1919] FIELD CROPS. 731 



examined yearly for possible genetic impurities, and reselection practiced if 

 the type is segregating. Meclianical mixtures are rogiied out. A bulk plat of 

 a cross is grown for several years, the cross becoming homozygous automati- 



(2?! — -ix 111 

 „^ 5 ' in which n is the number of genera- 

 tions elapsing since the ci'oss, and m the number of allelomorphic pairs by 

 which the parent sorts difter. After several yeai's individual selection is 

 practiced. 



Rod-row plats of a standard sort are grown every third to fifth plat and 

 the yelds determined. Tlie probable error of the average of thr^e plats sys- 

 tematically distributed as determined by the checks is multiplied by three, 

 and when subtracted trom the higher yielding sort a somewhat arbitrary figure 

 is obtained below which it is considered safe to discard all sorts for the season 

 in question without eliminating a high yielding strain. A similar method is 

 used in determining higher yielding sorts after a three-year test. Promising 

 sorts are also tested in rod rows in cooperation with several of the substations. 

 After a three-year test in rod rows, sorts of promise are placed in tiie plat 

 varietal test conducted by the section of farm crops. Pedigreed sorts produced 

 at this or other stations are ultimately distributed to the farmers. The in- 

 vestigational staff of the division acts as a board to determine, on the basis 

 of performance, which sorts to distribute. 



New introductions of winter wheat obtained from the Office of Cereal In- 

 vestigations, U. S. Department of Agriculture, proved less hardy than selec- 

 tions and crosses which were made at this station many years ago. Some selec- 

 tions and crosses stand up much better than others, including both subcorneous 

 and substarchy types. Odessa, a late-maturing sort, was crossed with Turkey, 

 which matures somewhat earlier, homozygous forms having been produced 

 which range in maturity from the early to the late parents. Several of these 

 crosses proved as hardy as the Odessa parent, or slightly hardier, and gave 

 much higher yields, while some of them were nonhardy. Other crosses were 

 made, none of which gave as hardy v.heat as when Odessa was used as a 

 parent, indicating the value of using as a parent a type which excels for the 

 character desired. Pure lines which excelled in hardiness include a Turkey 

 selection, 1-03-120 ; Big Frame I-OG-6 ; and 1-03-229, a bearded, smooth, red- 

 chaffed form with subcorneoiis kernels selected from an introduction from 

 Padui, Russia. 



Cooperative rust investigations conducted with the hope of producing a rust 

 resistant bread wheat of milling quality limited other agronomic breeding 

 studies with spring wheat, but recent pathological investigations (E. S. R., 40, 

 p. 642) in which many biologic forms have been found are believed to have 

 rendered questionable the value of further rust-breeding studies. Row tests 

 of bread wheat sorts indicated that Marquis is a very desirable variety for 

 INIinnesota conditions. Several crosses made in 1908 between Turkey Winter 

 and Wellman Fife are deemed better than Haynes Bluestem (Minnesota 169), 

 although none appeared to be of more value than Marquis. Based on a com- 

 parison with D-1 and D-4, North Dakota selections; Acme, a South Dakota 

 pedigreed sort; and other Arnautka and Kubanka pedigreed varieties, Mindum. 

 a selection from the Arnautka (durum) group, is said to have given evidence of 

 being a valuable sort for Minnesota. 



Oats breeding consisted of a study of the value of selection, it being found 

 comparatively easy to produce a higher yielding sort by selection from a com- 

 mercial variety. Results with Sixty Day oats are regarded as especially strik- 

 ing, although only a single year's test in rod rov.'s is available. One pure line, 



