252 Journal of Agricultural Research voi. xv, N0.4 



yields of different varieties when grown in plots surrounded by alleys is 

 not considered. Shape seemed to be of little importance in plots not 

 surrounded by alleys. 



Barber (/) noted that plants in the borders of plots surrounded by 

 alleys were more thrifty as indicated by a greater number of culms per 

 plant, a longer period of growth and a higher yield of grain. Yields 

 obtained at the Maine Agricultural Experiment Station for individual 

 oat plants utilizing 36 square inches of space as compared with others 

 occupying 6 square inches showed that the former yielded four times as 

 much grain as the latter. Additional evidence of the increased yield of 

 plants with greater space for development is quoted from Wacker {10). 



Based on the investigations of Ten Eyck (9) and Rotmistrov {8), 

 Barber (/) makes the estimate that plants growing in an area 6 inches 

 wide within the borders of plots receive benefit from the adjacent alleys. 

 Tables are given showing that the nearer a plot is to the form of a square, 

 the lower is the percentage a 6-inch strip around the border of a plot to 

 its total area. From this it is concluded that(j) shape as well as size of 

 plot must be considered in variety tests, and (2) that square plots give 

 more accurate results in variety testing than rectangular plots of the 

 same size. 



Mercer and Hall (5) conclude that, based on data secured from plots 

 removed from large fields and therefore not surrounded by alleys, there 

 is practically no difference in the variability of yields obtained from 

 oblong and square plots of the same size. 



Jardine (2, 3) mentions that the practice of renioving the outer drill 

 rows of field plots to eliminate alley effect is followed by the Kansas 

 Experiment Station. 



Love and Craig (4) show that the same varieties of wheat and oats 

 average 36.51 per cent higher in yield when grown in plots ^^ acre in 

 size as compared with the yields in rod rows from which the end borders 

 have been removed before harvesting. The shape of the plots is not 

 given. Differences in stands under the two systems are mentioned. 



MATERIAL AND METHODS OF EXPERIMENTATION 



In the final variety tests of oats, wheat, and barley conducted on 

 University Farm in 191 7, each variety was replicated three times, thus 

 making four plots of each. The plots were 8.5 by 132 feet with 16-foot 

 roadways seeded to grass between each two series and 18-inch alleys 

 between each two plots. The 18-inch alleys represented areas not 

 included within the margins of the adjacent plots. Of each variety, then, 

 there were available for the determination of alley effect four plots made 

 up o! seventeen 6-inch drill rows and 1/38.82 (approximately ^) of an 

 acre in size. Eleven varieties of oats, five varieties of wheat, and four 

 varieties of barley were included in the tests. 



