708 ON CONCRETIONS, SPICULES, ETC. [ch. 



by a layer of froth-like vesicles, uniform in size or nearly so. We 

 know that their mutual tensions will tend to conform them into the 

 fashion of a honeycomb, or regular meshwork of hexagons, and 

 that the free end of each hexagonal prism will be a little spherical 

 cap. Suppose now that it be at the outer surface of the protoplasm 

 (in contact with the. surrounding sea- water) that the sihceous 

 particles have a tendency to be secreted or adsorbed; the distribu- 

 tion of surface-energy will lead them to accumulate in the grooves 

 which separate the vesicles, and the result will be the development 



Fig. 321. '''' Reticulum plasmatique J" 

 After Carnoy. 



Fig. 322. Aulonia hexagona Hkl. 



of a dehcate sphere composed of tiny rods arranged, or apparently 

 arranged, in a hexagonal network after the fashion of Carnoy's 

 reticulum jplasmatique, only more solid, and still more neat and 

 regular. Just such a spherical basket, looking like the finest 

 imaginable Chinese ivory ball, is found in the siliceous skeleton of 

 Aulonia, another of Haeckel's Radiolaria from the Challenger. 



But here a strange thing comes to light. No system of hexagons 

 can enclose space; whether the hexagons be equal or unequal, 

 regular or irregular, it is still under all circumstances mathematically 

 impossible. So we learn from Euler: the array of hexagons may 

 be extended as far as you please, and over a surface either plane or 

 curved, but it never closes in. Neither our reticulum plasmatique 



