266 Journal of Agricultural Research voi. xvi, no. io 



Plots 2 and 3 are 5 rods long and 4 rods wide, while the others are of 

 the same length but of only half the width. Bach plot is separated from 

 its neighbors by a strip 6 feet wide. The plots have been seeded to the 

 center of this 6-foot strip, the outer edge around each plot being cut away 

 at harvest time. 



The soil has been classified by the Bureau of Soils of the United 

 States Department of Agriculture as Hempstead silt loam (jo, p. 26). 

 The silt loam stratum extends to a depth of from 39 to 50 inches, below 

 \\hich is a thick bed of clean gravel and coarse sand. The surface 

 stratum is very uniform in texture as may be seen from the moisture 

 equivalents reported in Table III. 



CULTURAL CONDITIONS 



The original plan of the experiment had been carried out until it was 

 interrupted in the spring of 191 5, when in order to determine the relative 

 productivity of the different plots we had all five plowed, prepared alike, 

 and planted to the same crops — viz, corn {Zea mays), sorghum (Andro- 

 pogon sorghum), turnip (Brassica rapa), alfalfa (Medicago saliva), and 

 mangels {Beta vulgaris macrorhiza) — these being so arranged that each 

 of the five appeared on every plot (fig. i, B). Weeds were very bad on 

 all the plots except No. 4 but, by frequent use of hoe and horse culti- 

 vator, these were kept down in all the crops except the alfalfa. On 

 account of the unusually cool weather, the crops made very slow growth 

 until early in July, and none of the plots at any time appeared to 

 suffer from a lack of moisture. 



DISTRIBUTION OF ORGANIC MATTER IN .SURFACE FOOT 



The ratio of organic carbon to nitrogen in surface soils is so nearly 

 constant that determinations of the latter serve to indicate variations 

 in the content of organic matter. Table I shows the nitrogen content 

 ' of the successive levels within the first foot of soil on the five plots. 

 The percentages reported in the first part of the table (a) are for com- 

 posite samples from six borings using a 4-inch plate auger and those in 

 (b) for composites of 24 samples taken by means of a 1.5-inch soil tube 

 of special construction. The latter are really the averages of the data 

 from two sets of samples, A and B, in each of which composites of 12 

 individual samples were employed, these being distributed as indicated 

 in figure i, C. Vae concordance of the data from these two sets, as 

 illustrated by Table II, is such that in the present discussion no purpose 

 would be served by reporting more than their averages. From these 

 data it is evident that any marked differences in nitrogen content, and, 

 hence, in organic matter found between samples from the same level 

 on different plots are to be attributed to differences in crop history and 

 not to the experimental errors of sampling. 



