502 THE FORMS OF TISSUES [ch. 



hexagonal "cells," which simulate an organic parenchyma very 

 closely; and by varying the experiment in ways which Leduc 

 describes, we may imitate various forms of tissue, and produce cells 

 with thick walls or with thin, cells in close contact or with wide 

 intercellular spaces, cells with plane or with curved partitions, and 

 so forth. 



James Thomson (Kelvin's elder brother) had observed nearly 

 sixty years ago a curious "tesselated structure" on a Hquid surface, 

 to wit, the soapy water of a wash-tub. The eddies and streaks of 

 swirling water settled down into a cellular configuration, which 

 continued for hours together to alter its details; small areoles 

 disappeared, large ones grew larger, and subdivided into small ones 

 again. With few and transitory exceptions three partitions and no 

 m^ re met at every node of the mesh work ; and (as it seems to me) 

 the subsequent changes were all due to such shifting of the lines as 

 tended to make the three adjacent angles more and more nearly 

 co-equal with one another : the obvious effect of this being to make 

 the pattern more and more regularly hexagonal*. 



In a not less homely . experiment, hot water is poured into a 

 shallow tin and a layer of milk run in below ; on blowing gently to 

 cool the water, holes, more or less close-packed and evenly inter- 

 spaced, appear in the milk. They shew how cooling has taken place, 

 so to speak, in spots, and the cooled water has descended in isolated 

 columns f. 



Benard's "tourbillons cellulaires"J, set up in a thin hquid layer, 



* James Thomson, On a changing tesselated structure in certain liquids, Proc. 

 Glasgow Phil. Soc. 1881-82; Coll. Papers, p. 136— a paper with which M. Benard 

 was not acquainted, but see Benard's later note in Ann. <Ie Chim. Dec. 1911. 



t See Graham's paper, quoted below. 



% H. Benard, Les tourbillons cellulaires dans une nappe hquide. Rev. gener. des 

 Sciences, xii, pp. 1261-1271, 1309-1328, 1900; Ann. Chimie et Physique (7), xxm, 

 pp. 62-144, 1901 ; ibid. 1911. Quincke had seen much the same long before: An7i. d. 

 Phys. cxxxix, p. 28, 1870. The "figures of de Heen" are an analogous electrical 

 phenomenon; cf. P. de Heen, Les tourbillons et les projections de I'ether, Bull. 

 Acad, de Bruxelles (3) xxxvii, p. 589, 1899; A. Lafay, Ann. de Physique (10), xiii, 

 pp. 349-394, 19.30. These various phenomena, all leading to a pattern of hexagons, 

 have often been studied mathematically: cf. Rayleigh, Phil. Mag. xxxii, pp. 529- 

 546, 1916, Coll. Papers, vi, p. 48; also Ann Pellew and R. V. Southwell, Proc. R.S. (A), 

 CLXXVi, pp. 312-343, 1940. The hexagonal pattern is a particular case of stability, 

 but not necessarily the simplest; it is only by experiment that we know it to be 

 the permanent condition in an unlimited field. 



