IMPERFECT SOCIETIES OF IXSECTS. 13 



This is clearly one of the modes by which their num- 

 bers are kept within due limits, as, doubtless, the great 

 majority of these adventurers perish in the waters. 

 Thus, also, a great supply of food is furnished to those 

 fish in the sea itself, which at other seasons ascend the 

 rivers in search of them ; and this probably is one of 

 the means, if not the only one, to which tlie numerous 

 islands of this globe are indebted for their insect po- 

 pulation. Whether tlje insects I observed upon the 

 beach Avetted by the waves, had flown from our own 

 shores, and falling into the water had been brought 

 back by the tide ; or whether they had succeeded in 

 the attempt to pass from the continent to us, by flying 

 as far as they could, and then falling had been brought 

 by the waves, cannot certainly be ascertained ; but 

 Kalm's observation inclines me to tlie latter opinion. 



The next order of imperfect associations is that of 

 those insects which feed together : — these are of two de- 

 scriptions — those that associate in tlieir ^r.?^ or last state 

 only, and those that associate in all their states. The 

 first of these associations is often very short-lived : a 

 patch of eggs is glued to a leaf; when hatched, the 

 little larvae feed side by side very amicably, and a plea- 

 sant sight it is to see the regularity v»'itli which this 

 work is often done, as if by word of coiiimand ; but 

 when the leaf that served for their cradle is consumed, 

 their society is dissolved, and each goes where he can 

 to seek his own fortune, regardless of the fate or lot of 

 his brethren. Of this kind are the larvcB of the saw- 

 fly of the gooseberry, whose ravages I have recorded 

 , before % and that of the cabbage-butterfly; the latterj 



' Vol. 1.2a Ed. 197. 



