452 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1959 
the 4,000-odd species of Lepidoptera recorded from the island may be 
assumed to feed in the canopy, where they are beyond the reach of the 
biologist. The rearing of tropical insects, too, is not made easy by the 
presence of their normal pathogens. To breed tropical saturniids 
in a European laboratory is paradoxically an easy business, for they 
will accept a wide range of substitute food plants and do not readily 
become infected with bacterial or virus disease. In the Tropics, the 
mortality from disease is high, and most species have specialized feed- 
ing habits which are not easily determined. 
For the greater part of the work, then, the moths were attracted to 
photoflood lights placed on the outside walls of the laboratories. 
While a few moths would generally come to light on most evenings, 
the peak emergences always occurred just before the new moon. They 
did not fly uniformly throughout the night; there appear to be bursts 
of activity at localized times, the main flights occurring in particular 
at about 1a.m.and4a.m. A given species, too, often has a preferred 
flight time, which in some cases may be highly restricted. Dirphia 
(Periphoba) hircia, for example, on Barro Colorado, tended to restrict 
its activity to the 4 a.m. flight, and on each night the whole flight of 
the small population near the laboratory was apparently completed 
within a 10-minute period. Most of the moths arriving at light 
were, from their perfect condition, quite clearly undergoing their 
first flight after emergence from the pupa, for when flying under 
tropical conditions these very large moths soon suffer conspicuous 
damage to their wings. 
Conveniently, these moths are highly resistant to a wide variety 
of quite drastic experimental procedures. Usually, after capture at 
an environmental temperature of some 26° to 29° C. they were stored 
in an icebox at 6° C. overnight, yet comparison with individuals which 
had not been so stored showed that their behavior was in no way 
impaired by this treatment. After they had been allowed to warm 
up to the surrounding temperature, their display behavior was ex- 
amined by gently stimulating them. Much of the actions that re- 
sulted was recorded on film and on 35-mm. color transparencies. 
The protective displays of saturniid moths, unlike those of mantids, 
do not require very specific stimuli for their release. Four subfami- 
lies are found on Barro Colorado: the Rhescyntinae, Citheroniinae, 
Hemileucinae, and Saturniinae. Of these four, only the Rhescyntinae 
and Saturniinae show any responsiveness when at rest to moving 
visual stimuli—shadows or solid objects in their vicinity, and the 
like. They will also respond to tactile stimuli, which alone elicit the 
protective displays of the other two subfamilies. 
The protective behavior falls into the following main categories: 
(1) Behavior which, teleologically speaking, is directed toward 
the quickest method of escape available to the insect. Thus, Copaxa 
