The fermentation reactions and some ancillary characters of 102 strains of virulent C. diphtheriae, 31 non-virulent strains and 150 diphtheroids have been studied.

With the possible exception of dextrin which proved to be an unsuitable substance for biological tests all the strains of virulent C. diphtheriae examined were found to have constant biochemical reactions, fermenting glucose, maltose and galactose and having no action on cane-sugar, lactose and mannite.

The strains classified as non-virulent C. diphtheriae had the same biochemical characters as the virulent strains and were indistinguishable from them either on morphological or cultural grounds.

The diphtheroids examined showed great diversity in their power of attacking the carbohydrates used, at least eleven biochemical groups being recognised.

By the use of three carbohydrates, glucose, galactose and cane-sugar 90 per cent, of the diphtheroids in this series could be distinguished from C. diphtheriae by their fermentation reactions. In the remaining 10 per cent, a clue as to the nature of the organism is often obtained from the character of the growth on agar.

Among the corynebacteria examined other than C. diphtheriae no constant relationship was found between source, biochemical, morphological and other characters.

Nevertheless in dealing with members of the genus in pure culture, the morphology, character of the growth on agar and fermentation reactions taken together give useful information as to whether the organism under consideration is C. diphtheriae (virulent or non-virulent) or a diphtheroid.

My best thanks are due to Dr Ledingham for much help and advice given throughout the work and to Dr Atkin and Dr Robison for their assistance in dealing with the lactose and dextrin problems respectively.