Recently added emphasis has been given to the importance of the constitutional factor in disease through the work of Draper,1 Kretschmer2 and others. The work of these men has been based largely on structural or morphologic observations. It was felt that an approach to this problem from the standpoint of the basic physiologic pattern, as suggested by Draper, might also prove of interest.
With this in view, a study of the distribution of the blood grouping was made in a series of tuberculous subjects, blood type being generally regarded as an outstanding constitutional quality, and tuberculosis as representing a somatic disorder of traditionally important genotypic relationship. The only previous work appearing in the literature is that of Alexander,3 who studied fifty cases with apparently negative results. Obviously, however, such a series is too small to warrant a valid conclusion.
The series utilized in this study was comprised of 400 cases,