The work done in various laboratories during the last few years has shown that starters, instead of being pure cultures of Skeptococcis lactis, as was at one time commonly supposed, contain two or more organisms, at least one of which produces considerable amounts of volatile acid. Good starters have regularly been found to yield a high and rather definite amount of volatile acidity but natural souring and sometimes poor starters also give a high volatile acidity, so that such an acidity is no guarantee as to the quality of a starter. This indicates that while the amount of volatile acid produced is important with a starter, the character of the volatile acids must also be considered and suggests the desirability of knowing what kinds of acids are produced in starters and in pure cultures of the starter organisms. The work herein reported represents an attempt to secure information along this line and also on the kinds of acids produced during prolonged natural souring such as occurs in old cream.