It was observed, in the course of blood grouping and matching, that certain bloods apparently belonging to group II did not match. In an attempt to find why these bloods did not match the serums were tested with known cell suspensions and some were found to be group IV (Moss). It was then discovered that the serum used as known group III agglutinated some group IV cells, and a thorough study of this serum was undertaken.
The agglutination reactions of this serum (III1) were tested with eight group I, seventy-two group II, fifteen group III and a hundred group IV cell suspensions. The cells from our subject were tested with seven group III, three group I and several group II serums. The results are given in table 1. The method used in all agglutination tests was as follows: Two drops of serum were mixed on a clean slide with