The Nuptial Flight 115 
I have said that these insects were attracted by the female 
cecropias; this statement needs to be amended somewhat, for 
in the last four experiments, 28 male cecropias came to the 
cages containing only the cynthia females, after the cecropia 
females were all dead; they were deceived by the similar or re- 
lated cynthia emanations. The behavior of the cecropia males 
showed, however, that the deception was not complete. 
A summary of the data shows (see Table No. 1), that out of 
735 marked males liberated in various directions and at distances 
varying from 25 feet to 3 miles, 170 returned to the females 
on the roof. This was 23.1 per cent of those liberated. If this 
result seems small to the reader, let me remind him that in the 
region where 791 wild males came to us voluntarily, there must 
have been an enormous number of females in the open to en- 
tice our experimental moths from their path; hence to me it 
seems really wonderful that any at all should return to a bleak, 
Second-story roof in an area of roofs and brick walls. The per- 
centage of returns of this species, a native of Missouri, com- 
pares well with the returns of the cynthias, which had the field 
completely to themselves, since in the wilds of this city there 
are no females of his species to waft her emanations on the 
breezes and thus swerve him from his course. Later pages will 
show that the per cent of returns of cynthia was 31. In the 
light of this comparison, I feel doubly assured that the returns 
of cecropia are good. 
A tabulation of the figures shows that the 170 which were 
liberated at various times during the evening came in during 
the dawn* of the first, second, third and fourth morning in the 
following proportion: 
1st morning, 106. 
2nd morning, 50. 
3rd morning, 11. 
4th morning, 3. ; 
The time of flight is -defined with surprising sharpness; it 
seems to be the established habit of this species to make the 
nuptial flight during one hour, just before and during the 
Period of dawn, from 3:30 to 4:30 o’elock, with deviations from 
this time seldom exceeding five or ten minutes, excepting in the 
©ases mentioned in the footnote. 
ed 
*In a few experiments the moonlight elicited an earlier response. 
