MECHANISM OK INSECT FLIGHT. 203 



are formed of india-rubber membranes connected with the 

 two wings by bent levers ; the air when compressed or rarefied 

 gives to these flexible membranes powerful and rapid move- 

 ments, which are transmitted to both wings at the same time. 



A horizontal tube, balanced by a counterpoise, allows the 

 apparatus to turn upon a central axis, and serves at the same 

 time to conduct the air into the drum, which produces the 

 motion. This axis is formed of a kind of mercurial gaso- 

 meter, which hermetically seals the air conduits, while it allows 

 the instrument to turn freely in a horizontal plane. 



Thus arranged, the apparatus shows the mechanism by 

 which the resistance of the air, combined with the movements 

 of the wing, produces the propulsion of the insect. 



If we set in motion the wings of the artificial insect by 

 means of the air-pump, we see the apparatus soon begin to 

 revolve rapidly around its axis. The mechanism of the mo- 

 tion of the insect is clearly illustrated by this experiment, 

 entirely confirming the theories which we have deduced from 

 optical and graphic analysis of the movements of the wing 

 during flight. 



It may be asked whether the figure of 8 movements de- 

 scribed by the wing of a captive insect are also produced when 

 the creature flies. We have just seen that the bending of the 

 main- rib is entirely due to the force which carries the insect 

 forward when it has become free. We might therefore sup- 

 pose that the main-rib of the wing does not yield to this force 

 when the insect flies freely, and that the resulting horizontal 

 force is shown only by the impulsion of the whole of the insect 

 forwards. 



If, after having gilded the wing of the artificial insect, we 

 look at the luminous image produced during its flight, we still 

 see the figure of 8 remaining, provided the flight be not too 

 rapid. In fact, this figure is modified by the movement of 

 the apparatus ; it becomes more extended, and resembles the 

 8 registered on a revolving cylinder, but it is not reduced to a 

 simple pendular curve, which would be the case if the main- 

 rib were always rigid. We can understand that this may be 

 caused by the inertia of the apparatus, which cannot be 

 affected by the variable movements which each stroke of the 



