48 



but these records neither add to nor subtract from the significance of 

 the data of difference in final yield already given, and hence need not 

 be reported in detail. 



The general result of this field experiment with repellents applied 

 with a fertilizer dropped in the hill leads us to prefer this to the direct 

 application of the repellent substance to the seed-corn, and encourages 

 us to believe that this may sometimes serve as a partial and temporary, 

 but highly profitable protection to the young plant. It may be used to 

 best advantage where corn is planted on old corn ground which is 

 heavily infested by ants having an abundance of- root-aphid eggs in 

 their possession. 



The use of repellents for the protection of young corn has at best, 

 however, some very serious drawbacks apart from the cost of the ma- 

 terials used. It is necessarily only temporary, since the repellent sub- 

 stances applied must evaporate in order to be repellent, and hence will 

 presently weaken and finally disappear. It is uncertain in its results, 

 very wet weather tending to wash away the fluids; and there still re- 

 mains the danger that under certain weather conditions it may injure 

 the seed or the young plant even when applied with the fertilizer as a 

 carrier, our experiments not having been sufficiently numerous and 

 varied to cover all possible or probable variations in moisture, temper- 

 ature, and the like. Since repellents do not kill or directly injure either 

 ants or aphids, they can do little or nothing to diminish immediately 

 the actual infestation of the field. Furthermore, they do not wholly 

 exclude these insects from the corn fields, but only diminish for a time 

 the number which find life tolerable among the roots in the treated 

 plots. Finally, our experiments with plots separated by untreated 

 strips or checks are open to the suspicion that the ants repelled are 

 simply driven temporarily to these check strips, with the result that 

 our data of difference of treatment between check and experimental 

 plots may be exaggerated. It is possible that if an entire field were 

 treated, the ants, having no untreated cover in which to take refuge, 

 would remain in the treated hills in larger numbers than they did un- 

 der our experimental conditions. Some method which will actually 

 destroy the aphids or ants, or both, in a heavily infested field, is there- 

 fore much to be desired, and if this method is simply a variation or 

 modification of ordinary agricultural practice, requiring no special ap- 

 paratus or materials, it will have great and obvious advantages. 



Rotation and Cultivation Methods 

 Two such methods have been especially advised : one, a quick 

 rotation of crops such that no field will be in corn more than one or 



