Milk protease, the naturally occurring proteolytic enzyme from fresh raw bovine milk, was purified 629-fold. Purification methods are described which included acid precipitation, ion-exchange and gel chromatography and ammonium sulphate fractionation. It was necessary to use a large-scale purification as the amounts of enzyme in milk are very low.

The basic properties of the enzyme were elucidated from experiments on this partially purified fraction. Under the conditions of assay these showed a wide pH optimum from 6·5 to 9·0 and a temperature optimum of 37°C. The protease was fairly heat stable, the stability being pH dependent.

The source of the enzyme appeared to be the mammary gland, where the enzyme was also detected.

Ceramide, an N-acyl derivative of sphingosine bases, has been isolated for the first time from bovine milk in the free form and its constitution studied. Sixteen component fatty acids were found, among which C23:0 (38·1%), C24:0 (29·5%), C22:0 (17·9%) and C16:0 (7·2%) were predominant. Seven component long-chain bases were detected. The principal bases were C18-sphingosine (35·0%) and C16-sphingosine (31·6%).

The kinetics of the release of glycopeptides by the action of rennin on bovine whole milk have been studied. The initial rate of release of glycopeptides was proportional to the concentration of rennin. Using the integrated form of the Michaelis–Menten equation, values have been obtained for Km and V. These values varied between milk samples from individual cows, and for a single milk sample the value for Km increased with increasing rennin concentration. It is suggested that the Km value is partly dependent on the carbohydrate composition of individual κ-casein molecules.

The rennin coagulation time of cow's milk was inversely proportional to the amount of glycopeptides released by rennin action on the milk and to the initial velocity of the release of the glycopeptides. It was shown that the rennin coagulation of milk is largely determined by the primary phase of rennin action and, therefore, by the amount of κ-casein that is susceptible to attack by rennin.

A large number of samples (433) from the milk of Friesian herds was taken during the period November 1968 – September 1969, and a smaller number (159) from Jersey herds during the dairy ‘season’ September 1969 – April 1970, and analysed for total nitrogen (TN) and for non-protein nitrogen (NPN). The Friesian milk sample varied in TN from 0·456 to 0·624% (average 0·531 ± 0·021%) and in NPN from 0·024 to 0·054% (average 0·035 ± 0·004%). The Jersey samples varied in TN from 0·467 to 0·718% (average 0·584 ± 0·036%) and in NPN from 0·022 to 0·051% (average 0·033 ± 0·006%).The average NPN value for all the milks was 0·034 ± 0·005%.

There was a seasonal trend in NPN and TN for the Jersey and in TN but not in NPN for the Friesian milks. The proportion of NPN as a percentage of TN varied widely – for the Friesian samples from 4·9 to 10·2% (average 6·46%) and for the Jersey samples from 3·7 to 8·7% (average 5·74%).

Urea N also was estimated in 125 of the Friesian and in all the Jersey milk samples. It varied from 0·012 to 0·029% (average 0·020 ± 0·004%) in the Friesian and from 0·009 to 0·031% (average 0·020 ± 0·004%) in the Jersey samples. For all the milks the urea N as a percentage of NPN varied from 33 to 79 (average 56·5%).

It is suggested that values for ‘protein content’ of milk should refer only to ‘true’ protein ((TN – NPN) × 6·38) and that in the absence of an analytical result for NPN the use of an average value such as that above would be justified.

The flavours of the following compounds were determined in synthetic butter prepared from 84% steam-deodorized butter oil and 16% water: C6, 8, 10, 12, 14 δ-lactones, C12 γ-lactone, C2, 4, 6, 8, 10, 12, 14n-alkanoic acids, phenol, m- and p-cresols, o-methoxyphenol, indole and skatole. C6, 8, 10 δ-lactones, C2, 4, 6, 8, 10, 12, 14n-alkanoic acids, phenol, m- and p-cresols, o-methoxyphenol and volatile fractions from butter oil were also evaluated in bland butter. None of the compounds listed constitute basic butter flavour on their own, but C8, 10 δ-lactones, C10 acid, phenol, p-cresol, indole and skatole contribute to butter flavour.

A method for the partial fractionation of the αs-casein complex is described. It involves removal of κ-casein from iso-electric casein by the method of Zittle & Custer (1963) followed by fractionating the residue (αs- and β-casein) by the urea procedure of Hipp et al. (1952). αs0- and αs1-caseins precipitate from 3·3 M urea at pH 4·6, but much of the αs2-, αs3- and αs4-casein remains in solution and co-precipitates with β-casein from 1·7 M urea at pH 4·9. Chromatography on DEAE cellulose clearly separates β-casein from the αs2-, αs3- and αs4-caseins, which are not resolved by the conditions employed. The method is suitable for the fractionation of iso-electric casein into all its major components (i) κ-casein; (ii) αs0- and αs1-caseins; (iii) αs2-, αs3- and αs4-caseins; (iv) β-casein; (v) γ-casein.

Two possible hypotheses for the mechanism of milk clotting were tested. The results obtained constituted strong evidence against one and suggested that the second is improbable. Milk was separated into a 5-fold-concentrated casein micelle suspension and milk serum. Pre-renneting of the serum did not reduce the rate of clotting on subsequent addition to the micelle suspension whether or not the conditions were such that the para-κ-casein became extensively aggregated. Washing of casein micelles up to 3 times with milk dialysate at 23°C extracted very little casein from the micelles and did not increase the clotting time of micelles resuspended to about the same concentration as in milk. The results appear to constitute decisive evidence against the hypothesis of milk clotting proposed by Parry & Carroll (1969). S-carboxymethyl-κ-casein, S-carboxymethyl-κ-casein containing 2·5 dimethylaminonaphthalene sulphonyl residues per mole, and rennin-treated dimethylaminonaphthalene sulphonated-S-carboxymethyl-κ-casein all bound Ca to the same extent at 30°C and pH 6·5, over the range 0·5–15·3 mM-CaCl2. This adds support to existing evidence that milk clotting does not involve formation of Ca bridges between casein micelles.

An isothermal jacket calorimeter was set up and apparent heat capacities of skim-milk, half-and-half milk, whole milk, and their concentrates up to about 30 % total solids were measured throughout the 5–70°C temperature range. Above the butterfat melting interval, the heat capacity of every sample was linearly related to temperature and a linear regression equation was found to relate heat capacity and total solids content for each one of the 3 milk sets. By means of an overall approximation, a single expression relating heat capacity, temperature and total solids content was also derived. To facilitate calculations of relative heat content for practical purposes, an average apparent heat capacity was computed for the whole temperature range and a nomogram was constructed for its determination from fat and solids-not-fat contents in any sample.

Ten different types of milk, arranged in 3 sets, were observed to be pseudoplastic fluids with linear behaviour at high shear rates in a coaxial cylinder rheometer; at low rates of shear all the samples could be considered sufficiently Newtonian and a Höppler viscometer was used for their study. Dynamic viscosity coefficients of skimmilk, half-and-half milk, whole milk and their concentrates up to about 30% total solids were determined every 5°C between 0 and 80°C. For every milk type a quadratic equation relating t to log η was obtained. At a given temperature the concentration dependence of η for the 3 milk sets was expressed by a single relation of the same mathematical model as in the previous case. A general expression was derived to enable η to be calculated at any temperature (0–80°C), and for any concentration (0–30% total solids) whatever the fat to solids-not-fat ratio (0·01–0·4). A nomogram was also constructed for the immediate computation of η without any practical loss of accuracy.

A method is described for determining alkaline phosphatase activity in cheese using phenolphthalein monophosphate as substrate. Results are read with a Lovibond comparator after 1 h incubation. Advisory standards are proposed based on a minimum pasteurization for 15 s at 71°C of the milk used for cheese-making.

Details are given of a spectrophotometric method for determining nitrite in cheese and other dairy products at levels down to 0·05 μg NO2-N/g and of a spectrofluorimetric method for levels down to 0·003 μg NO2-N/g. Nitrite-N levels in cheese varied from 0·003 μg/g to 0·021 μg/g. These levels increased up to 4-fold after UV irradiation of the cheese extracts. No nitrite (< 0·003 μg/ml) could be detected in raw milk.

The effect of ageing fresh milk for 24 h at 4°C on the composition of 4 fractions of milk fat globule membrane (FGM) – microsomal pellet membrane (MPM), deoxycholate soluble membrane (DOCM), high density membrane (HDM) and low density membrane (LDM) – prepared from washed cream treated with sodium deoxycholate (DOC), was studied for milk of individual cows. Total FGM and its fractions were solubilized by treatment with sodium dodecyl sulphate (SDS), EDTA and β-mercaptoethanol, and the dissociated FGM proteins were separated by polyacrylamide gel electrophoresis in the presence of SDS.

Ageing resulted in a greater loss of phospholipid during cream preparation than was found with fresh milk, and also in a reduction in the total amount of FGM isolated. Loss of FGM material was confined to MPM, DOCM and LDM and was accompanied by compositional changes in DOCM and LDM, quantitatively the most significant fractions. Ageing also produced changes in the gel electrophoresis patterns of DOCM, where the band of greatest mobility (mol. wt about 16000) was partially lost, and of LDM, where there was an increase observed in the major protein band.

The implication of the results on the proposed models of FGM structure is discussed.

Three groups of Friesian cows, one of cows in early lactation, one of cows in mid-lactation and one of cows in late lactation, received over successive 4-week periods according to a double changeover Latin square design, rations providing a low, a normal and a high plane of energy nutrition. With the group of cows in early lactation the treatments were repeated again in mid-lactation and in late lactation.

In all experimental groups there were increases and decreases in the treatment mean values for milk yield and solids-not-fat, protein and lactose contents, in response respectively to overfeeding and to underfeeding. The responses in lactose content in early and mid-lactation were small (overfeeding, 0·00 to +0·06 percentage units; underfeeding, −0·01 to −0·08 percentage units) and less than those in protein content, whereas in late lactation they were larger (overfeeding, +0·09 to +0·14 percentage units; underfeeding, −0·13 to −0·19 percentage units) and of the similar size to those in protein content.

The teats of quarters which were milked, and those in which milking had ceased, of 16 cows were exposed to bacteria twice daily for 2 weeks by dipping them in a suspension of streptococci and staphylococci.

Of 32 quarters in which milking had ceased 22 became infected in this period. Eighteen infections were with Streptococcus dysgalactiae, 2 with Streptococcus uberis, 1 with Staphylococcus aureus and 2 with organisms not used in the culture. Only 3 of 32 milked quarters became infected in the experimental period: 2 infections were with Str. aureus and 1 with an organism not used in the culture. Animals differed in their susceptibility to infection.

Quarters which had been unmilked for 2 weeks were as susceptible to new infectionas those in which milking had just ceased.

In this experiment, factors such as intramammary pressure, leakage of milk, concentration of bacterial inhibitors, composition of secretion and cell counts of the secretion did not appear to influence the rate of infection. Yield at drying-off, milking rates and teat patency were not related to the differences in susceptibility between animals. It is suggested that the main reason for the lower new infection rate in the milked quarters is that pathogens entering the streak canal or teat sinus are usually flushed out with the secretion during milking and do not cause infection.

A versatile continuously-stirred fermentation apparatus, suitable for batch and continuous culture studies, has been constructed from glassware normally available in the laboratory. With the aid of normal workshop facilities an effective stainless steel stirrer gland was constructed. This apparatus has been used successfully for batch and continuous culture studies on the growth of lactic streptococci in skim-milk medium.

Investigations on the growth characteristics of Streptococcus lactis ML8 in skim-milk medium have indicated that agitation of the medium may decrease the growth rate of the organism by decreasing the availability of CO2 (CO2 effect) and of soluble nitrogen (soluble nitrogen effect). The former effect was observed only when the CO2 concentration was depleted by heat sterilization; the latter occurred when a certain threshold level of agitation was exceeded. Results suggested that the soluble nitrogen affect was caused by an alteration in the rate of diffusion of substances formed in the immediate vicinity of the bacterial cell by the proteolytic enzyme system.

The behaviour of Streptococcus cremoris AM1 and AM2, and Streptococcus lactis ML8 in continuous growth in reconstituted skim-milk medium under controlled conditions of pH has been investigated.

At 20°C under aerobic conditions small quantities of oxygen were found to stimulate the growth rate of Str. lactis ML8 between pH 5·05 and 5·45. Growth between pH 5·45 and 6·35 became erratic due to the accumulation of hydrogen peroxide. Above pH 6·35 hydrogen peroxide could not be detected in the continuously growing culture and below pH 5·05 the growth rate of the culture was not affected by small quantities of oxygen. An effect of direct oxygen toxicity on growth, distinct from that of hydrogen peroxide accumulation, was observed over the limited pH range investigated. After 48 h growth in the presence of small concentrations of oxygen the effect disappeared.

At 30°C under anaerobic conditions a region in the dilution rate curves was observed with Str. cremoris AM1 and AM2 and Str. lactis ML8 which suggested the occurrence of a growth-limiting component. This region of slow growth could be correlated with diminished soluble nitrogen values, and was eliminated by enrichment of the medium with soluble nitrogen.

An unusual feature of the investigation was the dilution rates, indicative of extremely fast specific growth rates, experienced for Str. lactis ML8. It was thought unlikely that wall growth contributed appreciably to the dilution rate as a similar phenomenon was not observed with Str. cremoris AM1 or AM2.

Redox potential changes at 30°C in anaerobic reconstituted skim-milk cultures of Streptococcus cremoris AM1 and AM2 and Streptococcus lactis ML8 have been investigated in batch cultures and continuous cultures maintained at constant pH by the addition of fresh medium. Str. lactis ML8 was more strongly reducing than either Str. cremoris AM1 or AM2. The positive Eh values which can arise in dense continuous cultures of lactic streptococci are not likely to affect either the lactic acid formation or proteolytic activity of the resultant bacterial concentrate. An electrolytic reduction cell is described which allows variations in redox potential in cultures to be controlled simply and effectively.

The effects of daily supplements of 300 g native and 2 hydrogenated cod-liver oils (CLO) to the diets of lactating cows have been examined. Native CLO supplements increased the proportion of propionate and decreased the proportion of acetate in the rumen and also decreased the yields in milk of fatty acids synthesized within the mammary gland and those derived from plasma triglycerides. These effects were reduced or eliminated when the hydrogenated oils were given instead of the native CLO. With all 3 supplements about 15% of the dietary C20and C22 acids was secreted in milk.

Both native and hydrogenated CLO supplements increased the concentrations of the cholesteryl ester and phospholipid components of the α-lipoproteins in the blood plasma. These components contained most of the C20 and C22 acids of the CLO supplements that were incorporated into the blood plasma lipids. With the native CLO supplement about half of the C20 and C22 acids present in the blood plasma were polyunsaturated.

There were no differences in the effectiveness of high-density plasma lipoproteins as activators of triglyceride emulsions for lipoprotein lipase hydrolysis whether the lipoproteins were from cows receiving the native or the hydrogenated CLO supplements or the control diets. Significant rates of hydrolysis by lipoprotein lipase could not be demonstrated using as substrates low- or high-density plasma lipoproteins or activated emulsions of their extracted lipids.

Additional in vitro experiments using activated triglyceride emulsions as substrates for mammary lipoprotein lipase showed that replacement of a soybean oil emulsion by an emulsion of native CLO reduced the rate of hydrolysis considerably, whereas an emulsion of hydrogenated CLO was without effect unless it was present at levels in excess of 50%.

It is concluded that the change in rumen fermentation induced by the polyunsaturated fatty acids of CLO is implicated in the decreased intramammary synthesis of fatty acids, but that additional information is required before the reason for the reduced mammary uptake of plasma fatty acids can be established.