INSTIKCT OF INSECTS. 523 



trial that they could not reach the nectar from the top; 

 but that having- once ascertained by experience that 

 the flowers of beans are too strait to admit them, they 

 then, without further attempts in the ordinary way, 

 pierced the bottoms of all the flowers A^hich they 

 wished to rifle of their sweets. — M. Aubert du Petit- 

 Thouars observed that humble-bees and Xylocopa 

 xiolacea gained access in a similar manner to the 

 nectar oi^ Antirrhinum Linaria and majus^ and Mirabi- 

 lisJalappa', as do the common bees of the Isle of France 

 to that o^Canna indica ^; and I have myself more than 

 once noticed holes at the base of the long nectaries 

 of Aquilegia xulgaris, which I attribute to the same 

 agency. 



My second fact is supplied by the same ants, whose 

 sagacious choice of the vicinity of Reaumur's glass 

 hives for their colony has been just related to you. 

 He tells us that of these ants, of which there were such 

 swarms on the outside of the hive, not a single one was 

 ever perceived within ; and infers that, as they are 

 such lovers of honey, and there was no difficulty in 

 finding crevices to enter in at, they were kept without, 

 solely from fear of the consequences ''. Whence arose 

 this fear ? We have no ground for supposing ants en- 

 dowed with any instinctive dread of bees ; and Reau- 

 mur tells us, that when he happened to leave in his 

 garden, hives of which the bees had died, the ants then 

 never failed to enter them and regale themselves with 

 the honey. It seems reasonable, therefore, to attri- 

 bute it to experience. Some of the ants no doubt had 

 tried to enter the peopled as they did the empty hive, 



' JS'uuvcau Utilklin dcs Science', i. 4j. '' Rcaiiiu, v. 109. 



