1716. [left side] Eugenics laid down. The fundamental law, is, as already stated: Whenever the same unit defect exists in both parental protoplasms it will appear in all the offspring. The "unit defect" is not, as already pointed out, easily determined, nor is a given gross defect probably identical in the parental strain unless the parents are cousins. Despite these difficulties in its application the rule holds, by and large, as a valuable first approximation. Of unit defects the weakness of mucous membranes seems to be a good illustration. If both parents are subject to colds, catarrh, bronchitis, asthma or lung-tuberculosis all or nearly all of the children are liable to these diseases. The same is true in some forms of nervous disease and rheumatism. If the disease fails to appear in any child it is probably because the child died too early, or is still very young, or has been able through exceptionally favorable environment to avoid the incidence of the disease or, by strengthening other means of defence, to hold it down or eliminate it, even after attack. The expectation that is usually realized, however, is 22 [right side] Fit and Unfit Matings that all show a weakness to the same disease as their parents. If liability to the disease is found in the protoplasm of both parental strains but is shown in the soma of one only of the parents, then it will probably occur in one-half of the offspring. Examples are found in the families whose pedigrees are given in the diagrams. (C 5; B 3; C 4; C 7; C 10; C 8, C 9.) The total of the last generation in these examples gives 18 subject to the specific disease and 23 non-subject where 20 1/2 a dn 20 1/2 are the expectation. The excess of the non-subject may be explained on the same ground as the exceptions to complete incidence of disease referred to in the preceding paragraph. If both parents belong to strains having the same unit defect even though they have it not themselves we may expect either that one-quarter or none of the children will have the defect, depending on earlier ancestry. This rule is illustrated by some of my cases. (D 1, D3.) If one parent belongs to a strain with a unit defect while the other strain is without the defect then the children will be without the defect. 23 [end]