Wheat Investigations. 35 



protein content. Since the coefficient of correlation referred to 

 above was obtained by grouping pure strains originally selected 

 from different varieties it became a matter of interest to deter- 

 mine to what extent each of the respective parent varieties in- 

 fluenced the value of that coefficient. A tabulation of the differ- 

 ent strains within each variety with respect to their relative rank 

 in protein content in 1917 and 1918 brought out the fact that the 

 varieties Canada Red, Preston, Red Fife and Unnamed furnished 

 most of the strains which retained their relative rank in protein 

 content from one year to the next, while with the Minnesota lines 

 and the Marquis and Bluestem varieties the behavior of the lines 

 with respect to this character was more erratic. It may be of 

 interest to note in this connection that the varieties Preston, Can- 

 ada Red and Red Fife which contained a large number of lines 

 retaining their relative rank in protein content from year to year 

 have been grown for years in Aroostook and Eastern Canada 

 whereas the other varieties whose lines did not show a consis- 

 tent behavior with respect to protein content have only very re- 

 cently been introduced into Aroostook. This fact may serve 

 to explain the different behavior of these two groups of strains 

 with respect to their protein rank from one year to the next. A 

 commercial variety is a mixed population composed of a number 

 of different strains. These strains may possess a varied degree 

 of response to the factors of environment of a given locality. 

 If a commercial variety is grown for a number of years in one 

 locality its component strains may be expected to have an es- 

 tablished degree of reaction to the environment of that locality. 

 Strains selected out of such a variety will tend to retain their 

 relative rank in respect to a given quantitative character. On the 

 other hand, varieties introduced into a new environment can 

 not be expected in the first years of adaptation to segregate out 

 strains with a fixed type of response to the new environment 

 in regard to their quantitative characters. This brief considera- 

 tion may acount for the different behavior of the local, Aroos- 

 took grown and the Minnesota strains. 



Baking Tests of the Pure Lines. 



The wheat of thirty-one pure lines of the 1918 crop was 

 ground in a small experimental mill in the laboratory of the Rus- 

 sell-Miller Milling Co., Minneapolis, and the flour subjected to 



