No. 502] NOTES AND LITERATURE 



type in publications emanating from public health leagues that 

 ' ' Tuberculosis Is Not Hereditary." I do not know on what 

 scientific basis such a dogma rests. 



Professor Pearson has found cogent proof in the first of these 

 studies that the |>hthi>ir;i] <li;i11it'sis is just as hereditary as any 

 human characteristic we know about. It would take too much 

 space to completely review this paper. In a few words it may 

 be enough to say that he does not jump at the conclusion that 

 correlation coefficients necessarily show heredity. The question 

 of infection through members of the same family living in close 

 contact is discussed at length; but the analysis reveals no evi- 

 dence that direct infection is in any way important, as compared 

 to the heritable diathesis. 



For instance, in his second paper on this same subject he finds 

 that if a husband is tubercular, then there is a probability that 

 the wife will also be found tuberculous, and vice versa, but this 

 correlation is not nearly sq high as that between brothers and 

 brothers, sisters and sisters, and brothers and sisters. Yet 

 opportunities for direct infection in the case of husband and 

 wife are of course vastly greater than among brothers and 

 brothers, etc., who by the time of the average age of onset of 

 the disease (twenty to thirty years) have already ceased to live 

 in the same households. 



The question of assortative mating comes in to explain a cer- 

 tain amount of this observed correlation between husband and 

 wife. Assortative mating is a convenient name for the tend- 

 ency of like to mate with like, aside from any question as to 

 what causes may bring about the similarity in question. 

 It is a popular belief that tall men marry short women, and 

 blonds are attracted by brunettes, but the truth of the matter 

 seems to be quite the reverse. In nine series of physical char- 

 acteristics the correlations of resemblance between husband and 

 wife have been found to range between r=.20 and r = .28. 

 For physical characteristics nine series show a range between 

 r = .ll and r=AS, with an average of r=.2±.° The corre- 

 lation coefficient for insanity between husband and wife is 

 r=.30. 



