214 ON VARIABILITY AND ADAPTATION 



numerous atoms. Now Lord Kelvin, in his convincing investi- 

 gation into the size of the molecule of water (by a study of the 

 thinnest possible films of a bubble), has proved that in a line 

 0-5 fi in length there could be only about 150 molecules of water. 

 Let us be generous and compute the ids as being not on the average 

 0-5 but I'O fi in diameter, and, again being generous, let us com- 

 pute the ultimate molecule of living matter as being only thirty 



- v 



Fio. 18. 



A. Nucleus undergoing mitosis, the "skein" of chromatin broken up into V-shaped loops 



or chromosomes. 



B. Nucleus undergoing mitosis, in earlier skein stage, in which (as occasionally seen) 



the skein is seen to resolve itself into a double row of chromomeres (Weismann's 

 "ids") prior to breaking up into loops. 



times the size of the molecule of water from every considera- 

 tion an absurd underestimate. It will be seen upon calculation 

 that the id (supposing it to be spherical) can contain only about 

 as many molecules as presumably Weismann requires for one or 

 two biophores, or at the most economical rate for a single repre- 

 sentative determinant ; and not one, or two, or three, but several 

 hundreds of determinants ought to be compressed into a 

 respectable id. 



We have, in short, the reductio ad absurdum of Weismann's 

 theory. 



ADDENDUM 



[In the second edition of my Principles of Pathology, published 

 in 1910, with the publication of further studies by the physicists 

 upon the size of the molecule of water, I modified the treatment 

 of the subject as follows : 1 



" We employed previously Lord Kelvin's estimate of the size 

 of a molecule of water, pointing out that, according to his figures, 

 in the chromomeres, or bead-like granules seen in certain chromo- 

 somes, which have been taken to represent Weismann's ' ids,' 

 there could be stretched across the diameter only about 150 

 1 P. 136. 



