1. The denaturation of albumin and globulin took place rapidly in samples of milk heated at temperatures of 75°C. and above, and was complete in approximately 60, 30, 10–15, and 5–10 min. at 80, 90, 95, and 100°C. respectively.

2. There was no change in the non-protein nitrogen content of milk on heating at temperatures up to 100°C.; on continued heating at 95 and 100°C. extremely small amounts of proteose were produced by hydrolysis of protein.

3. In milk heated at 115 and 120°C. the denaturation of albumin and globulin was followed by appreciable hydrolysis of protein, which resulted in considerable increases in the proteose and in the non-protein nitrogen contents.

1. The literature on the soluble protein fraction of milk has been reviewed to show the difficulties that are encountered in attempting to decide whether the fraction consists, as is usually supposed, of lactalbumin and lactoglobulin only, or whether it consists partly of secondary proteins of a proteose-peptone nature.

2. It has been shown that on boiling the casein-free filtrate of milk, maximum coagulation of albumin and globulin occurs at a pH. of 4–75–4–80. At this pS. prolonged boiling does not cause hydrolysis, and the amount of albumin and globulin N coagulated represents about 70 per cent of the total soluble protein N of milk.

3.By precipitation of the casein-free nitrate of milk with various con- centrations of trichloracetic acid, a partition of the soluble proteins has been effected which strongly suggests the presence of proteose-peptone substances to an extent in agreement with the results of heat coagulation.

4. The maximum amount of albumin and globulin rendered coagulable by heating milk itself has been determined by precipitating denatured albumin and globulin along with casein at pH. 4–7. This maximum represented an average of 76 per cent of the total soluble protein N in a series of normal milk samples, and reasons have been given for considering this maximum as the total of true albumin and globulin present.

5. The conclusion has been reached that the soluble protein fraction of normal fresh milk is composed of approximately 76 per cent albumin and globulin, and 24 per cent proteose-peptone substances.

1. Acetylmethylcarbinol in cream is partly destroyed by neutralization and pasteurization, but the diacetyl content is practically unchanged.

2. Diacetyl and carbinol are retained in butter with the butter-milk fraction.

3. Diacetyl is present in butter: (a) sweet-cream 0·05 p.p.m., (b) “mild starter” 0·3–0·4 p.p.m., and (c) “high acid” 1·5 p.p.m. The diacetyl content remains fairly constant during 3½ months' storage at 14° F.

4. Diacetyl in butter does not affect the keeping quality or vitamin A content.

5. The diacetyl content tends to increase when butter is held at 40° F. for several days after manufacture. The change is not caused by oxidation with air, but is probably due to dehydrogenase activity of the “starter” organisms or their enzymes.

6. Experiments with butter “starters” are described.

Apart from mastitis milk there appear to be two types of soft-curd milks:

(1) Milks of individual cows which throughout a lactation or throughout their lives persistently give soft-curd milks. Such milks readily respond to additions of acid and appear to be due to the breed and individuality of the cow and not to any abnormality.

(2) Milks of cows or herds which, though normally giving a satisfactory coagulum, suddenly become rennet-resistant for a certain period. Such milks are more difficult to deal with as in general they do not respond to additions of acid unless these additions are large.

There is little evidence available to show whether the two types are regularly related, but some individual cows in a herd do produce soft curd milk of the type which is unresponsive to the addition of acid regularly throughout a season when the majority of the cows in the herd are giving normal renneting milk. In the work described above on individual cow's milk there has been no attempt to classify the milks according to the effect of acid on the coagulation, although it is now realized that such a course would have been desirable.

Traces of certain metallic contaminants (copper, iron, lead, zinc, aluminium and manganese), when added at the rate of 3 and 7 p.p.m. to milk used for the manufacture of Cheddar cheese, cause certain flavour (and colour) defects in the matured cheeses, of which, however, no indication is given when the cheese is plugged and graded at about 14 days after manufacture.

The distribution of copper, iron, manganese and lead between curd and whey during the cheese-making process is recorded and the action of traces of various metals on the acidity development in a typical "starter" culture has been followed.

The discolorations noted for copper- and iron-contaminated cheeses are considered to be due not to the formation of their sulphides (although this is the case for lead discoloration), but to the atmospheric oxidation of a colourless metal: protein complex.

The properties of commercial annatto solutions have been examined from the point of view of the cheese-making process.

(1) The effect of initial colour strength, pH of the medium and distribution during manufacture are recorded.

(2) Methods are described for the extraction of the annatto colour from milk products.

(3) The effects of old annatto solutions, metallic impurities (copper and iron) and sulphydryl are shown.

(4) It is suggested that the mechanism underlying the development of “bleached” and “pink” discoloration in cheese is an oxidation process, and the effect of oxidation on annatto solutions has been studied in some detail.

(5) Suggested standards for annatto extracts for use as cheese colours are appended.

It has been shown that the acidity developed in a vat of milk at the “running” stage has a very important influence on the quality of the resultant cheese. A high acidity causes increased mineral losses in the whey and a corresponding lowering of the mineral content of the cheese, while it also has a detrimental effect on the body of the cheese. The acidity at “running” also controls to a great extent the acidity of the cheese produced. Even where the acidities at the late stages of the process are kept the same, vats of cheese made at a high “running” acidity show more acid characteristics than cheese made at low “running” acidity.

The quantity of lactose incorporated in the cheese is also of importance in determining the acidity of the cheese. An increase in the lactose content, either by raising the lactose concentration of the cheese milk or by the addition of lactose to the curd, produces a more acid type of cheese at maturity, although some unfermented lactose is still present in the cheese at 4 months after manufacture.

The experiments described above have shown that on the whole there is a close relationship between titratable acidity and lactic acid concentration in wheys obtained under various conditions of cheese-making. Variations in the rate of acid development or in the lactose content of the cheese milk have no significant effect on the relationship. Variations in the mineral content of the curd brought about by differences in the composition of the milk or by different acidities up to the running stage do have a certain effect. In curd with a lower mineral content there is a less complete neutralization of lactic acid and the wheys show a higher titratable acidity for the same lactate content. The differences, however, are small and affect only the wheys coming off late in the cheese-making process.

A number of chemical substances that might be expected, or have been claimed, to affect the rate of chemical ripening have been incorporated in cheese at “salting”. The effect in most cases has been insignificant. It would appear that cheese is a remarkably stable system, the chief factors in the rate of chemical ripening being the concentration of rennet, salt and free water. The importance of the interrelationship between the last two factors has been discussed.