Our purpose in this paper is, first, to discuss the value to clinical medicine of a consideration of heredity, and, secondly, to show that sufficiently good genetic records can be obtained from hospital and private patients to be of use to future students. But the useful yield of dependable family histories, like certain chemical residues, is so small in proportion to the great mass of individuals who must be interviewed and discarded that the outlook is discouraging to the impatient investigator. For example, in our experience only about one fourth of all the individuals interviewed were able to give reliable histories. Furthermore, even these can seldom recall facts which concern ancestors or lateral connections more remote than grandparents. The term human constitution, it was previously stated, should connote that aggregate of hereditarial characters, influenced more or less by environment, which determines the individual's reaction, successful or unsuccessful, to the stress