Are there Laws of Heredity? 141 



way in which they are commonly proved is in the highest degree 

 illogical ; the usual course being for writers to collect instances 

 of some mental peculiarity found in a parent and in his child, 

 and then to infer that the peculiarity was bequeathed. 



By this mode of reasoning we might demonstrate any pro- 

 position ; since in all large fields of inquiry there are a sufficient 

 number of empirical coincidences to make a plausible case in 

 favour of whatever view a man chooses to advocate. But this is 

 not the way in which truth is discovered ; and we ought to inquire 

 not only how many instances there are of hereditary talents, etc., 

 but how many instances there are of such qualities not being 

 hereditary. 



Until something of this sort is attempted, we can know nothing 

 about the matter inductively ; while until physiology and chemistry 

 are much more advanced, we can know nothing about it 

 deductively. 



These considerations ought to prevent us from receiving state- 

 ments (Taylor's Medical Jurisprudence, pp. 644, 678, and many 

 other books) which positively affirm the existence of hereditary 

 madness and hereditary suicide ; and the same remark applies to 

 hereditary disease (on which see some admirable observations in 

 Phillips on Scrofula, pp. 101 120, London, 1846) ; and with still 

 greater force does it apply to hereditary views and hereditary 

 virtues ; inasmuch as ethical phenomena have not been registered 

 as carefully as physiological ones, and therefore our conclusions 

 respecting them are even more precarious. 



In this objection, however preposterous it may appear, we find 

 all the qualities of a thoroughly scientific mind that is, one which 

 receives evidence with caution. Yet it is difficult to see what 

 method Buckle would have us adopt in researches of this kind. 

 Is it the differential method, which consists in comparing the facts 

 of heredity with the exceptions to it, in accounting for the latter, 

 and in showing why they do not come under the law? Possibly 

 this might be done. Or is it the statistical method, which consists 

 in accepting the facts as they present themselves, in grouping, on 

 the one hand, those which have an hereditary character, and on 

 the other those which have not, and in estimating arithmetically 

 the relations of the two groups ? We shall see hereafter that this 

 has been attempted. 



