117 



towards the end of July, and again many hundreds about the 

 end of September. One of the two specmiens which I cap- 

 tured in July was injured and perished ; the other produced the 

 imago sometime in August. The kirvaa of the September 

 brood have passed into the pupa state. The larvae I found only 

 on the maple sprouts (^Acer rubruni) of three or four years' 

 growth. In consequence of its having escaped my observation 

 until last summer, I infer that its periods of occurrence in num- 

 bers must be remote, and the intervening years produce but 

 few specimens. The more I become acquainted with the econ- 

 omy of insects, the stronger becomes my belief that all insects 

 have their periods of increased numbers, which, in some in- 

 stances may be unfixed and irregular, but in others, their 

 periods of numbers are as fixed and regular as that of the 

 Cicada septemdecem. In the summer of 1832 I met with the 

 larva of Saturnia lo for the first time, and then in countless 

 numbers, on the top of oak saplings on a hill ; since that time I 

 failed not to visit that place, and similar localities, every sum- 

 mer, but was always disappointed in meeting with a single 

 specimen, either of the larva or perfect insect. 



