INSTINCTS OF BEES. 329 



of honey. This therefore seems to be a case in 

 which reason is taught by experience, and which 

 admits in all its particulars of a direct compa- 

 rison with human reason and human contrivance. 

 Moreover, on the cessation of danger, and when 

 honey-flowers were abundant, the colony pros- 

 perous and swarms prepared to issue, these saga- 

 cious engineers demolished the fortifications, in 

 order to give room for the exit and entrance of 

 the bees. A colony that had been thus attacked 

 in 1804, and was tardy in its defensive prepara- 

 tions, having derived instruction from the past, 

 constructed fresh ramparts speedily, on the re- 

 appearance of the Sphinx in 1 807, and thus guard- 

 ed itself from impending danger. 



From what has been said in page 296, it seems 

 probable that the lives of the working bees do not 

 extend beyond a year, at the utmost : if therefore 

 my inference be legitimate, the information of the 

 colony of 1807 must have been traditional, or 

 else derived from a queen which had reigned over 

 them from 1804. On the subject of traditional 

 information, see Memory of Bees. It is further 

 remarkable, as a confirmation of this process of 

 ratiocination and reflection, that if the apiarian 

 apply proper guards before the entrances to the 

 hives, when the Sphinx makes its appearance, the 

 bees, finding that they are anticipated, devise no 

 measures of security. 



