55 



3, 8,069 for Plot 4, and 10,096 for Plot 5— an average of 9,046 for the 

 three experimental plots, which is more than three times the number 

 seen escaping from the check. 



In order to show how much of the foregoing migrations ought to 

 be attributed to the merely ordinary underground movements of the 

 ants to and fro when undisturbed, a single furrow was plowed July 

 29, down the center of each of the five plots, its whole length, and the 

 movements of the ants across these furrows were observed nightly 

 from that date till August 24, a period of 27 days. During this time 

 migration movements were started across the furrow as follows : Plot 

 1, 22 lines; Plot 2, 14 lines; Plot 3, 19 Hues; Plot 4. 16 lines; Plot 5, 

 14 lines, — an average of 17 lines per plot. We ought evidently to sub- 

 tract this number from the experimental numbers given above, as an 

 allowance for the results of the usual movements of the ants under 

 ground, the residues remaining as the effect of the stirring of the 

 soil. So corrected the numbers for the lines of the experiment are: 

 Plot 1, 7 lines; Plot 3, 36 lines; Plot 4, 41 lines; Plot 5, 43 lines, — an 

 average of 40 lines for the experimental plots as against 7 for the 

 check ; five and a half times as much migration out of the former as 

 out of the latter. Indeed, the movements across the central furrows 

 of the plots plowed July 29, are the proper check on the results of a 

 disturbance of the soil, and so applied we find the lumibers of mi- 

 gration lines last mentioned may be taken as those due to the treatment 

 of the several lots. The numbers of ants involved in movements across 

 these five furrows averaged 4,383 to the plot, a number to be com- 

 pared with the average of 9,046 seen on migration lines between the 

 experimental plots and the intermediate unbroken strips. This com- 

 parison indicates that more than half the recorded active movements 

 of the ants were due to the treatment which the plots had received. 

 From these various experiments and comparisons it is evident that 

 the great majority of the migration movements of the ants were move- 

 ments of escape from the area in which they were being disturbed. 



Effect of Disking and Subsequent Rolling 



The freedom and activity of the migration movements of the ants 

 in the very loose open soil of our 1909 Bloomington field led me to 

 surmise that the so-called cultivation method might be improved upon 

 bv using a heavv roller to compact the earth after disking, and the ef- 

 fects of this treatment were tested the following year. 1910. at Gales- 

 burg. Here a plot forty-six rows wide and a hundred and ten rows 

 long was plowed May 5 to a depth of six inches, disked three times, 

 harrowed once, and rolled after planting, another plot of the same 



