VII 



OF HEXAGONAL SYMMETRY 



517 



thousand cells in such a plate has been found, Uke the Cycas-petiole, 

 to average out at six sides each, precisely* 



The cracking of a fine varnish may illustrate (as in Fig. 19.7) the 

 same phenomenon. A little water in the varnish tends to accumu- 

 late between the cells, interferes with their close-packing, and 

 complicates the arrangement of their partitions. 



The horny plates which form the carapace of a tortoise (different 

 as the case may seem) still obey the two guiding principles (1) that 

 the polygonal boundaries meet in three-way nodes, and (2) that the 

 three angles tend towards equality, always provided that no alien 

 influences interfere. These principles are of the widest apphcation; 

 the carapace of a Eurypterid, the dermal armour of an Old Red 



Fig. 197. Cellular patterns in varnish, a, dissolv'ed in dry acetone; 

 b, containing a little water. 



Sandstone fish like Hugh Miller's Asterolepis, exhibit them at a 

 glancef. The carapace of our tortoise is formed of a bony framework 

 of ribs and vertebrae, overlaid by superficial plates of horn or 

 tortoiseshell ; it is these latter with which we are about to deal. 

 They are arranged in three rows down the back, and a marginal row 

 of smaller plates surrounds the others ; there are (normally) twenty- 

 four plates in the marginal series, and five large ones in the median 

 longitudinal row. With these few^ facts, and our general principles 



* F. T. Lewis, Polygons in a film... and the pattern of simple epithelium, 

 Anat. Record, l, pp. 235-265, 1931. 



t The all but universal law of the triple corner, or triradiate suture, is now 

 and then enough to give a deceptive likeness to very different things. When 

 Cope marred his brilliant classification of the Ostracoderm fishes by seeing in the 

 carapace of Bothriolepis a likeness to the dorsal plates of the tunicate Chelyosoma, 

 it was this and this alone which led him astray. 



