45 
tares to which the insects were siib]'ected. Main' other reconls are 
questionable by reason of the fact that the t^eneration. or the nearness 
to the maximum of the generation, of the insects phiced in the cage 
was uncertain or unknown. 
Early in his studies of the life liistory of this insect the writer saw 
the necessity of linding some method by which the numbers of indi- 
viduals of a generation could l)e approximated at certain times. By 
an incidental study of the records of larvjv captured under bands, 
published by Professor Aldrich in lOno. it was noted that at a certain 
time in the season there were fewer larvie so caught than at i)eriods 
of time immediately following and preceding. By collecting as many 
records as were ol)tainable at that time, it was ol)served that these 
conditions werecjuite constant. The periods of the larger and siualler 
number were termed, respectively, the maximum and mininuun of 
larva? entering bands. 
In 1001 many fruit growers in Idaho, at the recpiest of the writer, 
kept and submitted records of the larva' killed under bands. Other 
records, many of which had been made without any idea of the future 
use to which thev might be put, were collected from many sources. 
These records were tabulated and curves were drawn upon cross-sec- 
tion paper, using the time as one factor and the number of larvae as the 
other. These curves give quite an accurate picture of the course of the 
insect in the orchards throughout the season. Not all of the records, 
however, were satisfactory, as a few of them from various causes gave 
data which were of no value. The curve showing the efl'ective tem- 
perature at the dates at which the larvc\? were killed under the l)ands 
was drawn upon the same charts and gives (juite accurately the etlVct 
of the tenq^erature upon this hai)it of the insect. A number of these 
records are reproduced (tigs. 5 to IG). 
INACCl'KACIKS OF THK KKCOKDS. 
There are many sources of possible inaccuracy in these records. 
The greatest inaccuracy is probably found in the weekly or biweekly 
band records, because these are composite records of many individutil 
trees and show only approximately the dates of the maxima and min- 
ima, ^lan}' of the records were commenced too late in the season to 
be of any real value: and when thev were started even a little late the 
curve ascends with rapidity, which would not have happened had the 
record been started earlier. In consequence of a series of warm days, 
the maxinumi number of larva.^ may enter the bands sooner than they 
w^ould if the tenq^erature had remained normal: and if the tenqxM-ature 
be low for many days, the maximum might be later than it would be 
normally. Spraying might seriously interfere with the accuracy of 
the record, as at certain ])eriods all of the larva^ entering the fruits 
might l)e killed and thus cause a fall in the curve of larvie entering 
