PHENOMENA OF FLIGHT IN THE ANIMAL KINGDOM. 285 



motions of the point of the wing. The glass plate moves from the right 

 to the left 5 the tracing is read from left to right. The head of the 

 bird is toward the left ; this flight is in the direction of the aiTOw. 

 We can divide this figure by vertical lines i^assiug through homologous 

 points, cutting it either at the top of the loops or at the summit of the 

 simple curves, as represented at the points a and e. Each of these divi- 

 sions incloses similar elements, although their development is unequal in 

 different parts of the figure. For the present we shall neglect these 

 details. 



It is evident that the periodical return of similar forms corresponds 

 to a return of the same phases in an evolution of the bird's wing. Tlie 

 division a e thus represents the different motions of the bird during an 

 alar evolution. 



Let us recollect that in the curve which we are analyzing all the mo- 

 tions are the reverse of those which the bird really executes. The two 

 vertical oscillations, the great and the small, should then be represented 

 by two downward curves. It is easy to recognize them in the great 

 curve ah c and the small curve c d e. Thus tbe bird rises from a to &, falls 

 from h to c, again rises from c to f/, and re-descends from d to e; but these 

 oscillations encroach on each other, producing the loop c d. The oscilla- 

 tion cde partly covers the first anteriorly. This is a proof that the indica- 

 tions of the curve are the reverse of the true motion ; for, at this 

 moment, the bird recedes, or, at least, relaxes its course. As the appa- 

 ratus is only sensible of changes of velocity, it is clear that the tracing 

 does not take the uniform rapidity of the bird into account, but indi 

 cates acceleration as a forward movement and retardation as a retro- 

 grade movement. This figure, then, sums up all the preceding experi- 

 ments which we have made on the motions of the bird in space. It is 

 here seen that the bird at each evolution of its wings rises and falls 

 twice, successively; that these oscillations are imequal ,• the larger, as 

 we know, corresponding to the depression of the wing, the smaller to 

 its elevation. It is also seen that the ascent of the bird during the 

 raising of the -svings is accompanied by a retardation of its speed, which 

 justifies the theory by which this ascent has been considered as made 

 at the expense of the bird's acquired velocity. But this is not all ; this 

 curve also shows us that the motions of the bird are not the same at 

 the beginning and end of flight. We have seen already (Fig. 20) 

 that the first strokes are more extended than the others ; we now see 

 that at first — that is, at the left of the figure — the oscillations produced 

 by the descent of the wing are also more extended. But theory fore- 

 told that the oscillation of the elevation of the wing being derived from 

 the acquired speed of the bird should be very feeble at the beginning 

 of flight when the animal has acquired but little impetus. The figure 

 shows us that this does happen, and that at the beginning of flight the 

 second oscillation (which forms the loop) is very insignificant. 



At last, then, we are in possession of the i^rincipal facts upon which 

 the study of the mechanical power developed by the bird during flight 

 can be established, and we see that it is during the descent of the wing 

 that the entire motive force which sustains and directs the bird in space 

 is created. 



