1. A study of the ability of pure cultures to reduce methylene blue and resazurin at 37·5° C., to reduce methylene blue at 22° C. and to cause milk spoilage as measured by precipitation with 68 % alcohol, and the clot-on-boiling test has revealed marked differences in behaviour between cultures.

2. At 37·5° C., with the exception of two cultures, there was good agreement between the time of reduction of methylene blue and resazurin. The two exceptions were a group B Streptococcus which reduced methylene blue but failed to reduce resazurin during the 9 hr. period of observation and an unclassified Streptococcus growing at 45° C. which reduced resazurin before methylene blue; the latter dye was found to retard the multiplication of the culture slightly.

Bacterial phosphatases of growing organisms can be simply demonstrated by incorporating di–sodium–p–nitrophenylphosphate (pNPP) into suitable media. The phosphatase content of a bacterial suspension can be assessed by incubating a measured amount of it with unbuffered or suitably buffered pNPP solutions. The yellow colour (in alkaline solution) of p-nitrophenol liberated from the substrate by phosphatase can be measured accurately and conveniently in a photoelectric absorptiometer, or estimated in a Lovibond comparator using permanent colour standards. The usefulness of the tests for differentiating between members of certain genera is demonstrated by the reactions of Bact. aerogenes and Bact. cloacae, for which a close relationship between phosphatase production and gelatine liquefaction times has been found. The optimal conditions for hydrolysis of pNPP by the phosphatase of a strain of Bact. aerogenes have been established.

Of thirty-five strains of streptococci isolated from milk produced in the Jura mountains, two were strains of Streptococcus cremoris. It appears, therefore, that this organism occurs in the wild state.

Although milk may be considered to be the normal habitat of the lactic streptococci, their adaptation to this environment must have followed the domestication of mammals, for these organisms are not pathogenic and do not normally occur in the udder. If this reasoning is accepted, then the small biological age of these organisms logically follows, and possibly they are the youngest members of this genus.

Using non-selective methods of isolation, the species that was detected with greatest frequency in raw milks of varying purity is Alcaligenes viscosus. The majority of the cultures do not produce a ropy condition in milk or other media. In some strains the oxidation of citrate, and hence the production of alkalinity in milk, is slow and erratic, being dependent on the occurrence of a mutation. The only Gram-negative rod detected in laboratory-pasteurized milk was A. tolerans n.sp., a description of which is given. That species is similar in general character to A. viscosus but differs principally in its greater resistance to heat, its inability to attack fats, and its erratic and usually poor growth on laboratory media. Greatly improved growth on nutrient agar is obtained by the addition of lactate and relatively high concentrations of ferrous or ferric salts.

Three strains of B. cereus were grown in milk at nine different temperature combinations. Broken cream could be prevented by holding milk at 5° C. or lower for 24 hr. before transferring to 15 or 18° C. for a further 24 hr., but with a subsequent period of storage at 22° C. the fault was not prevented. Broken cream developed after storage at 10° C. followed by 18 and 22° C., but not in milk kept at 10° C. and then transferred to 15° C. After preliminary storage at 15° C. the fault appeared within a further 24 hr. regardless of the subsequent storage temperature.

With two strains of B. cereus it was observed that pasteurization accelerated spore germination and hence the development of the fault. With low levels of inoculum there was delay in the appearance of the fault with one strain of the organism.

In twenty-four samples of commerically pasteurized milk it was shown that the fault could be controlled by refrigeration at 5° C. before distribution except when the subsequent temperature was allowed to rise to 22° C.

Contrary to the suggestion made by Davis (15) preliminary observations indicate that the growth of B. cereus and the development of broken cream is not inhibited by the presence of acid-producing organisms such as Str. lactis.

An effect simulating broken cream in whole milk was produced when Bacillus cereus was grown in washed cream at 22° C. A similar but less marked effect was obtained when a concentrated culture filtrate of B. cereus was added to washed cream.

Microscopic examination of washed cream to which B. cereus or its concentrated filtrate had been added showed that the fat-globule membranes had been broken down. The hydrolysis of lecithin in washed cream in which B. cereus had grown was demonstrated by qualitative estimation of free choline.

Since these results could not be reproduced when a non-lecithinase-producing strain of B. cereus or its concentrated culture filtrate was added to washed cream, it appears that the hydrolysis of the lecithin of the fat-globule membrane is at least partly responsible for the formation of broken cream.

With 1·0 p.p.m. of added copper, well-defined oxidized flavours described as cardboard, oily and metallic developed in pasteurized whole milk after ·2–3 days' storage at 3° C. With 0°1 p.p.m. copper only slight and irregular off-flavours, usually described as cardboard, appeared. Partial oxidation of the ascorbic acid of the milk with H2O2 did not increase its susceptibility to oxidized flavour development.

Marked off-flavours were always associated with peroxide values in the milk fat of 0·2 and more (m.equiv. oxygen/kg.) determined by the ferric thiocyanate technique. With fat peroxide values of about 0·5, usually obtained after 6–7 days at 3° C. with 1·0 p.p.m. copper, the oxidized flavour was very strong and the milk usually undrinkable.

The method of assaying rennin described by Berridge(1) requires the formation under reproducible conditions of a thin flowing film of milk. Although the skill needed to keep the film flowing at the optimum rate as the clotting point is approached is readily acquired, an automatic method has the advantage of complete reproducibility. A suitable film can be maintained if 10 ml. of milk are placed in a boiling tube of 2·2 cm. internal diameter and the tube, with its axis inclined at about 30° to the horizontal, rotated at 3·0 r.p.m. In the apparatus now used, the boiling tube is mounted on a rubber bung, which in turn is mounted concentrically on a spindle connected through Meccano gears to a 'Sangamo' electric clock motor. A special mechanical coupling allows the bung to be rotated by hand one revolution forwards, thus allowing 20 sec. for fitting or releasing the boiling tube before the drive is again taken up.

The milk from ten sows has been studied over one lactation. The density, viscosity, electrical conductivity and pH have been recorded tbroughout the lactation, and in two cases the study extended to the colostrum.

The fat-globule size distribution patterns have been determined throughout lactation for ten sows and through the colostrum phase for two.

1. Three samples of spray-dried whole milk were made from milk pre-heated at 190, 230° and 230° F. with regenerative cooling to 190° F., in an A.P.V. H.M. Paraflow Heat Exchanger and dried in a Kestner plant. The powders were packed in plain tin-plate cans and stored at 17, 37 and 47° C. The keeping quality of the powders was determined by examination at intervals for palatability, absorption of oxygen and peroxide value of the fat. The biological value and true digestibility of the protein of the powders were estimated in rat-feeding experiments initially and after 6 months' storage. The breadmaking properties of the powders were also examined.

2. The keeping quality, biological and baking tests showed that the three milk powders were very much alike, and that no advantage was gained by raising the preheating temperature of the liquid milk from 190 to 230° F.

1. Ratings for firmness which were carried out on 150 Cheshire cheese at 2 weeks old, by three judges, were examined so as to extract the expected psychophysical errors.

2. It was found that the value for firmness given to any particular cheese tended to be a function of the average firmness of the group of cheese in which it was judged, so that a cheese of a given objective value would be given a different rating according to its position in the series being judged.

3. A cheese of a given value presented after a firm one is judged softer than if it had been presented after a soft one. This contrast effect was masked by the influence of the group as described in (2). Nevertheless, it seems probable that contrast effects influence the judgements.

4. The frequency distributions of the subjective judgements were examined, and it was noted that the ‘semi-expert’ judge gave subjective values which approximated very closely to a normal distribution, more nearly so than the judgements of the other two judges.

5. Reasons are suggested for the differences in the values allotted to the cheese as judged by the different judges.

6. Some suggestions are made for eliminating the effects described.