237 



were seen crawling about outside, where their young might easily have 

 passed through the meshes of the cloth and reached the plant within. 

 This 3'ear, with a very large cage covering four planted hills thoroughly 

 stocked with root lice there has not been a trace of the aerial louse ; and 

 all attempts to evolve it by directly transferring the root louse to the 

 leaves have failed. 



Passing now to the other end of the season, we find that the aerial 

 lice continue to breed generation after generation of both winged and 

 wingless viviparous females until the autumnal cold and the perishing 

 of their food plants destroy them en masses leaving behind no trace or 

 remnant of a hibernating brood, nor evolving, so far as we have been 

 able to discover, any oviparous generation. These aerial lice pass rap- 

 idly and freely from plant to plant in the fall, concentrating thus on the 

 latest remnants of green vegetation about the corn, and spreading 

 likewise to the perennial grasses around the borders of the field. We 

 have dissected them by the hundred at this season, finding only fe- 

 males and these all viviparous, and have bred them in warm breeding 

 rooms throughout the winter, no less than nine generatians in succes- 

 sion occurring between October 7 and the 8th of March. In all these 

 winter generations no trace of oviparous females occurred, and no va- 

 riation of temperature or exposure made any appreciable change in the 

 form or habit of the louse. 



The aerial louse is extremely like the apple louse {AjyJiis mali), and its 

 general disappearance in autumn from the corn at a time when this 

 last species is laying its eggs freely on the apple twigs, led us to test 

 the hypothesis of a migration between these plants. These experi- 

 ments, persistently repeated, were, however, quite without result. The 

 corn lice inclosed in autumn under bell jars with fresh apple twigs 

 neither bred nor fed upon them, and invariably perished. Similar ex- 

 periments have not been made, however, with either grass or wheat, al- 

 though the aerial louse has been found in fall upon both. Besides a 

 repetition of these and similar experiments, the most promising still 

 to be made are those for the transfer of successive generations of the 

 winged root lice in June and July to the leaves and springing tassels 

 of corn and sorghum. 



As long as the connection between root and aerial forms remains in 

 doubt, all economic discussion must be of a provisional and tentative 

 character. Some observations and experiments on the root louse are, 

 however, worth reporting. 



In the first place, a long list of observations in the field in early 

 spring unite in showing that the corn-root aphis takes its start only in 

 fields where it occurred the year before, and that such fields are, as a 

 rule, most likely to suffer severely from the attack. The early evolu- 

 tion of a partly winged brood provides, however, for so general a dis- 

 persal that the expedient of rotation of crops can have only a second- 

 arv value. 



