Summary: Four out of seven influenzal meningitis strains isolated years apart, were proved by agglutinin absorption tests to be of one type; the remaining three were of different types. From the respiratory cases, five strains from different individuals were obtained which were of the same type. No other groups of more than two members were obtained in agglutination tests with any cultures. This was true of both the original pandemic and also of the new outbreak strains. Atypical hemoglobinophilic bacilli were found in 18 cases (normal individuals and cases of cold) and in five cases these were found in coexistence with typical influenza bacilli. These atypical organisms produced hemolysis, did not form indol and did not ferment glucose. About 75 per cent of typical influenza bacilli form indol; from 50 to 60 per cent ferment glucose somewhat irregularly and not energetically. The indol reactions did not correspond fully among the meningitis group of four serologically identical strains. Of these, three formed indol—one did not. In the family group, though the strains were not all identical serologically, all except the atypical hemoglobinophilic colonies formed indol. In general, it was found that strains having the same immunological characteristics had the same cultural reactions, but there were some exceptions. Many colonies were fished from each of a number of selected cases and tested with homologous and heterologous serums. In a few cases as many as 20 or 30 colonies from each case were tested with the homologous serum prepared from one of these colonies. In most cases the fishings of influenza bacilli proved to be of one type; at least one type greatly predominated. But in a minority of instances, the colonies from the same case were not of the same type as proved by agglutination absorption tests. In carriers or where the infection was mild, more than one type variety of the influenza bacillus was apt to be found in the same case. In several convalescent carriers, tests were made during periods of from several weeks to three months. These resulted in finding in all but one, the original type of the influenza bacillus. In one case, tests made one year apart proved the original type present.