VII] OF CELL-PARTITIONS 477 



mesh work of "cells," these lesser veins being all much of a muchness, 

 and exerting tensions insignificant compared with those of the 

 greater veins. Where (a) two ribs run so near together that only 

 one row of cells lies betw^een, these cells are quadrangular in form, 

 their thin partitions meeting the ribs at right angles on either side. 

 Where (6) two rows of cells are intercalated between a pair of ribs, 

 one row fits into the other by angles of 120°, the result of co-equal 

 tensions; but both meet the ribs at right angles, as in the former 

 case. W^here (c) the cell-rows are numerous, all their angles in 

 common tend to be co-equal angles of 120°, and the cells resolve, 

 consequently^ into a hexagonal meshw^ork. 



Many spherical cells, such as Protococcus, divide into two equal 

 halves, separated by a plane partition. Among other lower Algae 

 akin to Protococcus, such as the Nostocs 



and Oscillatoriae, in which the cells are a OCOOOTyCOD 

 embedded in a gelatinous matrix, we 

 find a series of forms such as are re- 

 presented in Fig. 163, which various OOCjOOOO 



conditions depend, according to what r^rv" — ^z- v^ I 



we have already learned, upon the ^^--^ AJ 



relative magnitudes of the tensions at 



the surface of the cells and the boundary q fTTTHnnnrTT'T^ 



between them. In some cases (Fig. 



163, B) the cells remain spherical, 



because they are merely embedded in D 1 1 I 1 I I I I I I I I 



the matrix, with no other physical Fig. 163. Filaments, or chains of 



continuity between them; even two cells, in various lower Algae. 



soap-bubbles do not tend to unite, \f:l J"T-' ^fln'^nT'"' 



^ '(C) Rivulana; (D) Oscillatoria. 



unless their surfaces be moist or we 



put a drop of soap-solution between them. In certain other cases, 

 the system consists of a relatively thick- walled tube, subdivided by 

 more dehcate partitions, which latter then tend (as in D) to become 

 plane septa, set at right angles to the walls. Or again, side- walls 

 and septa may be all alike, or nearly so ; and then the configuration 

 (as in C, on Fig. 163) is that of a linear cluster of soap-bubbles*. 

 In the spores of liverworts, such as Pellia, the first partition 



* Cf. Dewar, Studies on liquid films, Proc. Roy. Inst. 1918, p. 359. 



