VIII] OF THE PERIDINIANS 6-15 



are thus in three series, four in touch with six neighbours each, 

 four with four, and four with three; and this arrangement leads 

 to curved contour-lines which may be seen in some of Cohn's figures. 

 Sometimes the interstitial substance is not enough to keep the four 

 cells apart ; then they come together in the usual rhomb or lozenge, 

 and the rest group themselves around in the simplest and most sym- 

 metrical way. That the whole arrangement is as compact as possible 

 under the conditions, that it is in accord with the principle of minimal 

 areas, and is "such as-would result from surface-tension and adhesion 

 between viscous colloidal globules" is now well known to botanists*. 



The tiny plates which form the microscopic shell of a Peridinian 

 illustrate over again by their various collocations the principles 

 which we have been studying in the partition- walls of the segmenting 

 egg; and. if the one case has shewn us pitfalls in the way of the 

 embryologist, the other shews how the systematist, in his endless 

 task of describing the forms and patterij^s of things, may sometimes 

 base distinctions on what seem trivial differences from the physical 

 or mathematical point of view. 



On the upper half (or epitheca) of the globular test of a Peridinium 

 we have fourteen Httle plates, or fourteen "cells," to use the word 

 in a mathematical rather than a histological sense, whose boundary- 

 walls always meet in three-way nodes. We may reproduce the 

 identical arrangement of the Peridinial plates by blowing bubbles 

 in a saucer; but to deal with so many bubbles at once needs more 

 patience than do the other similar experiments which we have 

 described. That the cells are fourteen in number is, from the 

 physical point of view, the merest accident, but from the zoologist's 

 it is a criterion of the genus; when there are more cells or fewer, 

 the organism is called by another name. The number of possible 

 arrangements of fourteen polygonal cells, linked by three-way nodes, 

 is very large; but the "characters of the genus" exclude many of 

 the variants. Many of them occur — it is quite possible that all 

 occur — in Nature; but they are not called Peridinium. The fol- 

 lowing arrangement defines the genus. There is a central or apical 

 cell, around which are grouped six others ; of these six, one extends 

 to the boundary of the figure, that is, to the equator of the globular 



♦ R. A. Harper, The colony in Gonium, Trans. Anier. Microsc. Soc. xxxi, pp. 65- 

 84, 1912; cf. F. Cohn, in Nova Acta Acad. C.L.C., xxiv, p. lOJ, 1854. 



