FOOD OF INSECTS. 387 



yefusing even tobacco ; and from the impossibility that 

 one of a million of the innumerable swarms of gnats 

 which abound in swampy places, particularly in re- 

 gions which but for them would be lost to sensitive ex- 

 istence, should ever taste blood, it seems clear that they 

 are usually contented with vegetable aliment. Indeed 

 the males, as well as those of Tabanus of which even 

 the females readily imbibed the sugared fluid offered to 

 them by Reaumur % never suck blood at all; so that 

 they must either feed on vegetable matter, which in 

 fact I have observed them to do, or fast during their 

 whole existence in the perfect state. 



Though insects, generally considered, have thus a 

 much more extensive bill of fare than the larger ani- 

 mals, each individual species is commonly limited to a 

 more restricted diet. Many both of animal and vege- 

 table feeders are absolutely confined to one kind Of 

 food, and cannot exist upon any other. The larva of 

 CEstrus Equi can subsist no where but in the stomach 

 of the horse or ass, which animals therefore this insect 

 might boast with some show of reason to have been cre- 

 ated for its use i-ather than for ours, being to us useful 

 only, but to it indispensable. The larvae of Si/rphus 

 Pyrastri (Miisca, L.) according to De Geer eat no 

 other Aphis but that of the rose''. Most Ichneumons 

 and Spheges prey each upon a single species of insect 

 only, which therefore they would seem to have been 

 formed for the express purpose of keeping within due 

 limits. Reaumur mentions having once found in a 

 parcel of decaying wood the nests of six different kinds 

 of Sphex, each of which was filled with flies of a di- 



" Reaum, iv. 2S0. " De Geer, vi. 112. 



2 C 2 



