Bramble-bees and Others 



each sex? Again, not so; for in the first 

 series, where the females are housed, instead 

 of the interval of 1 1 millimetres, which oc- 

 curs at the beginning and the end, we find, 

 in the middle of the series, an interval of i6 

 millimetres, while, in the second series, re- 

 served for the males, instead of the interval 

 of 7 millimetres at the beginning and the 

 end, we have an interval of 5 millimetres in 

 the middle. It is the same with the other 

 series, each of which shows a striking dis- 

 crepancy in its figures. If the Osmia really 

 studied the dimensions of her chambers and 

 measured them with the compasses of her 

 body, how could she, with her delicate 

 mechanism, fail to notice mistakes of 5 milli- 

 metres, almost half her own length? 



Besides, all idea of geometry vanishes if 

 we consider the work in a tube of moderate 

 width. Here, the Osmia does not fix the 

 front partition in advance; she does not even 

 lay its foundation. Without any boundary- 

 pad, with no guiding mark for the capacity 

 of the cell, she busies herself straightway with 

 the provisioning. When the heap of Bee- 

 bread is judged sufficient, that is, I imagine, 

 when her tired body tells her that she has 



82 



