HABITATIONS OF INSECTS. 499 



it downwards ; those on the sides widen it to right 

 and left ; and those which are employed above the 

 thickest part extend its dimensions upwards. The 

 more a comb is enlarged below, the more it is neces- 

 sary that it should be enlarged upwards to the top of 

 the hive. The bees that are engaged in lengthening 

 t\ie comb, work v^ith more celerity than those which 

 increase its width ; and those that ascend or increase 

 its width upwards, more slowly than the rest. Hence 

 it arises that it is longer than wide, and narrower to- 

 wards the top than to\vards the middle. — The first 

 formed colls are usually not so deep as those in the 

 middle ; but Avlien the comb is of a certain height, they 

 are in haste to lengthen these cells so essential to the 

 solidity of the whole, sometimes even making them 

 longer than the rest. — The cells are not perfectly ho- 

 rizontal ; they are almost always a little higher to- 

 wards their mouth than at their base, so that their axis 

 is not perpendicular to the partition that separates the 

 two assemblages. They sometimes vary from the hori- 

 zontal line more than 20°, usually four or five. When 

 the bees enlarge the diameter of the cells preparatory 

 to the formation of male cells, the bottoms often con- 

 sist of two rhomboids and two hexagons, the size and 

 form of which vary, and they correspond with four in- 

 stead of three opposite cells. — The works of bees are 

 symmetrical less perhaps in minute details than consi- 

 dered as a whole. Sometimes, indeed, their combs 

 have a fantastic form ; but this, if traced, will be found 

 to be caused by circumstances : one irregularity occa- 

 sions another, and both usually have their origin in 

 the dispositions vvhich we make them adopt. The in- 



2 K <2 



