FLIGHT OF INSECTS. 



189 



Graphic method employed for the determination of the move- 

 ments of the wing. — The preceding experiments throw great light 

 on the traces which we obtain by the friction of the insect^ s 

 wing against the blackened cylinder. Although the figures 

 thus produced are for the most part incomplete, we are able, 

 by means of their scattered elements, to reconstruct the figure 

 which has been shown by the optical method. 



It is to be remarked that without sensibly interfering with 

 the movements of the wing, we can obtain traces of seven or 

 eight millimetres when the wing is rather long. The slight 

 flexure to which the wing is subjected allows it to remain in 

 contact with the cylinder to that extent ; we thus obtain a 

 partial tracing of the movement ; so that if we are careful to 

 produce the contact of the wing with the cylinder in different 

 parts of the course passed through by the limb, we obtain a 

 series of partial tracings which are complementary to each 

 other, and thus allow us to deduce from them the form of a 

 perfect curve of the revolution of a wing. Suppose, then, 

 that in fig. 71, the curve described by the gilded wing is 

 divided by horizontal lines into three zones : the upper one, 

 formed by the upper loop ; that in the middle, comprehending 

 the two branches of the 8, crossing each other and forming a 

 sort of X ; the lower one including the lower loop. 



By registering the movement of the middle zone, get 



Fig. 72.- -Tracing of the middle region of the course of the wing of a bee, 

 showing the crossing of the two branches of the 8. One of the branches 

 is prolonged rather far, but still the tracing of the lower loop has not been 

 produced. 



figures somewhat resembling each other, in which the lines, 

 placed obliquely with respect to each other, cut each other. 

 This is the case in fig. 72, the middle region of the tracing 

 of a bee, and in fig. 73, the middle portion of that of a 

 humming-bird moth. 



