Exceptions to the Law of Heredity. 207 



either the effect is produced by mechanical laws or by chemical 

 laws. In the case of mechanical laws each cause is found in the 

 complex effect, precisely in the same way as though it alone had 

 acted : the effect of concurrent causes is exactly the sum of the 

 separate effects of eac"h. On the other hand, the chemical com- 

 bination of two substances produces a third, the properties of which 

 are entirely different from each of the other two, whether taken 

 separately or together : thus, a knowledge of the properties of 

 sulphur and oxygen does not imply a knowledge of the proper- 

 ties of sulphuric acid. 1 But psychological laws are analogous, 

 now to mechanical, now to chemical laws. It is even prob- 

 able that the greater number of them are chemical. Hence 

 it is impossible to proceed by deduction from causes to effects. 

 Here experience alone can guide us. It is curious to notice that 

 prior to the discoveries of modern chemistry the idea of a total 

 dissimilarity between causes and effects, and, what is still more 

 striking, between the composite and its component parts, seems to 

 have been unknown to science, except perhaps the dreams of 

 alchemists about the transmutation of metals. It would surely 

 have been a surprise for the scientific men of that epoch had they 

 been told, Here is oxygen, a gas without colour or odour, com- 

 bustible, and the active agent of all combustion ; and here is 

 hydrogen, another and a very different gas. Combine the two in 

 definite proportions, and you will get a liquid which may be either 

 the water you drink, or the mist on which is painted the rainbow. 

 The chemistry of life, by showing us how inorganic matter is trans- 

 formed into the plant, the plant into the animal ; how in the 

 animal the organic matter returns by death to the inorganic world 

 to recommence its course, has revealed to us metamorphoses far 

 more astounding than those whose explication we seek. 



We may, then, regard it as certain that in the domain of life 

 (including thought) a disproportion often exists between cause 

 and effect which cannot be foreseen by reasoning, which is given 

 us only by experience, and that it is a wholly gratuitous assertion 

 to say, There is too much difference between such a fact and 

 such another between the simplicity of the one and the com- 



1 Mill's Logic, book vi., iv., and book iii., vi 

 10 



