BIOLOGY OF SATURNIID MOTHS—BLEST 463 
regulated flight behavior which the present theory demands. Simi- 
larly, in the period immediately before death, a mere 2 to 3 days after 
hatching, the procryptic Lonomia cynira shows similar unregulated 
activity. And, as we would expect, the rate of decay of the inhibitory 
“clock,” as measured by the changes in the rocking potential, is ex- 
tremely fast. Overt rocking responses in this species may disappear 
by the end of the second day from hatching. Conversely, in the long- 
lived and aposematic D. (Periphoba) hircia, the rocking potential 
declines extremely slowly, and strong rocking responses can still be 
observed as much as 10 days after capture. 
However, we still require exact measurements of the characteristics 
of these clocks; and, as yet, there is nothing known as to the way that 
sexual activity may modify their performance, although there seems 
to be a possibility that it may do so. Certainly this is a very special 
case: even in the closely related Citheroniinae and Saturniinae there 
seems to be no parallel to this clock system. For ourselves the con- 
comitants of the aging process present a familiar picture of failing 
sexual powers, increasing rigidity of outlook, and overall physical 
deterioration; the hemileucines, on the contrary, expire in a final 
blaze of hyperactive glory. Why then, should their aging and death 
so particularly interest us? 
The answer to this is twofold: First, as has already been mentioned, 
no previous biological material has been able to suggest precise paths 
through which natural selection can act to alter longevities. It has 
generally been assumed that selection, where it is able to act, must of 
necessity tend to lengthen lifespans (Williams, 1957). The recognition 
that, given an initially restricted lifespan, the mechanisms of aging 
may be of such a kind as to allow selection to modify the rates at which 
they proceed in either direction, may open the way to a more ration- 
ally planned approach to some of the problems of causation. Sec- 
ond, the apparent limitation of the lifespans by a neural clock implies 
that in this case the phenomena of specific lifespan are perhaps con- 
trolled by a unitary leading process. Now this process is certainly 
one which is unlikely to have any very general application in the 
animal kingdom. Attempts are still being made to provide general 
theories of senescence; examples such as that of the Hemileucinae 
stress the difficulties attending these oversimplified assaults upon what 
is beyond doubt a complex and specifically variable problem. 
This report has had to cover a good deal of ground in a somewhat 
perfunctory manner, and it may be felt that some of the speculations 
go too far beyond the existing evidence; it is, in fact, less a report 
than a blueprint for future research. The justification for these 
extravagances lies in the way in which the whole course of this work 
illustrates the often fortuitous advantages that may be derived from 
a broad evolutionary study conducted in a tropical environment. 
536608—60-——31 
