TwO Latin-square experiments using lactating Jersey cows have been made to determine the response to 4 levels of acetic acid given as a supplement to diets fed at Woodman's Standards. The response in milk yield and composition to daily supplementary infusions of acetic acid is curvilinear, the point of inflexion occurring between the levels of 450 and 600 ml acetic acid/d. The direction of responses to the 150-ml level of infusion varied between the experiments. At the 600-ml level of addition the non-significant increases in milk fat, protein and lactose yield led to a significant increase in solids corrected milk yield. The effects of acetic acid infusion on the composition of the milk fat were determined.

Possible mechanisms by which the degree of acetylation of hepatic CoA might be regulated were examined in lactating and in non-lactating dairy cows. This involved the measurement of the hepatic steady-state concentrations of free CoA, acetyl-CoA, free carnitine, acetylcarnitine and total acid-soluble carnitine in freezeclamped biopsy samples, the measurement of the hepatic release of acetate from acetylcarnitine in vitro, and the measurement of the rate of hepatic output of acetate and ketone bodies in vivo.

The hepatic ratio of [free CoA]/[acetyl-CoA] was 0·11 in lactating cows and 0·59 in non-lactating cows. There was a significant rise in the hepatic concentration of acetyl-CoA and a significant fall in that of total acid-soluble carnitine in the lactating cows as compared with the non-lactating cows.

There was a net in vivo output of acetate and of ketone bodies from the livers of both lactating and non-lactating cows. The rates of output amounted to 7·77 and 2·00 mmole/min for acetate and ketone bodies respectively in the lactating cows, and 4·95 and 2·14 mmole/min in the non-lactating cows.

The rate of enzymic release of acetate from acetylcarnitine in vitro amounted to 0·93 and 0·34 μmole min−1 (g wet wt)−1 at 37 °C for liver homogenates derived from lactating and non-lactating cows respectively. The activity observed in the case of the lactating cows was sufficient to account for the rate of hepatic acetate output observed in vivo.

Milk samples from 3 groups of cows were taken at regular intervals throughout lactations following 2 spring and 1 autumn calving. The average radius of casein micelles in milk was determined either from the wavelength dependence of turbidity or from the diffusion coefficient as measured by a quasi-elastic light scattering method. A 2-fold variation in naturally occurring average radii was found both in milks from the same cow at different times and between milks from different cows at the same time.

Samples of bulk silo milk from 5 creameries in the south west of Scotland were taken over a 16-month period. The average radii of casein micelles were determined by measuring the wavelength dependence of the turbidity of skim-milk diluted with its own ultrafiltrate. In addition, the concentrations of casein, soluble and colloidal Ca and P1 were measured. The average size of casein micelles followed a pronounced seasonal trend with smaller average sizes in the summer compared to the winter period, confirming earlier observations made on milks from individual cows. The average size of casein micelles correlated positively (r = +0·777, P < 0·001) with the amount of colloidal P1 per unit weight of casein and negatively (r = −0·77, P < 0·001) with casein-bound Ca. Average micelle radius was negatively correlated with heat stability at the natural pH of milk (r = −0·61, P < 0·001), but there did not appear to be any causal relation between these 2 variables.

Bovine casein micelles in pre- and post-partum colostrum have been characterized by both electron microscopy and light scattering. Attempts have been made to relate the changes in average micelle size to the partitioning of Ca and inorganic orthophosphate betwen the colloidal and aqueous phases. Pre-partum colostrum, and to a decreasing extent post-partum colostrum, contain casein micelles several microns in diameter. These giant micelles often contain closed cavities suggesting that they are formed by aggregation of particles comparable in size to normal casein micelles. Electron microscopy of mammary gland alveoli from nearterm pregnant heifers shows clear differences in average size between micelles immediately before and after secretion from Golgi vesicles. It appears that micelles aggregate in the alveolar lumen, possibly as a result of continuing co-precipitation of casein with Ca phosphate.

An improved method for preparing β-lactose crystals was developed by heating a supersaturated solution of lactose under reflux. The product obtained by this method yielded large well-defined crystals.

Differential thermal analysis was used for the characterization of β-lactose. The thermograms of β-lactose obtained when heating conditions were similar to those used in the conventional measurement of the melting point gave an endothermic peak at 229·5–230 °C. A marked difference was observed between the peak temperature on the thermogram and the melting point (m.p.) of β-lactose (252–252·2 °C) given in the literature. The method for determining the m.p. of β-lactose was reexamined in detail and previously reported values could not be confirmed. By application of the conventional method for determining m.p. of organic substances, 229·5 °C (dec.) is proposed as the m.p. of β-lactose.

The amount of damage caused to the milk-fat globule by mechanical agitation of raw milk was assessed by several methods and related to the level of induced lipolysis resulting from the agitation. The results obtained by measuring the amount of ‘free’ fat formed, the percentage of fat in the skim-milk phase after separation, and the increase in the proportion of alkaline phosphatase associated with the skim-milk phase, showed little correlation with the level of lipolysis induced by agitation over a range of conditions. However, the level of induced lipolysis showed good correlation with the amount of lipolysis caused by addition of an exogenous lipase (from Candida cylindraceae) to the agitated milk. It is concluded that classical methods of assessing milk-fat globule damage are unsuitable for predicting the amount of fat available for lipase action and the lipolytic potential of raw milk subjected to mechanical agitation.

The retention of the protein nutritional value during storage of lactosehydrolysed dried milk at different water activities (aw) was studied and compared with that of ordinary milk. In the lactose-hydrolysed milk biologically available lysine decreased much more rapidly than in the ordinary milk at all the different aw studied. Thus, at conditions normally accepted for ordinary dried milk (aw ~0·2; moisture ~ 4 %) there was an available lysine loss in the hydrolysed milk of about 25 % after 2 months and about 40 % after 6-months storage at room temperature. This occurred without any visible browning. It is concluded that drying to very low aw (≦ 0·11) is necessary to obtain good stability of the protein nutritional value (PNV) in lactose-hydrolysed milk. The drying, however, must be done so that losses in nutritional value are minimal during that process. It should also be noted that fat oxidation might be a problem at such low aw. A chemical method for available lysine assay (guanidination and assay of homoarginine) gave values in good agreement with the biological evaluations with rats.

The effect of heat treatment of milk on low molecular weight, volatile compounds was studied in order to relate changes in the flavour of milks to changes in chemical composition. Milks were heat treated in a UHT plant for 3 or 90 s at 140 °C and stored at ambient temperature for periods up to 112 d. Volatile compounds in raw milk and in heated milks were isolated by a low temperature spray distillation technique and identified using gas chromatography and mass spectrometry. Cabbagey defects in heated milks are correlated with total volatile sulphur and it is concluded that the compounds H2S, COS, CH3SH, CS2 and (CH3)2S could be responsible for this defect.

Separated milk was treated with rennet at 30 °C for known proportions of the visually-determined clotting time. Samples were fixed with glutaraldehyde, which inhibited aggregation of casein micelles, mixed with agar, and stained sections examined. Bias in the preparation of the electron micrographs was minimized by replication, particularly in the early stages of sample preparation, and, for thin (60–90 nm) sections, by systematic selection of the areas examined.

Whole casein micelles, viewed in stereo micrographs of thick (about 1 μm) sections, had open, diffuse structures and their appearance did not change on treatment with rennet. Micrographs of thick sections taken at 90° to each other showed that casein micelles form a network-type of gel, with the same structure in all dimensions. Counts of micelles larger than about 20 nm diam. in fields of thin sections showed that each of 3 sets of fields was similar and representative for each sample. The number of micelles/μm2 appeared to increase between 60 and 130 % of the clotting time, but this change was not statistically significant.

The aggregation of casein micelles in milk by the action of rennet, examined by electron microscopy and viscometry, was correlated with visually-observed coagulation. The spatial relationships between micelles were quantified from measurements made on micrographs. Aggregation did not start until about 60% of the rennet-clotting time, when the enzymic action was almost complete. The micelles tended to form chains which then linked into a loose network. This became more extensive with time as the average size of the aggregates increased. By the clotting time, most micelles were in contact with another and a network was beginning to form. Initially, micelles probably linked by bridges. These then appeared to contract, bringing the particles into contact and eventually causing partial fusion.

Whole cells of Streptococcus lactis C10, when incubated with an energy source, converted fumarate to succinate and malate to lactate. Cell-free extracts of Str. lactis C10 contained fumarate reductase, but no aspartase, adenylosuccinate synthetase and lyase or argininosuccinate synthetase and lyase activity could be detected. Cells grown in the presence of [14C]bicarbonate produced labelled succinate during the synthesis of purine bases. However, the amount of succinate produced by this pathway only accounted for approximately one-sixth of the succinate produced by the cells.

The fixation of [14C]bicarbonate into aspartate by Streptococcus lactis C10 was achieved by the combined reactions of pyruvate carboxylase (E.C. 6.4.1.1) and glutamate-oxaloacetate transaminase (E.C. 2.6.1.1). The pyruvate carboxylase from Str. lactis C10, which was most active at pH 8·0, was activated by the divalent metal ions Mn2+, Mg2+ and Co2+, and inhibited by sulphydryl reagents. The enzyme was inhibited non-competitively by aspartic acid and competitively by oxaloacetate.

Discovery of an endopeptidase by gel chromatography and separation of 3 exopeptidases (a dipeptidase, an aminopeptidase and a specific carboxypeptidase) from Lactobacillus casei NCDO 151 by affinity chromatography is described. The 3 exopeptidases were strongly inhibited by the metal chelators EDTA and 1,10-phenanthroline but were reactivated with Co2+ and Mn2+. The pH optima for aminopeptidase, dipeptidase and carboxypeptidase activities were 6·5, 7·6 and 7·2, respectively. Maximum activity was obtained at 45 °C for the aminopeptidase, at 30 °C for the dipeptidase and at 40 °C for the carboxypeptidase.

The substrate specificities of the 3 enzymes were also studied. The properties of these 3 enzymes are compared with those of other bacteria.

A microcalorimetric method has been developed for the identification of 9 representative strains of lactic acid bacteria used in the dairy industry. A chemically defined medium containing several carbohydrates gave reproducible and characteristic growth patterns. The technique is proposed as a rapid alternative method for the identification of bacteria selected for commercial use.

Cheddar cheese was produced free of non-starter bacteria, acidified with starter or glucono-δ-lactone and containing active coagulant (chymosin or pepsin) or inactivated coagulant (pepsin). The level and type of proteolysis in the experimental cheeses was monitored by protein solubility at pH 4·6 and in 12 % TCA, polyacrylamide gel and high voltage paper electrophoresis, gel filtration and paper chromatography. The results show that the coagulant was primarily responsible for the formation of large peptides while small peptides and free amino acids were produced principally by the starter, possibly from coagulant-produced peptides.

Cheddar cheeses were pressed from a batch of curd each month over a period of 1 year to study the production of headspace volatiles and the development of flavour in ripening cheese. At 2-monthly intervals, samples taken from selected sites in the cheese were presented to a trained panel for assessment of quality and intensity of Cheddar flavour. Using a headspace technique samples of vapours from bore holes in different parts of the cheese were analysed for H2S, methanethiol, acetone, butanone, methanol, ethanol and 2-pentanone. Most of the compounds increased in concentration during the ripening period, but ethanol and butanone varied in concentration in an unsystematic way. Possible precursors of the headspace volatiles are discussed.

The effect of added gastric lipases on the ripening and flavour development in Ras cheese was studied. Ras cheese was made from pasteurized milk with the addition of Italase, Capalase K or Capalase KL at the rate of 12 g/100 kg milk before renneting and analysed at different intervals for fat, moisture, salt, pH, soluble tyrosine and tryptophan and Shilovich ripening index. The organoleptic properties of the cheeses were also assessed. The addition of lipases was found to accelerate the development of flavour, particularly when Capalase K was used. These lipases had no effect on the gross composition of cheese during ripening.