MOTIONS OF INSECTS. 355 



and asserts, that if either a poiser or winglet be cut off, 

 the insect will fly as if one side overbalanced the other, 

 till it falls to the ground ; and that if both be cut off, they 

 will fly awkwardly and unsteadily, as if they had lost some 

 very necessary part^. Shelver cut off the winglets of a 

 fly, leaving both wings and poisers, but it could no longer 

 fly. He next cut off the poisers of another, leaving the 

 wings and winglets, and the same result followed. He 

 found, upon removing one of these organs, that they 

 were not properly compared to balancers. Observing 

 that a common crane-fly (Tijmla crocata) moved the 

 knee of the hinder tibia in connexion with the wing and 

 poiser, he cut it off, and it could no longer fly : this last 

 experiment, however, seems contradicted by the fact, 

 which has been often observed, that the insects of this 

 genus will fly when half their legs are gone. He after- 

 wards cut off both its poisers, when it could neither fly 

 nor walk. Hence he conjectures that the poisers are 

 connected with the feet, and are air-holders ''. I have 

 often seen flies move their poisers very briskly when at 

 rest, particularly Seioptera vihrans, before mentioned. 

 This renders Shelver's conjecture — that they are con- 

 nected with respiration — not improbable. Perhaps by 

 their action some effect may be produced upon the 

 spiracle in their vicinity, either as to the opening or 

 closing of it. 



There are three classes of fliers in this order, the form 

 of whose bodies, as well as the shape and circumstances 

 of their wings, is different. First are the slender flies — 

 the gnats, gnat-like flies, and crane-flies (Ttpularue), 



* Phi/s. Theof. 13th Ed. 366, note («'.) 

 ^ Wiedemann's Archiv. ii. 210. 



2 A 2 



