DEATH'S HEAD SPHINX. 39 



venient, and they are removed, until new alarms 

 demand their reconstruction. 



" The entrances formed in 1804, were destroyed in 

 spring 1805. The Sphinx did not appear that year, 

 nor was it seen in the year following ; but it returned 

 in great numbers in autumn 1807. By speedily 

 barricading themselves, the bees prevented their 

 threatened ravages ; but before the departure of 

 swarms, in May 1808, they demolished the fortifica- 

 tions, whose narrow passages prohibited free egress 

 to the multitude. 



* When the entrance to the hive is itself restricted, 

 on care being taken to contract it soon enough to 

 prevent the devastation of their enemies, bees dispense 

 with walling themselves in. We cannot explain 

 their conduct otherwise, than by admitting the evolu- 

 tion of their instinct according as excited by circum-* 

 stances. 



" But how can the Sphinx alarm colonies so 

 warlike ? How can a moth, the dread of supersti- 

 tious people, also exercise a secret influence over 

 insects, and have the faculty of paralyzing their 

 courage ? Does it emit any emanation pernicious to 

 bees? 



" Other species of sphinges subsist on the nectar of 

 flowers alone ; they have a long, slender, flexible, 

 spiral trunk, and seek their food at sunset. But the 

 Atropos is later on the wing ; nor does it hover 

 about the hives until night is farther advanced. It is 

 provided with a thick, short proboscis ; is endowed 

 with great strength ; and when seized, some unknown 



