61 
>ut in fields of small grain, Setaria and Polygonum may con- 
inue to support the corn root louse at least until the second 
generation is well matured. Indeed, I have found this insect on 
he roots of smartweed more than a foot high as late as June 
.7. Crab-grass (Panicum) also becomes infested, but less abund- 
tntly than the Setaria, and from the latter part of June through- 
>ut the rest of the season the aphis breeds abundantly on the 
:ommon purslane ( Portulaca oleracea) . We have occasionally 
ound it so abundant on purslane plants far removed from corn 
ields—beside paths in lawns and in other similar situations— 
hat one might well regard this as a purslane aphis, if it were 
lot for the fact that this weed starts too late in the season to 
ierve as food for the earlier generations. 
We have, further, experimental evidence that the corn root 
iphis can live on roots of ragweed (Ambrosia), having trans- 
erred May 8, 1889, half-grown young of the second generation 
rom smartweed roots to this plant, where they lived and fed 
intil they acquired wings, five days later. The fall oviparous 
generation and the one preceding it have been repeatedly re¬ 
ported by my field assistants—who were constantly dealing with 
he root aphis and knew its characters perfectly—to have abund- 
mtly infested dock ( Kumex crispus ), fleabane ( Erigeron cana- 
lense), mustard ( Brassica nigra), sorrel (Oxalis stricta), plantain 
Plantago major), Hungarian grass ( Setaria germanica) , pigweed 
Amarantus hybridus), and squash; but as these statements 
vere not verified by successful transfers from these various plants 
;o corn, they rest only on determinative evidence, notoriously 
mreliable with respect to the plant louse species. Indeed, an 
ittempt at transfers of the supposed corn aphis found on squash, 
sent me from Ohio by Prof. C. M. Weed, entirely failed. A sim- 
lar result was reached in an attempt to transfer known corn 
•oot lice from corn to wheat and oats, begun April 22, 1889. 
insects placed on roots of wheat in breeding cages April 22 con¬ 
tinued to live there until May 5, but without producing young, 
day 11, however, all had left"the plants. An earlier experiment. 
Degun April 10, had a like ending, and a precisely* similar result 
vas obtained in a parallel experiment with oats. 
The repugnance of this insect to the roots of small grain was 
’epeatedly shown also by field observations. Fields of oats and 
vheat on old corn ground, sometimes known to have been badly 
nfested by the root aphis the preceding year, often contained in 
ipril and May large numbers of these root lice and their asso¬ 
ciated ants, the former feeding on the roots of smartweed and 
Digeon-grass growing with the grain, but never being seen on the 
*oots of the grain even where these and the grass roots were 
closelv interlaced. 
The relation of this louse to other plants than corn has an 
mportant economic bearing. For example, infields on old corn 
ground, the first generation of plant lice are very noticeably 
nore abundant early in spring in the lower parts of the field 
bhan elsewhere, especially in those parts so situated as to re- 
