ON THE STATISTICAL VIEW OF NATURE. 613 



whole, formed by the joint agencies of a host of what 

 appear to us to be unconscious or barely conscious 

 elements. . . . The doctrine of Pangenesis gives excellent 

 materials for mathematical formulas, the constants of 

 which might be supplied through averages of facts." : 

 Mr Galton does " not see any serious difficulty in the 

 way of mathematicians in framing a compact formula, 

 based on the theory of Pangenesis, to express the com- 

 position of organic beings in terms of their inherited 

 and individual peculiarities, and to give us, after certain 

 constants had been determined, the means of foretell- 

 ing the average distribution of characteristics among a 

 large multitude of offspring whose parentage was 

 known. 2 ... In short, the theory of Pangenesis brings 

 all the influences that bear on heredity into a form that 

 is appropriate for the grasp of mathematical analysis." 



Evidently in the mind of Mr Galton the problem of 40. 



Problem of 



heredity divides itself into two distinct problems ; and Heredity. 

 he has himself laboured at the solution of both. We 

 may call the one the " historical " or the " mechanical " 

 problem, the other the " statistical " problem, following 

 the distinction which Maxwell drew when dealing with 

 the kinetics of gases. The historical problem would 

 involve a more detailed account of the nature of those 

 organic units which the theory of Pangenesis, in common 

 with other similar theories, like those of Buffon and 

 Nageli, assumes, and of the mechanism by which they 

 unite and are transmitted. If this is impossible, or at 

 all events highly hypothetical, the actual following up 

 by observation and experiment of the phenomena of 



1 'Hereditary Genius' (1892), p. 356. 2 Ibid., p. 358. 



