1. The influence of acidity, developed in cream that had been pasteurized while sweet and subsequently ripened, was studied over a pH range (in the buttermilk) from 4.5 to 7.0.

2. Three series of creams, viz., 20, 30 and 37.5 percent fat, were investigated.

3. The losses (calculated as percentage of the total fat) for all three series varied little in the pH range 7.0 to 5.5. In this region the least variation was encountered with 30 percent cream; such tendency in loss changes as was exhibited by 30 percent cream was toward a decreasing loss with decreasing pH, while the 37.5 percent cream losses tended to pass through a. minimum at pH 6.3 to 6.4. From pH 5.5 to pH 4.8 or 4.9 the losses rose to maxima (at 4.8 to 4.9) with 20 and 37.5 percent cream; a slight rise with no definite maximum at pH 4.8 to 4.9 occurred with 30 percent cream. With all three creams a marked change of function in the curves (loss vs. pH of buttermilk) occurred at pH 4.8 to 4.9; the loss dropped sharply and in practically linear fashion from that point to pH 4.5.

4. The above facts (especially the maximum at pH 4.8 to 4.9) were interpreted as indicating that casein plays an important role in the protection of the fat globules in cream, if the churning loss is taken as a measure of protective action.

5. The churning loss data correlated very well with electro-kinetic potentials of the fat globules, determined by Sommer and North and re-presented here.

6. Churning times show closer correlation with pH of buttermilk the lower the fat test of the cream. Other factors such as change in protein to fat ratio, increased viscosity, greater ease of whipping, lower specific gravity, etc., may be involved in affecting the churning times of the richer creams.

7. Churning time data in this and the third bulletin of this series indicate that, if the fat and serum in cream are in proper physical state and chemical equilibrium, no hard and fast rule can be drawn that long or short churning times must be associated with high losses.

8. Data show that the fat test of the buttermilk in low fat (18 to 20 percent), highly ripened creams (pH 4.5 to 4.6) is considerably lower than those for high fat (30 to 37.5 percent), sweet cream (pH 6.5). Calculated as the percentage of the total fat churned, however, the low fat, highly ripened cream losses are approximately equivalent to those for 30 percent sweet cream and are slightly higher than those for 37.5 percent sweet cream. This shows that the American, Australian and New Zealand churning losses compare very favorably with those obtained in Denmark, Germany and Holland.

9. Based on the data presented and others from the literature it was hypothecated that the protective action at the fat globule interface was caused by two types of protective materials- one labile and one non-labile. The latter is closely associated with the fat, presumably on the fat side of the interface, and consists of a protein-phospholipin complex. The former is oriented from the water side of the interface and is composed of all the surface tension lowering constituents of the serum. Of the serum constituents casein probably plays the most important protective role as indicated by certain dairy phenomena.

10. If the validity of the hypothesis presented is assumed, the following explanation of the churning process seems logical: Utilization of the labile protective materials, to stabilize foam interfaces, decreases their concentration at the fat-serum interface. When the labile to non-labile protective material ratio is sufficiently small that the fat globules are in an unstable state, they merge and lose their identity. This merger weakens the forces at the force centers of the fat globules to such an extent that the non-labile materials are released from the fat globule surfaces and are incorporated in the buttermilk, while the fat unites to form butter.