between emmer and 

 Marquis which was grown in 19 16 appeared resistant, while in the Fj 

 generation which was grown in 19 17 the number of resistant plants was 

 much smaller than would be expected if resistance were a dominant 

 character. This led to the belief that very likely more than one biologic 

 form of stemrust was present in 191 7. 



Each F2 plant which produced viable seed was tested in 1918. As 

 previously mentioned, the barberry bushes were removed from the 

 immediate vicinity of the rust plot early in the spring of 1918. An epi- 

 demic was obtained with a known strain of Puccinia graminis triiici 

 which had been cultured in the greenhouse by the Section of Plant 

 Pathology for several generations. This strain had been tested repeatedly 

 on varieties of wheat and proved to be constant. 



The 19 1 8 results have been used to determine the resistance or sus- 

 ceptibility of F2 plants which have been grown the previous season. 

 The Fg as well as the Fj data have been used as a basis for placing the Fj 

 plants in certain botanical groups, for the problem was chiefly to deter- 

 mine the mode of inheritance and correlation of resistance or suscepti- 

 bility with those botanical characters which are commonly used in 

 differentiating wheat species, some of which are also of economic im- 

 portance. 



STERILITY IN THE CROSSES OF MARQUIS WITH DURUM AND EMMER 



VARIETIES 



In order to determine whether there is any interrelation between 

 botanical characters and rust resistance, the data for each V^ plant were 

 taken in a correlated manner. The practical significance is obvious. 

 For example, if durum head and seed characters were ralher closely 

 linked with rust resistance in inheritance, it would be necessary to grow 

 a larger Fj population to obtain the desired form than if each character 

 were inherited independently. In deciding regarding possible linkage 

 it is important to know whether sterility is involved in the crosses. 



