INSTINCT OF INSECTS. 481 



might happen, and appeared to profit by the warning 

 given by the fall of one of the combs to consolidate the 

 others and prevent a second accident of the same nature. 

 These last had not been displaced, and appeared solidly 

 attached by their base ; whence Huber was not a little 

 surprised to see the bees strengthen their principal points 

 of connexion by making them much thicker than before 

 with old wax, and forming numerous ties and braces to 

 unite them more closely to each other and to the walls 

 of their habitation. — What was still more extraordinary, 

 all this happened in the middle of January, at a period 

 when the bees ordinarily cluster at the top of the hive, 

 and do not engage in labours of this kind *. 



You will admit, I think, that these proofs of the re- 

 sources of the architectural instinct of bees are truly ad- 

 mirable. If, in the case of the substitution of mit^'s for 

 the first range of waxen cells, this procedure invariably 

 took place in every bee-hive at a jf^r^f? period — when, for 

 example, the combs are two-thirds filled with honey — it 

 would be less surprising: but there is nothing of this in- 

 variable character about it. It does not, as Huber ex- 

 pressly informs us^, occur at any marked and regular 

 period, but appears to depend on several circumstances 

 not always combined. Sometimes the bees content 

 themselves with bordering the sides of the upper cells 

 with propolis alone, without altering their form or giving 

 them greater thickness. And it is not less remarkable 

 that, from the instances last cited, it appears that they 

 are not confined to one kind of cement for strengthen- 

 ing and supporting their combs, but avail themselves of 



. ' Huber, ii. 280. " Ibid. ii. 284, note*, 



VOL. II. 2 T 



