538 



THE FORMS OF TISSUES 



[CH. 



nor the cells strictly horizontal. The base is always thicker than 

 the side- walls; ,its solid angles are by no means sharp, but filled up 

 with curving surfaces of wax, after the fashion, but more coarsely, 

 of Plateau's bourrelet. Hence the Maraldi angle is seldom or never 

 attained; the mean value (according to Vogt) is no more than 

 106-7° for the workers, and 107-3° for the drones. The hexagonal 

 angles of the prism are fairly constant; about 4° is the limit of 

 departure, and about 1-8° the mean error, on either side. 



r^-t^ 



Fig. 208. Brood-comb, with eggs. 



The bee makes no economies; and whatever economies lie in the 

 theoretical construction, the bee's handiwork is not fine nor accurate 

 enough to take advantage of them*. 



The cells vary little in size, so little that Thevenot, a friend of 

 Swammerdam's, suggested using their dimensions as a modulus 

 or standard of length; but after all, the constancy is not so great 

 as has been supposed. Swammerdam gives measurements which 

 work out at 5-15 mm. for the mean diameter of the worker-cells, 

 and 7 mm. for those of the drones ; Jeffries Wyman found mean 

 values for the worker-cells from 5-1 to 5-2 mm.; Vogt, after many 

 careful measurements, found a mean of 5-37 mm. for the worker- 



* All this Heinrich Vogt has abundantly shewn, in part by making casts of the 

 interior of the cells, as Castellan had done a hundred years before. See his 

 admirable paper on the Geometric und Oekonomie der Bienenzelle, in Festschrift 

 d. Universitdt Breslau, 1911, pp. 27-274. 



