with stronger resistance survive. The proof of these propositions has been abundantly given by Karl Pearson, G. Archdall Reid and others. Their main points may be indicated. In the first place it must be shown that the morbidity from tuberculosis is largely due to heredity—a point on which most medical men are still uninformed. Measurement of the direct correlation between phthisis in parent and child shows it to be about .5, i. e., what one expects if it is a matter of heredity. This is the coefficient for most physical and mental characters: it is the coefficient for such pathological traits as deafness and insanity, which are obviously due in most cases to inheritance rather than infection.
But, one objects, this high correlation between parent and child does not prove inheritance,—it obviously proves infection. The family relations are so intimate that it is folly to overlook this factor in the spread of the disease.
Very well, Professor Pearson replied, if the relations between parent and child are so intimate that they lead to infection, they are certainly not less intimate between husband and wife, and there ought to be just as much infection in this relationship as in the former. The correlation was measured in thousands of cases and was found to lie around .25, being lowest in the poorer classes and highest in the well-to-do classes.
At first glance this seems partly to confirm the objection—it looks as if there must be a considerable amount of tubercular infection between husband and wife. But when it is found that the resemblance between husband and wife in the matter of insanity is also .25, the objection becomes less formidable. Certainly it will hardly be argued that one of the partners infects the other with this disability.
As a fact, a correlation of .25 between husband and wife, for tuberculosis, is only partly due to infection. What it does mean is that