NOISES OF INSECTS. 407 
domen, is observable?.. In this is the true drum, the 
principal organ of sound, and its aperture is to the 
Cicada what our larynx is to us. If these creatures are 
unable themselves to modulate their sounds, here are 
parts enough to do it for them: for the mirrors, the 
membranes, and the central portions, with their cavi- 
ties, all assist in it. In the cavity last described, if you 
remove the lateral part of the first dorsal segment of the 
abdomen, you will discover a semi-opaque and nearly 
semicircular concavo-convex membrane with transverse 
folds—this is the drum>. Each bundle of muscles, be- 
fore mentioned, is terminated by a tendinous plate 
nearly circular, from which issue several little tendons 
that, forming a thread, pass through an aperture in the 
horny piece that supports the drum, and are attached to 
its under or concave surface. ‘Thus the bundle of mus- 
cles being alternately and briskly relaxed and contracted, 
will by its play draw in and let out the drum: so that 
its convex surface being thus rendered concave when 
pulled in, when let out a sound will be produced by the 
effort to recover its convexity; which, striking upon the 
mirror and other membranes before it escapes from un- 
der the operculum, will be modulated and augmented 
by them*. I should imagine that the muscular bundles 
are extended and contracted by the alternate approach 
@ Reaum. ubi supr. f. 3. U1. BTbidf. 6: ¢¢. f. 9. 
¢ Prare VIII. Fic. 19. cc. The figure given in this plate does 
not show the drums clearly; but the principal object of it was to 
exhibit the bundles of muscles, which are of a different form from 
those in Reaumur’s figures. In the above figure, a. is the mzrror ; 
bb. the bunches of muscles; cc. the drums; d. the back of the ub- 
domen; e. the belly. 
