12 



MISSOURI AGR. EXP. STA. RESEARCH BULLETIN 49 



The rows in some cases ran east and west, and in some cases north 

 and south. 



All these plots were seeded with a 5-row nursery drill, built from 

 plans furnished by Professor T. A. Kiesselbach of the Nebraska Sta- 

 tion. This is a hoe drill designed for rapid and thorough cleaning 

 between plots. Photographs of it have been published in reports of 

 earlier work on field plot technic at the Nebraska Station (Mont- 

 gomery 14 page 57, and Kiesselbach page 16). Its use resulted in uni- 

 form seeding and covering and accurate spacing between rows, with a 

 close approach to ordinary field conditions in the state in which the 

 field was left after seeding. Each field was seeded in a single day. 



All plots were harvested by hand with sickles, a foot at each 

 end of each row discarded, and the remainder (usually 16 feet) tied 

 in a bundle and hung in a ventilated shed to dry. In 1919 and 1920 

 each row was bundled and threshed separately; in 1921 the border 

 rows of each 5-row block were bundled separately and the three in- 

 terior rows bundled together. Yields were determined by weighing 

 in grams at the time of threshing. All final yields were converted 

 to bushels per acre and are so expressed. 



Work of 1919. In 1919 tests were conducted with barley and 

 oats. Thirty varieties of barley were grown, each in 3 replicate plots. 

 The test comprised three ranges of 185 rows each, including 21 check 

 plots, or one in every sixth plot. The barley was drilled at the rate 

 of eight pecks per acre, on March 21, in rows running north and south. 

 The rows were 14 feet long and 10 inches apart. They were cut to 12 

 feet in harvesting. The planting plan is shown in figure 2. Conditions 



FIGURE 2. PLANTING PLAN OF BARLEY VARIETY TEST 1919. Legend: B, 

 border. CK, check. Numbers 1-30, planting numbers of varieties tested, as 

 given in Table 1. 



were fairly favorable, and the yields of the adapted varieties were 

 slightly higher than the average obtained under the conditions at Co- 

 lumbia. Two varieties, Italian and Australian White, gave extremely 

 low yields and were excluded. Another, Sandrel, was represented 

 only in two series, and was also excluded. The yields of the remain- 



