2o8 THE INTEGUMENTAL SKELETON OF THE IMAGO. 



most exhaustive. He gives an elaborate description of the 

 wings and wing-joints. The structure of the wing-joints and 

 the mechanism by which they are moved differs entirely from 

 that observed in the Coleoptera, , Lepidoptera, Hymenoptera, 

 and Diptera. In the Dragon-flies the great wing muscles, as 

 was long ago observed by Meckel and Chabrier, are inserted 

 directly into the wing roots, and not into the thoracic wall. 

 Except in the existence of three roots to each wing, I have been 

 unable to compare the structures exhibited in the Dragon-flies 

 with those of other insects ; and the nomenclature employed by 

 Lendenfeld is special. He makes no attempt to compare the 

 wings of the Dragon-flies with those of other insects. Three 

 sets of muscles are, however, found to each wing, corresponding 

 with the three wing-roots. 



Lendenfeld took instantaneous photographs of the wings in 

 motion, and claims an exposure of o-qW of a second. From a 

 series of such photographs he has constructed a projection 

 giving the curves described by the wings. The curve is similar 

 to the one I have given (Fig. 36), but he has reversed the 

 direction of the motion. So far as I can judge, there was 

 nothing in Lendenfeld's method to determine the direction, and 

 in a curve given by him [93, p. 368] he apparently agrees with 

 me completely, not only in the direction of movement, but — if 

 the arrows are reversed in his large diagram to correspond with 

 those given in this curve — in the rotations of the wing plane. 

 The curves are constructed from photographs of the wings of 

 Agrion puella. 



Lendenfeld, like myself, rejects Marey's theory of passive 

 rotation. He has amplified Marey's tables, giving the relative 

 wing surface in birds and insects. His results show that the 

 wing surface in insects is from twenty to a hundred times greater 

 than in birds for equal body weights. This relation enables the 

 insect to compete with a bird in actual as well as in relative 

 velocity. Leeuwenhoek's old observation on the relative 

 velocity of a swift and a dragon-fly, in which the insect out- 

 stripped the bird, is well known and often q^jpted. 



