316 ON THE INTERNAL FORM [ch. 



the mathematical expression of a physical equiUbrium — and is the 

 not too remote counterpart of the arrangement of the electrons in 

 an atom. Be that as it may, it is found that a group of three, 

 four or five Httle magnets arrange themselves at the corners of an 

 equilateral triangle, square or pentagon; but a sixth passes within 

 the ring, and comes to rest in the centre of symmetry of the 

 pentagon. If there be seven magnets, six form the ring, and the 

 seventh occupies the centre ; if there be ten, there is a ring of eight 

 and two within it ; and so on, as follows * : 



When we choose from the published figures cases where the 

 chromosomes are as nearly as possible alike in size and form — the 

 condition necessary for our parallel to hold — then, as LilHe pre- 

 dicted and as Doncaster and Graham Cannon have shewn, their 

 congruent arrangement agrees, even to a surprising degree, with 

 what we are led to expect by theory and analogy (Fig. 96). 



The break-up of the nucleus, already referred to and ascribed 

 to a diminution of its surface-tension, is accompanied by certain 

 diffusion phenomena which are sometimes visible to the eye; and 

 we are reminded of Lord Kelvin's view that diffusion is implicitly 

 associated with surface-tension changes, of which the first step is 

 a minute puckering of the surface-skin, a sort of interdigitation with 

 the surrounding medium. For instance, Schewiakoff has observed 

 in Euglyphaf that, just before the break-up of the nucleus, a system 

 of rays appears, concentred about it, but having nothing to do with 

 the polar asters: and during the existence of this striation the 

 nucleus enlarges very considerably, evidently by imbibition of fluid 

 from the surrounding protoplasm. In short, diffusion is at work, 

 hand in hand with, and as it were in opposition to, the surface- 

 tensions which define the nucleus. By diffusion, hand in hand with 

 surface-tension, the alveoli of the nuclear meshwork are formed, 

 enlarged and finally ruptured: diffusion sets up the movements 



* H. Graham Cannon, On the nature of the centrosomal force, Journ. Genetics, 

 XIII, p. 55, 1923. 



t Schewiakoff, Ueber die karyokinetische Kerntheilung der Euglypha alveolata, 

 Morph. Jahrb. xiii, pp. 193-258, 1888 (see p. 216). 



