206 MENTAL EVOLUTION IN ANIMALS. 



a transition takes place from one class of cell to the other, a 

 complex geometrical problem arises how to imite hexagonal 

 cells of a small with others of a large diameter, without 

 leaving any void spaces or interfering with the regularity of 

 the comb. Without occupying space with what would 

 necessarily be a rather lengthy exposition of the manner in 

 which the bees solve the problem, it is enough to say that 

 in passing from one form of cell to the other, they require to 

 construct a great many rows of intermediate cells which 

 differ in form, not only from the ordinary cells, but from each 

 other. When the bees arrive at any stage in this process of 

 transition, they might stop at that stage and continue to build 

 the whole of their comb upon this pattern. But they inva- 

 riably proceed from one stage to another until the transition 

 from small hexao'ons to larc^e hexao'ons, or vice versd, is 

 effected. On this subject Kirby and Spence remark : 

 ''' Eeaumer, Bonnet, and other naturalists cite these irregu- 

 larities as so many examples of imperfections. What would 

 have been their astonishment if they had been aware that 

 part of these anomalies had been calculated (? adaptive) ; 

 that there exists as it were a moveable harmony in the 

 mechanism by which the cells are composed ! ... It is 

 far more astonishing that they know how to quit their 

 ordinary routine when circumstances require that they should 

 build male cells : that they should be instructed to vary the 

 dimensions and the shape of each piece so as to return to a 

 regular order ; and that, after having constructed thirty or 

 forty ranges of male cells, they again leave the regular order 

 in which they were formed, and arrive by successive diminu- 

 tions at the point from which they set out .... Here 

 again, as observed in a former instance, the wonder would be 

 less if everj/ comb contained a certain number of transition 

 and of male cells, constantly situated in one and the same 

 part of it ; but this is far from being the case. The event 

 which alone, at whatever period it may happen, seems to 

 determine the bees to construct male cells, is the oviposition 

 of the queen. So long as she continues to lay the eggs of 

 workers, not a male cell is provided ; but as soon as she is 

 about to lay male eggs, the workers seem aware of it, and 

 you then see them form their cells irregularly." 



Here, then, w^e have concerted variation in the mode of 

 constructing the cells of a normal and definite kind, and we 



