10 BULLETIN 398, U. S. DEPARTMENT OF AGEICULTUEE. 



date-of-seediiig, and rate-of-seeding tests of winter and spring wheat, 

 and spring oats, barley, and flax. In different years the number of 

 varieties under trial has varied from 75 to 125. The number of plats 

 in the date-of-seeding and rate-of-seeding tests has ranged from 40 

 to 60, so that the total number of experiments has varied from 115 

 to 185. 



Tlie nursery work at Moccasin covers two general lines: (1) The 

 introduction and testmg of new varieties, and (2) experiments in the 

 unprovement of cereals. New varieties are usually tested in nursery 

 rows before bemg grown in plats. This makes possible the testing 

 of a large nimiber of varieties with much less time and labor than 

 would be necessary m larger plats. 



The effort in the improvement work has been to select individual 

 heads representing desirable types. These selections usually are 

 made from the varieties that have given the best results in the plat 



Fig. 3. — Buildings at the Judith. Basin substation, Moccasin, Mont., in 1915. (From a photograph 

 lent by the Office of Exhibits, U. S. Department of Agriculture.) 



tests. These selections are multiphed and tested for yield and other 

 desirable characters, such as length and stiffness of straw and freedom 

 from shattering. The poor selections are discarded and the prom- 

 ising ones increased. 



The number of rows has varied from 256 in 1909 to 840 m 1911. 

 Most of these are devoted to the improvement work. Since 1911 the 

 number of rows has been reduced as the poor selections are dis- 

 carded and the promising ones increased. 



EXPERIMENTAL METHODS. 



That portion of the substation farm used for plat work is divided 

 into blocks 132 feet wide and 617 feet long, which are separated by 

 roads 1 rod wide. These blocks may be divided into tenth-acre, 

 twentieth-acre, or fiftieth-acre plats, each 132 feet, or 8 rods, long. 



Plat Experiments, 

 size of plats. 



AU of the varieties in 1908 and 1909 and nearly all in 1910 were 

 grown on tenth-acre plats. In 1911 and 1912, because of the short- 



