The size distribution of fat globules was determined microscopically in cream samples taken from different parts of the Vacreator during treatment of cream.

A considerable increase in numbers of small globules (splitting) was found in samples which had passed the equilibrium valve. Smaller increases took place in the pasteurizing chamber and in the extractor pump. Increase in numbers of large globules (clumping) occurred in the cooler section.

Splitting of globules could be reduced: in the pasteurizer by reducing velocity of incoming steam; at the equilibrium valve by reducing the vacuum drop across the valve or by introducing part of the steam at the infuser instead of into the pasteurizer.

It is concluded that splitting of globules in the Vacreator is due to a homogenizing effect where steam and cream travel at a high velocity. Flash-boiling causes clumping but not splitting of globules. These conclusions should apply to any equipment for pasteurization or deodorization of cream with direct steam.

The thanks of the author are due to Mr J. Singleton (Superintendent) and Mr B. Le Heron (Buttermaker) of the Institute factory for arranging and carrying out Vacreator treatments of cream, to Dr F. H. McDowall for helpful advice and criticism and to Mrs R. P. Haverkamp for assistance in measuring fat globules.