THE HARVEST-FLIES. 205 



are alternately tightened and loosened, recovering their nat- 

 ural convexity by their own elasticity. The effect of this 

 rapid alternate tension and relaxation is the production of a 

 rattling sound, like that caused by a succession of quick 

 pressures upon a slightly convex and elastic piece of tin 

 plate. Certain cavities within the body of the insect, which 

 may be seen on raising two large valves beneath the belly, 

 and which are separated from each other by thin partitions 

 having the transparency and brilliancy of mica, or of thin 

 and highly polished glass, tend to increase the vibrations of 

 the sounds, and add greatly to their intensity. In most of 

 our species of Cicada the drums are not visible on the out- 

 side of the body, but are covered by convex triangular 

 pieces on each side of the first ring behind the thorax, which 

 must be cut away in order to expose them. On raising the 

 large valves of the belly, hoAvever, there is seen, close to 

 each side of the body, a little opening, like a pocket, in 

 which the drum is lodged, and from which the sound issues 

 Avhen the insect opens the valves. The hinder extremity 

 of the body of the female is conical, and the under side 

 has a longitudinal channel for the reception of the piercer, 

 which is furthermore protected by four short grooved pieces 

 fixed in the sides of the channel. The piercer itself consists 

 of three parts in close contact with each other ; namely, two 

 outer ones grooved on the inside and enlarged at the tips, 

 which externally are beset with small teeth like a saw, and a 

 central, spear-pointed borer, which plays between the other 

 two. Thus this instrument has the power and does the work 

 both of an awl and of a double-edo;ed saw, or rather of two 

 key-hole saws cutting opposite to each other. No species of 

 Cicada possesses the power of leaping. The legs are rather 

 short, and the anterior thighs are armed beneath with two 

 stout spines. 



The duration of life in winged insects is comparatively 

 very short, seldom exceeding two or three weeks in extent, 

 and in many is limited to the same number of days or hours. 



