FOOD OF INSECTS, 389 



however, that when some of these have fed for a time 

 on one plant they will die ratlier than eat another, 

 which would have been perfectly acceptable to tliem if 

 acciistonied to it from the first''. Here too it must be 

 borne in mind, that by far the greater part of insects 

 feed upon different substances in their different states 

 of existence, eating one kind of food in the larva and 

 another in the imago state. This is the case with the 

 whole order L,epidoptcra, which in the former eat plants 

 chiefly, in the latter nothing but honey or the sweet 

 juices of fruit, which they have often been observed to 

 imbibe ; and the same rule obtains also in regard to 

 most dipterous and hymenopterous insects. Those 

 which eat one kind of food in both states, are chiefly 

 of the remaining orders. 



I have said that insects, like other animals, draw 

 their subsistence from the vegetable or animal king- 

 doms. But I ought not to omit noticing that some 

 authors have conceived that several species feed upon 

 mineral substances. Not to dwell upon Barchewitz's 

 idle tale of East Indian ants v*'hich eat iron% or on the 

 stone-eating caterpillars recorded in the Memoirs of 

 the French Academy*^, which are now known to erode 

 the walls on which they are found, solely for the pur- 

 pose of forming their cocoons ; Reaumur and Swara- 

 merdam have both stated the food of the larvifi of 

 Ephemerce to be earth, that being the only substance 

 ever found in their stomachs and intestines which are 

 filled with it. This supposition, which if correct ren- 

 ders invalid the definition by which Mirbel (and my 

 friend Dr. Alderson long before him) proposed to di« 



* Reaum. ii. 324. ^ J^esser, L. i. 259. ^ x. 45S, 



