32 
conditions suitable to the development of most insects, and which 
hatched from eggs deposited September 16, died after descending into 
the sand and while still in the larval condition. Another lot of larvae 
received October 4 from Athens, Ga., gave forth adults during the 
second week of November. The consensus of opinion of writers on 
the subject is that the insect probably hibernates as pupae, with a 
smaller percentage of individuals passing the winter months as adults. 
A large number of larvae were obtained from the field and divided 
into two equal lots, one being placed on grasses and weeds on the 
Department grounds, and confined b}^ means of a large cover. The 
remainder were placed upon sand in a large rearing cage having the 
same sized cover and kept out of doors, where they were fed on 
fresh g*rasses and weeds from day to day. About a dozen of the sec- 
ond lot died while feeding, as a result of a cold spell with rain, which 
lasted for some little time, and the remainder, as far as could be ascer- 
tained, all perished as pupae in their earthen cocoons before trans- 
forming to moths. Similarly, the lot placed on the ground transformed 
and entered the earth, but no moths issued. 
It is practically proved that all egg masses deposited late in autumn 
produce larvae, and in the most northern limits of the insect, where 
cold weather sets in much earlier, it is probable that of these larvae 
only a very small percentage survive the average winter, and that the 
latest appearing generations in northern localities are the progeny, 
at least in many instances, of moths which have developed in, and 
migrated from, more southern localities earlier in the season. 
From experience with four lots of the fall army worm under vary- 
ing conditions and in different soils, the following deductions have 
been made in regard to pupation: 
PUPATION. 
When full grown the larvae enter the earth to a depth of from a 
quarter of an inch to not more than an inch and a quarter, the average 
appearing to be about half an inch, the depth not A T aiying at all accord- 
ing to the brood or the season, those transforming to pupa in late 
October and early November doing so at the same depth as others 
observed in August and September. 
Pupation of all individuals observed in early fall occurred in little 
earthern cells, which were nearly erect or more or less inclined, the 
head in all cases being nearest the surface, and those observed in com- 
pact earth had smooth walls, made by the constant revolving of the 
pupa within; but occasionally pupae were not thus protected. Most 
individuals which came under observation in cool October weather, 
and representing a later generation — in short, the last that can be 
produced in a region as cool as the District of Columbia — spun up 
more or less substantial cocoons of earth, lining them with an unusual 
