144 ILLINOIS STATE ACADEMY OF SCIENCE 



of borings per plot necessary to get a representative aver- 

 age. The following formula was used for computing the 

 probable error of the average. 



P. E. = (-)O. 8453 . ^^rlv^^^ 



I n(n-l) 



In which ^ (— ) (V) = the sum of the deviations from the mean, 

 their sign being disregarded 

 n = number of borings entering the average. 



It was at once found by inspection that three borings 

 per plot were an insufficient number to give a reliable aver- 

 age, and the possibility of using this small number was at 

 once abandoned. Tables 1 and 2 give the data secured in 

 this preliminary work. 



It will be noted that in Plot 4 the highest probable error 

 occurred in the 7 to 12 inch borings, in Plot 7 in the surface 

 borings, and in Plots 10 and 13 in the 19 to 24 inch 

 borings. No explanation is apparent for the high variation 

 found in the above named strata in Plots 4 and 7. In the 

 case of Plots 10 and 13, the wide variations found in the 

 moisture content of the 19 to 24 inch borings seem to be ac- 

 counted for by the fact that the tight clay stratum does not 

 occur at a uniform depth, and consequently some borings 

 contained more of this material than others. This conclu- 

 sion led to the abandonment of this plan of taking the bor- 

 ings at these arbitrary depths and the substitution, in all 

 subsequent work, of the following depths: Surface, 0-8 

 inches, subsurface, 8 inches to the tight clay, subsoil, 6 

 inches of tight clay. These depths were chosen because 

 they coincided with the very typical strata as they occur 

 in this field. 



During the spring of 1919, six sets of samples of 5 borings 

 per plot were taken from Plots 1, 2, 3, and 4 of Series 100 

 and Plots 13, 14, 15, and 16 of Series 400. Series 100 had 

 been plowed the preceding fall for corn and was undis- 

 turbed the following spring due to being too wet, until just 

 prior to taking the last set of samples on June 19. 



The plots of this series had received two tillage and dyna- 

 mite treatments prior to this season's sampling, the first in 

 the fall of 1914 and the second in the fall of 1918. Ap- 



