334 THE GRAMMAR OF SCIENCE 



then nearly all the statements we have made with regard 

 to physics will serve as canons for determining the validity 

 of biological ideas. In particular, any biological concept 

 will be scientifically valid if it enables us to briefly 

 summarise without internal contradiction any range of our 

 perceptual experience. But the moment the biologist 

 goes a step further, and asserts on the ground of the 

 validity of his concept that it is a reality of the pheno- 

 menal world, although no perceptual equivalent has yet 

 been found for it, then he at once passes from the solid 

 ground of science to the quicksands of metaphysics. He 

 takes his stand with the physicist who asserts the pheno- 

 menal existence of the concepts atom and molecule. 



S 3. — MccJianism and MetapJiysics in Tlieorics of Het'edity 



I cannot bring home to the reader the difficulties with 

 which the projection of conceptions into the phenomenal 

 world is attended better than by briefly referring to two 

 well-known biological theories of heredity. Of the change 

 in those groups of sense-impressions which the biologist 

 sets himself to describe there are two prominent features 

 which at first sight might seem to correspond to nomic 

 and anomic changes (p. 95, footnote), to routine and to 

 breaches of routine. These features are the recurrence in 

 our experience of the offspring of sense-impressions 

 associated with the parental organism, and the occurrence 

 in our experience of the offspring of sense-impressions not 

 associated with the parental organism. These features 

 are termed inheritance and variation. The apparent 

 anomy, involved in variation, is very probably like the 

 anomy of the weather, a result of our not yet having 

 formed a sufficiently wide or fundamental classification of 

 facts. Be this as it may, inheritance and variation form 

 the basis upon which biologists construct the evolution of 

 life. Theories which endeavour to resume inheritance and 

 variation under a single and simple formula are termed 

 theories of heredity, and two of the most important of these 

 theories are due respectively to Darwin and Weismann. 



