number of Ichneumon flies, we cannot conceive how it is poffible that these infects fhould live feveral months under ground in a place fo clofe and fo impenetrable to the air if they needed to breathe in it. If indeed they were to breathe even the air which is inclofed along, with them, fo fmall a portion, paffing fo often through their spiracles, and being tainted with the exhalations it could not but receive from them, would be of no use to the insect.

Neither would I maintain the refpiration of chryfalids; at least one experiment convinces me that there are some which do not always breathe. I took the chryfalis of the Sphinx Liguftri, which being one of the greatest is the most proper for experiments. It had, betides, the two anterior. ftigmata fo open that with a common magnifier I could fee into the fubftance of its body and obferve a final vacuum between it and the cover. This made me hope that if respiration took place in chryfalids, this would give me certain proofs of it. Two or three months before the perfect infect iffued from its cover, I dug it up, and covered, feveral times, first one, and then two, and afterwards fucceffively the whole of its ftigmata, with foap-water. At each time I obferved for a confiderable fpace, with the glafs, the ftigmata, thus covered, to see if any bubble of air was formed above; which would naturally have happened, had thefe ftigmata ferved as conduits to the air in refpiration; but with all my attention I could perceive none. Some days afterwards, I repeated the fame experiment in a manner which appeared to me still more decifive. Inftead of covering the ftigmata with foap-water, I covered each with a little bubble of air, taken from the froth of the fame water, that the air might enter and go out more freely. But my curiofity was not the more gratified; these bubbles which ought to have risen or funk at each expiration or infpiration of the Chryfalis preferved conftantly the fame appearance, till, their pellicle becoming dry, they burst.

When the perfect infect had iffued from this chryfalis Į took it up inftantly. I washed the infide of it, and obferved at the ftigmata of its fegments little bundles composed of a great number of very white threads of which the longeft were about two lines in length. Thefe appeared to me the exuviæ of the pulmonary organs. I blowed on each of the ftigmata with all my force, by means of a very slender tube, but my utmoft efforts could neither fwell nor move

any

any of thofe fragments of the veffels, which were attached to them internally; but this muft neceffarily have happened, had the communication of the external air by thefe ftigmata in the bronchiæ remained open, or had the infect, when inclosed in its chryfalis, been able to breathe through them.

If the reader is unwilling to draw a more general conclufion from these laft experiments, it may at leaft, I think, be inferred from them, that the chryfalis of the sphinx liguftri lives for fome time without refpiration, and that its two anterior open ftigmata ferve only to facilitate the evaporation of the fuperabundant humours, and to permit the external air to fupply their place.

PAGE 55, 1. 15,

It is not to be doubted. It is not on the fingle experiment here related, that our knowledge of the refpiration of infects is founded. That experiment does not appear to me fo decifive, as fome may perhaps imagine. Even though an infect should not refpire naturally, it might happen, if its parts are tender, and not apt to ftretch, that, when placed in an exhaufted receiver, it would be killed. It would be fufficient for this purpofe, that the air, which is difperfed in different parts of its body, fhould be there confined, in fuch a manner as to have no convenient iffue. Thus, as foon as the air which furrounds the animal, and preffes on it in all directions, is removed, the air within its body will not fail, in confequence of its natural elasticity, to dilate itself exceedingly, and thus to burft the membranes and veffels which keep it confined: this would certainly occafion the death of the animal, though the want of refpiration had no fhare in it. We have proofs lefs doubtful of the refpiration of infects. The fact appears to me fufficiently demonftrated, with refpect to many fpecies of aquatic infects: I fpeak of thofe which are often seen thrusting the extremity of their abdomen to the surface of the water, and remaining there, as if fufpended. Thefe extremities are with them. the organs of refpiration, and they keep them thus in the air, for the purpose of breathing. To be certain of this, we need only cover the furface of the water where they are, with fomething which may prevent them from thus bringing their tail to the furface. They will immediately be feen greatly agitated, and feeking with extreme inquietude, some

opening

opening through which to push it. If they find no opening, they are seen foon afterwards, to fall to the bottom and die, often in much lefs time than would be neceffary to drown the most delicate terreftrial infect: an evident proof of the refpiration of thefe animals, and that it is abfolutely neceffary to their exiftence. It is proper, however, to inform thofe who would make this experiment, that all aquatic infects, which refpire at the pofterior part of the body, do not die equally foon when deprived of the air. The dytifci can relift this proof for a long time; but their larvæ can fupport it but for a few minutes.

As to lungs, it is eafy to affure ourselves that infects have, them, even without the trouble of diffection ; we need only examine, in the water, the greater part of their exuviæ, where we fall fee floating, great quantities of white veffels, which are connected, by their principal trunks, with what was the orifice of the organs of respiration. These veffels are the fragments of the lungs. Thefe lungs, in infects, divide into fuch an infinite number of bronchiæ, difperfed over their whole body, that all the parts are, as it were, embarraffed with them, and it is often difficult, when we diffect an infect, to lay afide all those filaments, which confound by their number, every thing that appears. After this, would we not be furprised to learn, that thefe pulmonary veffels are not tubes compofed of a single membrane; but veffels always open, composed of a cord or thread, the turns of which are like those of a spring wound round a cylinder, and which thus form hollow cylinders, which open a paffage to the air ? But this is not the only wonderful circumftanee in the mechanifm of thefe lungs. M. de Reaumur found, that the ftrings which form them, have, in fome infects, fix raised fides, fo that they feem to be compofed of fix threads, nearly cylindrical, fastened to one another. Who would have imagined, that fuch small veffels were conftructed with fo much art?

PAGE 55, 1. 24.

It is found at the mouth. It is feldóm at the mouth or at the head, that the refpiratory organs of infects are to be found: nay, perhaps, there is not a fingle instance of their being fituated in that place. As it is on the authority of Frifch that the author advances this fact, I think myself obliged to obferve, that the experiment which induced Frifch to believe, that the aquatic larvæ of the fmaller dra

gon-flies

gon-flies refpire by the under part of the mouth, is by no means decifive. Thefe animals have the mouth, and all the under part of the head covered with a mafk, attached to a kind of arm, which, iffuing from the under fide of the thorax, and making a bend, is folded back, and touches the animal under the chin. It is by means of this arm, if I may call it fo, that the animal puts on, or takes off the mafk from its mouth at pleasure. When this larva is seized between two fingers, even when it is dead, and its abdomen a little preffed, it is furprising to fee it often raise its muzzle, project the angle of the arm from under the thorax, lower its mafk, and, when the preffure is removed, put it on again. Frisch, who, like me, had made the experiment, draws from it a confequence which I cannot approve. He affirms, that this mask, and the arm by which it is attached to the chin, is the organ by which the animal refpires, and the motion made by this arm when the abdomen of the larvæ is preffed, is a proof of this; becaufe, according to him, the motion fhews, that the air is communicated. from the abdomen to the arm by the chin. But I am afraid, he is deceived in the conclufion which he thus draws.” It is my opinion, that these animals, while they continue in the state of larvæ, respire water and not air; and that they infpire the water, not by the mafk, but by their pofterior part, through which alfo they difcharge it. The expiration is more vifible than the infpiration, but it is easy to convince one's felf, that both are performed by the pofterior extremity. Let a filk thread, fuch as is fpun by the filk worm, be taken, and its extremity rolled up between the fingers into a little clew. This clew, when moistened, finks to the bottom, and being fufpended by fo flender a thread, it receives all the motions which the larva communicates to the water. Let it be brought near to the orifice at the posterior part of the animal, and it will be feen, that the clew is alternately repelled and attracted, fo as to fhew distinctly, that its return to the body of the animal, is the effect of an actual attraction, as it is much more quick than it would be, were it the effect, merely of its own weight. Thefe infects, therefore, refpire by their pofterior extremity, and it is water, and not ai which they refpire. Thus, when their abdomen is preffed under water, they are made to lower the mafk: but no bubble of air issues, any more than it does from the other extremity; and we never fee that these a

1

nimals mount to the surface of the water to take in air, which many aquatic infects do which refpire air. Befides, the action of lowering the maík, which they often do when their abdomen is preffed, by no means proves that they refpire by that organ. On the contrary, it fhews, that the fluid, which at that time is very vifibly accumulated in a part of the arm, towards the infide of the angle, has no exit, as the turgefcency lafts as long as the preffure is continued; and this may be reiterated at pleasure, even in a dead animal, which would not happen, did the fluid escape by that way. The mafk has another ufe, much more certain; it is cleft in the fame direction with the mouth, and re-cleft by another fiffure, which fails perpendicularly on the first from the front of the muzzle. Upon the approach of an animal fuited to their tafte, thefe infects instantaneoufly lower the mask, and opening the clefts, they lay hold of their prey, and keep it inclofed and firm between thefe clefts, while they devour it at their leifure. The principal organs, which are to ferve the purposes of refpiration, when they are to breathe air, after they are changed into dragonflies, inftead of water which they breathe at this time, have their origin or openings on the thorax, where they are diftinguished by two ftigmata, and in these terminate all the pulmonary veffels which already contain air, even while the infect ftill lives in the water. It is difficult to fay how the air gets into thefe; for, as was faid before, the animal is never feen coming to the furface of the water to receive it. But that they do contain air is a certain fact, of which it is easy to be fatisfied. We have only to put the water, which contains a few of these animals, over a gentle fire, and when it begins to grow warm, we fhall fee the air within them expand and efcape in bubbles, fometimes with noife, by the two ftiginata of the thorax. Thefe remarks, I think, fhew fufficiently, that dragon-flies, at least those of the fpecies I am'fpeaking of, before their transformation, refpire by the hinder part, and afterwards by the thorax, and that therefore, the organs of their refpiration, are by no means placed near the mouth in them, more than in any other insect that I know.

PAGE 55, 1. 25.

At the extremity of their body, towards the tail. It is not only at the extremity of the abdomen, and at the thorax,

that