12 BULLETIN" 297, U. S. DEPARTMENT OF AGRICULTURE. 



harrow and disk harrow, pulverize the soil to a greater extent and 

 increase the danger from soil blowing. 



The only varietal tests here reported which were not conducted on 

 fallowed land are those with oats, barley, and flax in 1913. The 

 varietal test of oats was on corn ground, which was disked and har- 

 rowed into good condition before seeding. The barley varieties were 

 grown on land cropped to wheat the previous season, which was 

 plowed about 10 inches deep soon after the wheat was taken off and 

 was worked down with disk and harrow immediately after plowing. 

 The flax varieties were sown on land on which small grain was grown 

 in 1912, the preparation being the same as that for barley. 



The rate-of-seeding test with spring wheat in 1910 and with oats 

 in 1913 and the date-of -seeding test with flax in 1913 were on land 

 on which corn was grown the previous year. The land was disked 

 and harrowed but was not plowed before seeding. The rate-of-seeding 

 test with spring wheat in 1913 was on land cropped to small grain 

 the preceding season and prepared the same as that for the barley 

 varieties. 



RATES AND DATES OF SEEDING. 



The usual rate of seeding for spring wheat has been from 4 to 5 

 pecks to the acre; for winter wheat, 3 pecks; for oats, 6 pecks; for 

 barley, 5 pecks; and for flax, 2 pecks. 



Spring grains have been sown as early as seemed practicable, seldom 

 before April 1 or later than April 15. Winter grains have been sown 

 at what was thought to be the most favorable date each year. This 

 has usually been between September 15 and October 1. 



NURSERY EXPERIMENTS. 



The nursery tests at Newell have included varieties newly intro- 

 duced and those of which there was not sufficient seed for sowing in 

 the field plats, and also pure-line selections from the better commer- 

 cial varieties. The latter has been the most important feature of 

 the nursery work. The varieties and selections have been grown in 

 short rows, thus making possible the economical testing of a very 

 much larger number of varieties and races than could have been 

 included in the plat tests. 



NATURE OF THE WORK. 



Although the improvement of the cereal crops by selection lias 

 required a considerable outlay of time and money, the results are 

 less than were expected at the outset. This is due to several cause-, 

 the most important of which are the partial or complete crop failures 

 resulting from the extremely dry seasons. This has prevented 

 sufficient increase of the more desirable strains for a thorough test in 



