THE CORN ROOT-APHIS AND ATTENDANT ANT. 37 



An excei^tion should, perhaps, be made to this last statement with 

 respect to the aphis ant itself. The mysterious disappearance lasr 

 May of all the ant and aphis population of a 20-acre field of oats in 

 Avhich the food plants of the aphis had nearly all been killed led me 

 to experiment with ants and aphides kept in confinement imder star- 

 vation conditions. A colony of ants, well established in an artificial 

 nest of the kind devised by Miss Adele M. Fielde, were given a 

 group of aphides on a fresh corn plant. This was then allowed 

 gradually to dry up until it no longer afforded food to the aphides, 

 and these, of course, soon ceased in turn to make their usual offerings 

 to the attendant ant. Seeming to despair, at length, of making any 

 further use of their aphide proteges, the ants finally ate them up, 

 and continued so to do with all the fresh ones given them. 



I strongly suspect that the same thing happens sometimes in the 

 field, and that, when the food plants of an aphis family fail them 

 and no more can be readily found, their masters and owners — for 

 such their ant attendants really are — do the best thing possible under 

 these hard conditions and convert their milch cows into beef. 



PRACTICAL ECONOMIC MEASURES. 



The most obvious means of lessening or preventing injury by this 

 insect is a rapid rotation of crops, leaving no land in corn more than 

 a single year; but this measure is very inadequate, because, however 

 the field is finally cropped, it will at first spring up to smartweed 

 and pigeon grass and other common w'eeds, on the roots of which the 

 aphides live and breed abundantly until, say, the latter part of May. 

 Then if the field is sown to oats, the shading and sapping of the 

 ground by the crop plant will dwarf the weeds and thus reduce the 

 food supply of the aphides. One important consequence is shoAvn 

 by a comparative observation made last May on the composition of 

 the aphide colonies in hills of corn, as related to the density of the 

 aphide population. Fift}- considerably infested hills contained an 

 average of 105 aphides to a hill, and 225 lightly infested hills con- 

 tained an average of 5 aphides to a hill. In the lightly infested hills 

 21 per cent of the adults were winged, or Avere pupa> about to acquire 

 wings, while in the w^orst infested hills the winged and pup* 

 amounted to 64 per cent of the adults. That is, as the food supply 

 fails the aphides escape starvation, in great measure, by getting wings 

 and flying away. Many of these winged migTants doubtless perish, 

 but many of them are found by the ants, which are mining every- 

 where as if in preparation for their advent; and whenever one is 

 found a new family presently appears on a newly-infested plant, and, 

 multiplying at that time at a rate of 40 to 1 for every week, a fresh 

 start is soon made and the game goes merrily on. The effect of rota- 

 tion is thus a merely temporary check on increase, followed by a wider 



