23 
It is possible that this list might be largely increased by examin¬ 
ing the various collections throughout the country, but as it is, it 
will indicate very clearly the dates at which the moths usually make 
their appearance north of the latitude of Washington City, or in what 
may properly be called the northern portion of the United States. 
NOTICES OF THE LARVZE. 
Illinois —Jackson Co.: 
First seen in 1861, about April 29, continued to appear in 
various parts of the county until May 15, and finally disap¬ 
peared about June 10. 
Latter half of September, 1874. 
Latter part May to first week in June, 1875. 
Carbondale, Jackson Co.: 
June 4 (1877) all the first larvae below the dirt to pupate. 
June 18, two moths appeared. 
June 21, four moths appeared. 
June 22, one moth appeared. 
Centralia: 
June 10, “not all gone." 
Pana: 
June 2, “worms at work;” -June 17, left about this date. 
Champaign: 
June 10-12, growth about completed.—(Emery.) 
Knox Co.: 
June 18, worms at work, not fully grown. 
Missouri —St. Louis : 
“Early part of May.” “During the latter part of April and 
throughout the month of May. 
“Early in June mowing down meadows.” “All through the 
month of August, and a few full grown individuals as late as 
the 23d of September.—(Riley.) 
Hannibal: 
June 8, here in myriads destroying the grass.—(Trabue—teste 
Riley.) 
Maine .—“About the middle of August, and soon after the cater¬ 
pillars entered the ground to transform.”—(Packard.) August 
2, 13.—(Packard.) 
Massachusetts t —Was noticed first of July.—(Packard.) 
The plan we propose to adopt in this investigation is as follows: 
Taking the facts ascertained by direct experiment with specimens 
in confinement, we use these as a basis, assuming that they, as a 
rule, indicate the minimum length of the stages. We then compare 
the dates at which the worms appear and disappear in different 
sections, the dates at which the moths are seen, and such other 
facts as we have relating to their life history, in a state of nature. 
The results thus obtained are compared with and checked by the 
data obtained by experiments. As a second check upon our calcu¬ 
lations as to the length of the various stages, we endeavor to ascer- 
