616 



THE FORMS OF TISSUES 



[CH. 



shell, and seven other "equatorial cells" complete the boundary. 

 In other words, the apical hexagon is surrounded by two concentric 

 rows, originally of six cells each, whereof one cell of the inner row 

 has (as it were) burst through a cell of the outer row, so reaching 

 the boundary itself, and so dividing into two the equatorial cell 

 which it encroached on and bisected. 



In any such collocation as this, the number of sides and the 

 number of nodes or corners are strictly determined ; there are here 

 fourteen cells, all conjoined by three-way nodes, and it follows that 

 there are just 39 separate walls or edges, and just 26 nodes or 

 corners. Many of these last are already defined for us; for six of 

 them are the corners of the central hexagon, eight lie on the equator, 

 and six more are at the inner ends of the radial partitions which 

 separate the equatorial cells from one another. Six remain to be 

 determined, those, namely, where the 

 partitions running outwards from the 

 apical cell meet the walls of the 

 equatorial cells. The diagram (Fig. 

 265) shews us two sets of radiating 

 partitions, six running inwards from 

 the equator (a, 6, c, d, e, /) and six 

 running outwards from the central 

 hexagon (A. . .F), those of the one 

 set being nearly opposite to those of 

 the other; but near as they may be 

 they never meet, for to do so would 

 be to make a four-way node, which 

 theory forbids and which observation tells us does not occur. In 

 every case, one partition must be slewed a little to one side or 

 other of its opposite neighbour; and the whole range of possible 

 variations depends on whether the shift be to the one side or to the 

 other. We have six pairs of partitions, and in each of the six there 

 is this possible alternative of right or left ; there are therefore 2® or 

 64 possible variations in all. Whether all six may vary, I do not 

 know; there is no obvious reason why they should not. But 

 alternative variation does occur in the two anterior and two posterior 

 pairs of partitions; and these four give us 2^ or 16 possible arrange- 

 ments. 



Fig. 265. Dorsal view of a 

 Peridinium : diagrammatic. 



