Whole milk from cows was concentrated 2-fold by reverse osmosis in a small pilot scale apparatus. All the main constituents of the milk were retained in the concentrate.

The experiment showed that protein was a major cause of blocking of the membrane. Laboratory trials were therefore made on the ultrafiltration of milk to remove the protein with the aim of developing a 2-stage process for the concentration of milk by ultrafiltration followed by reverse osmosis with subsequent recombination of the concentrates from the 2 operations. The rates of both processes and the retentions of the constituents of milk by the membranes were measured.

Thirty-six lactating cows were used in a 3×2 factorial design experiment to study the effects of level of energy and fibre intake on milk yield and composition over an 8-week experimental period. Three levels of fibre intake — 1g of crude fibre per 25, 17·5 and 10 kcal of estimated metabolizable energy (ME) requirements — were given in diets supplying both 80 and 100% of energy requirements. The level of energy intake significantly affected milk yield, milk energy output and the percentage solids-not-fat (SNF) and protein in the milk. The effects of fibre intake on both milk yield and milk energy output were marked but not significant. Fibre intake had a significant curvilinear effect on both the SNF and protein in the milk with the highest fibre intake resulting in a significant decrease in both fractions. A significant linear decrease was obtained in the lactose fraction as the fibre intake increased.

The total volatile fatty acid (VFA) level in the rumen was significantly affected by both energy and fibre intake. The level of energy intake only significantly affected the proportion of propionic acid. Fibre intake significantly affected the proportion of both acetic and butyric acids resulting in mean proportions of acetic acid of 66, 70 and 72 at the low, medium and high fibre intakes. The correlations between the rumen acids and lactation efficiencies are also presented.

Multiple regression analysis within each fibre level has been used to partition the ME available for production between that used for milk energy output and liveweight change. The results showed efficiencies of utilization of ME for milk output of 66, 65 and 56 on the low, medium and high fibre diets respectively.

Nitrogen balance data are presented.

When the pH of milk was varied within the range 7·1 to 6·3 by addition of acid or alkali or through bacterial action, the reflectance of the milk after subsequent ultra-heat-treatment (UHT) was at a maximum of about pH 6·70. Below this value the reflectance dropped rapidly with decrease in pH. The cause of this decrease was the development of increasing amounts of sediment in the product. At pH 6·4–6·5, at least 90% of the casein and 40% of the whey proteins had been precipitated. The addition of 0·1% sodium di-hydrogen phosphate or 0·1% sodium citrate to the raw milk prevented the formation of the sediment. The role of calcium appeared important as small additions of calcium chloride or EDTA altered the patterns of sediment formation and reflectance with changing pH. Addition of 0·3% EDTA prevented sediment formation as the pH dropped.

The distribution of protein nitrogen was determined in milk treated by an experimental ultra-high-temperature (UHT) plant, operating alternatively as an indirect or as a direct heating plant.

The extent of denaturation of β-lactoglobulin can be used to assess the relative severity of UHT heat treatments; this criterion was used to compare the indirect and direct arrangements of the plant with each other and with other published results.

The time-temperature profiles of the plant were used to calculate the expected extent of denaturation of β-lactoglobulin during treatment of the milk. The results of the calculations are in fair agreement with the analytical results.

The protein, phospholipid and neutral lipid composition of a deoxycholatesoluble fraction (DOCM) from bovine milk fat globule membrane (FGM) was determined, at intervals during lactation, in 3 cows maintained at a constant level of food intake. It was suggested that the appearance of free fat in washed cream during the first 2 days post partum was related to membrane stability in colostral secretion. Differences between the cows in DOCM yield (mg/100 g cream lipid) and composition were greatest during the first 25 days post partum. DOCM yield increased markedly as lactation proceeded in one animal, and this was thought to be due to a decrease in globule size. Variation in DOCM composition was principally due to an alteration in neutral lipid content. There were few differences, however, between the quantities of phospholipid and protein in DOCM for all 3 animals, and similarly the percentage composition of the major DOCM phospholipids varied little, phosphatidyl choline being the predominant phospholipid.