NOISES OF INSECTS. 487 



may cause some motion in them, enough to occasion friction. With 

 respect to Dij^tera, Latreille contends that the noise of fiies on the wing 

 cannot be the result of friction, because their wings are then expanded ; 

 but though to us flies seem to sail through the air without moving these 

 organs, yet they are doubtless all the while in motion, though too rapid for 

 the eye to perceive it. When the aphidivorous flies are hovering, the ver- 

 tical play of their wings, though very rapid, is easily seen ; but when they 

 fly off" it is no longer visible. Repeated experiments have been tried to 

 ascertain the cause of sound in this tribe, but it should seem with different 

 results. De Geer, whose observations were made upon one of the flies 

 just mentioned, appears to have proved that, in the insect he examined, 

 the sounds were produced bj' the friction of the root or base of the wings 

 against the sides of the cavity in which they are inserted. To be con- 

 vinced of this, he affirms, the observer has nothing to do but to hold each 

 wing with the finger and thumb, and stretching them out, taking care not 

 to hurt the animal, in opposite directions, thus to prevent their motion, — 

 and immediately all sound will cease. For further satisfaction he made 

 the following experiment. He first cut off' the wings of one of these flies 

 very near the base; but finding that it still continued to buzz as before, 

 he thought that the winglets and poisers, which he remarked were in a 

 constant vibration, might occasion the sound. Upon this, cutting both off", 

 he examined the mutilated fly with a microscope, and found that the 

 remaining fragments of the wings were in con>tant motion all the time that 

 the buzzing continued ; but that upon pulling tliem up by the roots ail 

 sound ceased.^ Shelver's experiments, noticed in my last letter, go to 

 prove, with respect to the insects that he examined, that the winglets are 

 more particularly concerned with the buzzing. Upon cutting off" the wings 

 of a fly — but he does not state that he pulled them up by the roots — he 

 found the sound continued. He next cut off' the poisers — the buzzing 

 went on. This experiment was repeated eighteen times with the same result. 

 Lastly, vvhen he took off" the winglets, eitlier wholly or partially, the buzz- 

 ing ceased. This, however, if correct, can only be a cause of this noise in 

 the insects that have winglets. Numbers have them not. He next, there- 

 fore, cut off" the poisers of a crane-fly {Tipula crocafa), and found that it 

 buzzed when it moved the wing. He cut off" half the latter, yet still the 

 sound continued ; but when he had cut off" the whole of these organs the 

 sound entirely ceased.^ 



Dr. Buraieister, however, was led by his experiments to a different con- 

 clusion. Finding that the buzz of a lai-ge fly {Eriskdis tcnax) still con- 

 tinued after the winglets, the poisers, and even the wings had been quite 

 cut off" except their very stumps (only in this last case the sound was 

 somewhat weaker and higher), he conceived that the spiracles lying between 

 the meso- and meta-thorax nmst be the instruments of the sound, which 

 accordingly he found to cease entirely when they were stopped with gum, 

 though while the wings were in vibration. Pursuing his researches, he 

 extracted one of these spiracles, and opening it carefully, found its poste- 

 rior and inner lip, which is directed towards the commencement of the 

 trachea, to be expanded into a small flat crescent-shaped plate, upon which 

 are nine parallel very delicate horny laminse, the central one being the 



1 De Geer, vi. 13. 2 Wiedemann's Arcldv. ii. 210. 217. 



114 



