226 HUMAN DISEASES 



and frequently perish without leaving offspring. It is conceivable, 

 of course, that insusceptibility to this or that disease may be 

 correlated to some other and apparently unrelated character such 

 as height or colour or peculiarity of disposition, which therefore 

 would undergo concurrent evolution as the race grew resistant to 

 the disease. But, speaking comparatively, correlated characters of 

 this sort are very rare, and the study of disease has revealed none, 

 though the opportunities for observation are excellent ; for we are 

 able to compare races which are very susceptible to this or that 

 disease with races which are highly resistant. Very certainly the 

 comparison of such races reveals no such correlations. 



373. Some biometric investigations, it is true, seem to indicate 

 lines of progression and retrogression other than that related to 

 disease ; but the data on which they are founded are very scanty 

 as compared with the evidence that can be gleaned from tables of 

 mortality ; they cover only a few generations, and make no very 

 serious attempt to distinguish real variations from acquirements. 

 In the case of disease we are able not only to obtain precise 

 numerical information on a large scale as to the causes of death, but 

 it is in our power to test such theories of evolution as we may found 

 on that information in a very decisive way, by comparing races 

 which have long and severely suffered from any disease with races 

 that have been newly introduced to it. By this means we may 

 discover, for example, whether the selective elimination which 

 we suspect to be a cause of progression is a true cause, and 

 whether the progression, if any, has correlated to it the growth of 

 some other trait. 



374. Of deaths due to disease by far the greater number are 

 caused by microbic maladies, or by the abuse of such narcotics as 

 alcohol and opium. Probably, with the exception of violent death, 

 which is non-selective, and cancer, which occurs mainly during 

 advanced life, and therefore exercises little or no selection, there 

 are no other very important causes of elimination. Thus of the 

 76,844 deaths which occurred in England and Wales during the 

 year 1905 from diseases of the heart and blood-vessels, it is pro- 

 bable that the great majority had a microbic or alcoholic origin, 

 especially in the case of people who died before the end of the 

 reproductive period. Doubtless, also, the deaths of very many of 

 the 18,871 infants who perished because prematurely born, or who 

 died of 'convulsions '(12,53 3), or of teething (2343), were due to the 

 same causes acting directly on the children or through the parents. 



375. Half a century ago the nature of microbic diseases was 



