Sh ee 
227 
The experiments were carried on in three series, kept as nearly 
independent as possible so that each would serve as a check on the others. 
I shall designate them as the large-cage series, the life-history series, 
and the band-record series. 
Fic. 4. Two oviposition cages in place on tree. 
LARGE-CAGE SERIES 
Cages used in this series were large enough to cover entire trees, 
the largest cages being 18 feet square and 18 feet high. (Fig. 5, 6, and 7.) 
The plan was to use two cages each for the first and the second genera- 
tions and one cage for the third generation—if a third generation appeared. 
The framework of the cages, consisting of 24 scantling, was 
erected over the trees before they were in bloom, and wire cloth, 12 wires 
to the inch, was tacked over the frame as soon as the time of full bloom 
was past. Each of the cages had a tight-fitting screen-door provided 
with hasp and padlock. 
The trees thus screened in were examined very carefully for larve 
and two inches of the soil in each cage was removed. 
The first moths to emerge from the hibernating larve were placed 
in cage No. 1, and the last moths to emerge in cage No. 2, to rear the 
first and the last individuals of the first generation; the first moths reared 
