356 
Psyche 
[Vol. 85 
degree rule” is intriguing but far from universal and applies only 
partially to the lacewing calls described here; why the rule should 
apply at all, to any insect call parameter, is still unknown. 
It should not be assumed from the discussion above that linear 
regressions characterize the temperature relationships of all insect 
song parameters. In fact, it is likely that all temperature regressions 
are ultimately exponential in form when based upon data taken over 
a sufficiently wide temperature range, since the kinetics of the phys- 
iochemical processes underlying song production are non-linear 
with respect to temperature. One conspicuous example is the pulse 
(or chirp) duration of the French tettigoniid Ephippiger provincia- 
lis (Yers.), which varies inversely with temperature in a “hyperbolic” 
manner (Dumortier 1963, fig. 229). The volley duration typical of 
calling Ch. plorabunda also decreases with temperature (Fig. 3E), 
but the function describing that decrease seems more linear than 
hyperbolic or logarithmic over the chosen range of temperatures, 
and the slope of the relationship is twice as steep as that shown for 
the tettigoniid. Unless thermoregulation by the larger-bodied katy- 
dids accounts for these differences, one can conclude only that dis- 
tinct physiological mechanisms determine pulse, chirp, or volley 
durations in different insect groups. 
Recently, Michelson et al. (1982) published a comprehensive 
theoretical and empirical study treating general aspects of the phys- 
ics, transmissibility, and energetics of the vibrational songs of 
insects. This study provides a rationale for the observed frequency 
modulation of Ch. plorabunda volleys, and by implication helps to 
explain why lacewing singers strictly control the frequencies of all 
portions of their volleys, rather than simply allowing their abdo- 
mens to oscillate at their frequencies of resonance. Viewing the plant 
substrates of vibrating or tremulating insects as acoustical filters, 
Michelson et al. concluded that signals consisting of multiple fre- 
quencies should propagate more uniformly (and effectively) through 
their substrates than narrow-bandwidth calls, since a signal of rela- 
tively pure tone tends to excite a certain pattern of standing waves in 
its substrate and will therefore vary tremendously in its intensity 
from place to place on that substrate. Broad-bandwidth or frequency- 
modulated signals, on the other hand, are ideally suited for penetrat- 
ing such acoustical filters, so that at least a portion of the call’s 
energy reaches the receptors of the recipient individual or partner. 
Although the authors reported frequency changes on the order of 
