240 



PHENOMENA OF FLIGHT IN THE ANIMAL KINGDOM. 



obtain tracings of any extent, because the wing cannot remain long in 

 contact with the blackened cylinder, which it leaves and approaches 

 successively. Under these special conditions it is necessary to have 

 resource to an artifice, and since it is impossible to obtain a satisfactory 

 trace at a single stroke, we should try to divide the difficulty and sepa- 

 rate the operation into several periods. The preceding experiments sim- 

 plify the interpretation of the tracings very much, and we can recon- 

 struct the figures which the optic method has indicated from the slender 

 elements which they aftbrd. I have considered in the complete course of 

 the wing of an insect, such as is represented in Fig. 4, three distinct 

 zones, of which 1 have obtained the tracings separately; an inferior 

 zone, corresponding to the lower i^ortion of the figure of eight, a median 

 zone, and a superior zone corresponding to the middle and upper parts 

 of this figure. Bringing together the tracings obtained in these three 

 zones I have been able to reconstruct the entire curve. In registering 

 the tracings of the median zone, figures much resembling each other 

 are obtained, presenting the two crossed lines shown in Fig. 5. 



Fig. 5. 



Trace of the median course of the wing of the Macroglossa galium, (Bedstraw sphynx- 



moth.) 



The multiple tracings of the figure are formed by the fringed extrem- 

 ity of the wing, which presents many small points. The upper portion 

 is' in the form of a loop, as well as the part which corresponds to the 

 lower course of the wing, and these three parts successively obtained 

 give, when united together, the complete representation of a figure of 

 eight, such as is obtained in acoustics in registering by Kcenig's method 

 the vibrations of a Wheatstone's octave rod ; that is, a rod which vi- 

 brates twice transversely for each longitudinal vibration. The slower 

 motion of the cylinder produces the condensation of the end of the 

 tracing. 



Fiff. G. 



Trace of a Wheatstone's octave rod. 



The experiments can also be varied by obtaining, not the tracing of 

 the point of the wing, but that of the anterior border of this membrane, 

 striking laterally against the cylinder. It is clear that in describing the 

 upper loop, this edge will approach the cylinder, then deviating ; in a 

 similar manner it will descrilje the lower loop, so that in its complete 

 course it will rub twice against the blackened surface, and leave two 

 white traces sejiarated by an interval. This is observed in Fig. 7. 



