39 
field had been heavily infested by root-lice the preceding year was 
shown by the number of aphis eggs and young aphids in possession of 
the ants — the aphids on roots of smartweed and grasslike weeds in the 
field. The ants themselves lived mainly under ground, at least by day, 
rarely opening their burrows to the surface, until the 4th of May, 
when, as the weather warmed up after a heavy rain, they made their 
presence known by a deposit of pellets of earth at the surface around 
the openings of their burrows. The oats at this time were about four 
inches high, and smartweeds and ragweeds about half as tall. Both 
the latter had many root-lice on them, but none were on the roots of 
the oats, and the aphis eggs were not yet all hatched. Nests of the 
ants were very abundant in the grass outside the borders of the field, 
and several of these which were opened up contained corn root-aphids 
on the roots of the grass. From the 8th to the 10th of May the ants 
were actively running over the ground as if in search of food, but, so 
far as could be seen, with meager result. The second generation of 
aphids had by this time begun to appear, and many of the smartweeds 
were dead at the root. The ground was very dry for the next few 
days, and both grain and weeds grew very slowly, many of the oat 
plants being dead. Thus matters went on, with no material change 
except that the ants were found feeding freely on dead June-beetles 
and other insects, until June 5, when it was noticed that the ants were 
leaving the field. 
"As I was returning home last evening," says Mr. Kelly, "I no- 
ticed crossing the lane beside this field a colony of Lasiiis niger amcri- 
canus, composed of ants going in both directions between the field of 
oats and the corn field on the opposite side of the road. I followed 
the column thru a hedge fence and about twenty feet into the oats, 
where they were coming out of their nest. Some of those going out- 
ward were carrying larvae, but none of those returning in the other 
direction. The line of march was indirect, and about fifty-seven feet 
long, ending in the grass on the opposite side of the road, where a 
new nest was being formed. The colony was still moving at 6 the 
following morning, evidently having been at work all night. They 
had finished the transfer before 8 a. m., nothing remaining in the nest 
from which they had emerged. No root-lice were in their possession, 
but only their own larvse." 
Finally, on the 22d of June, a thoro search was made of the site.-? 
of all the nests marked as originally occupied by ants. Forty-nine of 
the fifty were identified, the remaining one being lost. Forty-four of 
these nests had been completely deserted, and only five were still in- 
habited by ants. These were in rather open spots in the field, with an 
abundance of grass about them. There were no corn root-lice on this 
grass, however, but only the grass root-louse {Schizoneura panic ola) 
and a few Geoica squamosa. The oats at this time reached about to 
the knee, and were beginning to head. 
From these data, combined with those previously published, we 
may infer a gradual but general migration of the ants from the old 
