1. The influence of the age of cows on the milk yield and on the milk solids is traced. The milk yield reaches a maximum at about the sixth lactation and then declines. The concentration of fat and solids-not-fat shows a gradual decline from the first lactation onwards.

2. During lactations in which the yield is below the normal productivity of a cow, the quality of the milk also tends to be subnormal.

In a previous communication (Allen (1)) it was shown that Cheddar cheese made from milk of a very low bacterial count (Grade A (T.T.)) exhibited a retarded rate of ripening, as evidenced by proteolysis and development of flavour and texture, when compared with similar cheese made from ordinary market milk, and the view was advanced that the miscellaneous flora of market milk produces sufficient proteolysis of the casein in the early stages of manufacture to provide the starter organisms with necessary nitrogen compounds for rapid growth.

1. Some cases of sudden failure in the growth of cheese starter cultures are described.

2. It is shown that the cause lies in the method of pasteurisation of the milk in which the cultures are grown.

3. Aeration of the milk immediately before the inoculation of the culture is shown with certain cultures to have an inhibitory efEect on the growth of the organisms. Normal starter cultures are not affected by aeration of the milk; they may, however, suddenly develop this sensitivity to aeration and may just as suddenly revert to the normal state.

4. The mere absorption of atmospheric oxygen while milk is being cooled after pasteurisation in a wide-mouthed cylindrical vessel may constitute an aeration sufficient to inhibit sensitive cultures of lactic streptococci. Milk pasteurised and cooled in conical flasks plugged with cotton-wool does not reabsorb sufficient oxygen to lead to inhibition, provided that the flasks are not violently agitated.

5. The possible mechanism of the inhibition of the sensitive streptococci is discussed. Evidence is brought forward which tends to show that the action is connected with the oxidation-reduction reactions by means of which the organisms normally obtain their growth energy.

6. The practical significance of the phenomenon in the process of cheese manufacture is emphasised.

1. The process of ripening in experimental cheeses was followed by periodic estimations of the amounts of various types of protein degradation product formed.

2. No connection between the type of nitrogen partition and the commercial quality of the ripened cheese could be traced.

3. The proteolytic powers of the various pure cultures of streptococci used as starters in the cheese vats seemed to have little influence on the rate of proteolysis in the cheeses during the ripening period. The main trend of the protein breakdown was governed by the rate and total extent of acid formation in the milk and curd during the manufacturing process. This suggests that the rennet enzyme is the principal proteolytic agent, and that the most significant function of the streptococci is the formation of acid which in turn influences the rate of action of the rennet enzyme.

4. No specific connection between the numbers of streptococci present in the cheese milk at the time of renneting and the rate of ripening of the cheese could be traced.

5. Some evidence was found tending to suggest that minor differences in the type of protein degradation occurring during the ripening period influenced the changes of bacterial flora in the cheese. Lactobacilli appeared to develop earlier where the proportion of simpler protein derivatives was high.

A description is given of the characters of certain streptococci which in small numbers are sometimes found in the milk of individual cows. Although the organisms possessed high haemolytic activity, they showed no evidence of pathogenic power, either for the cow's udder or for ordinary small laboratory animals. Not only are they different from the streptococci of bovine mastitis, but they are also to be distinguished from streptococci of the Str. pyogenes group which are occasionally found in milk. One of their principal characters is the formation on solid media of minute colonies, consisting of relatively minute cocci, which grow in broth as very short chains.

The fifty-eight strains studied could be divided into two groups on biochemical grounds: (a) those fermenting glucose and lactose with production of small amounts of acid, (b) those which ferment mannite and salicin as well. The former are referred to as “low-acid” strains and the latter as Str. infrequens, as they bear a resemblance to the organism given this name in Holman's classification. Fermentation of sorbite and trehalose occurred with streptococci of both groups.

The two groups could be distinguished from Str. pyogenes and the ordinary mastitis Streptococcus by direct agglutination. For distinguishing the two groups from each other, agglutinin absorption methods were required. It was then found that the low-acid Streptococcus possessed two antigens of which one only was represented in Str. infrequens.

1. The effect of increasing concentrations of milk solids on the degree of ionisation of electrolytes cannot be measured by determinations of electrical conductivity, owing to the influence of viscosity on ionic mobility.

2. An approximately linear relationship exists between concentration of milk solids and the freezing-point depression. Above a certain concentration (equivalent to 34 g. milk solids per 100 g. of water), however, the solutions are saturated with respect to lactose, and any increase in freezing-point depression beyond this point is limited to that due to the other osmotically active constituents of milk.

3. There is no evidence of the existence of “bound” water in solution of milk solids up to concentrations equivalent to 65 g. milk solids per 100 g. water.

4. An approximately linear relationship exists between concentration of milk solids and hydrogen-ion concentration, increase in the concentration of milk solids resulting in a markedly increased acidity.

5. The applications of these results to the heat coagulation of evaporated and dried milks are briefly discussed.

Iodometric methods have been applied to the determination of lactose in milk by Kolthoff(1), who used freshly precipitated zinc ferrocyanide as protein precipitant, and by Fine(2) who made use of colloidal iron. The reduction of ferricyanide by reducing sugars has been used as a semi-micro method by Issekutz and von Both(3) and has been applied to the estimation of lactose in milk by Gohr(4) who also used ferrocyanide as protein precipitant.