Ewes' milk from two breeds of Greek sheep, Karagouniki and Serron, was analysed throughout mid and late lactation to investigate the salt distribution between colloidal and soluble phases. Separation of the phases was obtained by centrifugation at 48000 g. Effect of stage of lactation was significant only for total Mg in both breeds (P < 0·05) and for colloidal citrate in Serron ewe milk (P < 0·01). Breed differences were not significant (P < 0·05), except for colloidal casein.

In the colloidal phase there was a strong positive correlation of casein with Ca and P, but correlation with Mg was poor. Molar sums of [Ca + Mg] and [P + Cit] were also significantly correlated. The Ca to P molar ratio averaged 2·00 and was higher than that reported for cows' milk. The inclusion of Mg as well as citrate in the colloidal calcium phosphate is indicated. In the aqueous phase there was a strong negative correlation of Ca with pH and a positive one with citrate. Correlation with phosphate was less. It seems that citrate is the main factor influencing Ca concentration in the soluble phase.

The chymosin-initiated coagulation of casein micelles was followed by monitoring light transmission using a Platelet Aggregometer. The release of macropeptide by chymosin was monitored using fluorescamine. The lag period in the clotting reaction was proportional to clotting time and the reciprocal of enzyme concentration. The average rate of coagulation, which was approximately equal to the reciprocal of clotting time (Tc), increased in proportion to enzyme concentration at low enzyme concentrations and reached a limiting value at high enzyme concentrations. The percentage hydrolysis at the Tc was 47 ± 5% in the presence of 20 mM-CaCl2 and it was calculated that a 5-fold decrease in the speed of the enzyme-catalysed reaction would decrease this value at the Tc to 43 ± 5%. The possible uses and limitations of the Platelet Aggregometer for determining the influence of the chemical environment on the velocity of the chymosin-catalysed reaction and para-casein micelle aggregatability are discussed.

Coagulation of para-casein micelles was monitored using a Platelet Aggregometer at 37·8 °C and pH 6·7. The macropeptide released by chymosin was determined quantitatively with fluorescamine. The aggregatability of para-casein micelles integrated over the complete hydrolysis of κ-casein was calculated from the amount of κ-casein hydrolysed during the clotting time (Tc). Low concentrations of CaCl2 enhanced the rate of κ-casein hydrolysis but increases in CaCl2 concentration above 8 mM caused marked decreases in this rate. Calcium chloride enhanced the ability of para-casein micelles to aggregate. Little aggregation of para-casein micelles occurred at 0·6 mM-CaCl2 even after all of the κ-casein was hydrolysed. As the concentration of CaCl2 was increased from 3 to 60 mM, aggregation of para-casein micelles took place at progressively lower levels of κ-casein hydrolysis and the percentage hydrolysis at the Tc decreased markedly from 71±7% to 26±8%. The combined influence of CaCl2 on the velocity of the chymosin-catalysed reaction and on para-casein micelle aggregatability accounted for its effect on the average rate of coagulation (calculated by the reciprocal of the Tc). Results are consistent with the hypothesis that electrostatic repulsions and ionic bonding are involved in the interaction between chymosin and κ-casein.

The effect of varying the casein composition of artificial micelle milk on rennet coagulation time and syneresis was examined in order to determine whether either of these processes is dependent on the concentration of particular casein components. It was found that the levels of κ-and β-caseins had a significant effect on coagulation, whereas syneresis was only affected by the level of β-casein. Partial dephosphorylation of preformed micelles or the incorporation of dephosphorylated or partly dephosphorylated β-casein into artificial micelle milk was found to have an adverse effect on both coagulation and syneresis. It was concluded that the phosphate groups of casein, particularly those of β-casein, are directly involved in the micelle-micelle interactions which occur during coagulation and syneresis.

For milk from four herds of cows, maintained under different conditions of feeding and management, the natural Cu content and the stability of the ascorbate were highly correlated. Low Cu levels in milk from cows at pasture at farms A and B during the summer were associated with low storage losses of ascorbate. During this period, the milk of cows at farms C and D (on forage and Cu-supplemented concentrate) was richer in Cu, and losses of ascorbate were high. Heat treatment of the milk stabilized the ascorbate. Thus, in ‘high Cu’ milk (60 µg Cu/l), loss of ascorbate in the raw milk was 58% at 2 d, as against 17% after pasteurization at 72 °C and no loss after treatment at 82 °C. Storage of milk in light caused rapid destruction of ascorbate, equally with 72 and 82 °C heat treatments. The effects were examined of milk pasteurization temperature (72–82 °C) on flavour stability, bacteriological quality and vitamins of the B-complex. Heat treatment at 82 °C increased the susceptibility of vitamin B12 to destruction by light, but otherwise caused no greater losses of B-complex vitamins than did treatment at 72 °C. Taste panel ratings showed an initial preference for milk heated at 72 °C, but on storage of this milk in darkness the flavour score fell progressively and at 5 d it was judged ‘stale’. Treatment at 82 °C gave a faint ‘cooked’ flavour although, unlike that of the 72 °C-treated milk, the flavour remained stable throughout 14-d storage and after d 8 was increasingly preferred. On exposure to light after treatment at 72 °C milks rapidly acquired an unpleasant‘oxidized’ flavour, but after treatment at 82 °C, exposure to light had no such adverse effect on flavour during the early days of storage. Pasteurization at ∼ 80 °C offers a potential for improvement in the oxidative stability of milk and its contribution of vitamin C to the diet.

Milk samples from three groups of cows were taken at frequent intervals throughout lactation following autumn-, winter- or spring-calving. The ethanol (EtOH) stability/pH profile was determined for each sample and its characteristic parameters calculated. The lactational trends in these parameters were examined. Asymptotic maximum EtOH stability (Smax) was low in early lactation but rose rapidly to a value which showed no further lactational trends. Asymptotic minimum stability (Smin) for samples from autumn- and winter-calving cows showed a decrease which could be associated with the transition to summer grazing but no obvious lactational effects. The slope parameter increased slowly during lactation. The profile pK value decreased in early lactation, but thereafter increased throughout lactation giving the most obvious effect observed in direct measurement, namely an alkaline shift in the profile as lactation progressed. The EtOH stability calculated at a fixed pH of 6·6 passed through a maximum, characteristic for each cow, in the first weeks of lactation but declined steadily thereafter. This behaviour mirrors the lactational behaviour of the soluble salt balance ratio calculated from the original data of White & Davies (1958).

A sensitive, specific radioimmunoassay for bovine α-lactalbumin (α-la) suitable for measurement of serum, tissue culture media and milk was developed. α-La was not detected in serum from prepubertal male or female cattle, but was detected as early as 60 d of gestation in nulliparous Holstein heifers, the level being greatly increased during the last one third of gestation. Serum from cross bred beef heifers contained less α-la and it was not detected until late gestation. Concentrations of α-la in serum samples from pregnant multiparous Holstein cows decreased at drying-off and subsequently increased just before parturition. Secretion of α-la by mammary tissue explants from steroid-primed prepubertal Holstein heifers was induced by the addition of bovine prolactin, ovine prolactin or human growth hormone to tissue culture media.

Bovine whole milk was concentrated up to 5-fold by ultrafiltration and separated conventionally to produce protein enriched creams with fat contents ranging from 20 to 50 % (w/w). The measured viscosity of the concentrates increased with the degree of concentration. This led to inefficient centrifugal separation, as measured by the loss of fat into the skimmed milk, which was further worsened when damage to the fat globules occurred during ultrafiltration. The viscosity of the creams, at constant fat percentage, increased as the concentration factor of the milk from which it was separated rose.

Three groups of coryneform bacteria isolated from the rind of nine soft and semi-hard cheese varieties, i.e. the orange, white and yellow strains, were studied with regard to morphological and physiological features. The 115 orange isolates of the Brevibacterium linens type formed only one homogeneous phenon. The 96 white and 161 yellow pigmented strains were grouped into 15 phena. Five phena (58% of the strains) belong to the species ‘B. ammoniagenes’, Arthrobacter variabilis and A. nicotianae. The other strains were arranged into five Arthrobacter spp. and four Corynebacterium spp. groups as well as a Rhodococcus sp. group. Usually specific forms for the individual cheese varieties were involved. In cheeses of poor surface-smear quality B. linens was found in reduced numbers.

A typing scheme was used to test 15 strains of Streptococcus uberis according to their production of (P-type) and sensitivity to (S-type) bacteriocin-like inhibitory substances. Twelve of the strains were inhibitor producers and nine different P-types were detected. All of the strains were typable according to inhibitor sensitivity, ten different S-types being distinguished. Both the P-type and S-type designations of the strains were reproducible on repeated testing. By combination of P-typing and S-typing, highly discriminatory inhibitor ‘fingerprints’ of the strains could be obtained. This scheme would appear to have considerable potential for typing isolates of Str. uberis as an aid to investigations into the epidemiology of Str. uberis mastitis in dairy cattle.

Pseudomonas fluorescens NCDO 2085 produced a single heat-stable extracellular proteinase in Na caseinate medium at 20 °C and pH 7·0. The proteinase was purified to electrophoretic homogeneity using chromatofocusing, gel filtration and ion-exchange chromatography. The purification procedure resulted in a 158-fold increase in the specific activity and a yield of 3·5% of the original activity. The enzyme is a metalloproteinase containing Zn and Ca, with an isoelectric point at 5·40±0·05 and a mol. wt of 40200±2100. It is heat-stable having D-values at 74 and 140 °C of 1·6 and 1·0 min respectively; 40 and 70% of the original activity remained after HTST (74 °C/17 s) and ultra high temperature (140°C/4 s) treatments respectively. The amino acid composition of the proteinase was determined and compared with those from other Pseudomonas spp.

Deproteinated whey, prepared from sweet whey powder, was digested anaerobically in a downflow fixed-bed reactor at 35 °C. Different pH control treatments were applied over a period of 210 d. Initially NaOH was added at a concentration of 68 mequiv. l-1 substrate. This was successfully replaced by 80 mequiv. Na2CO3l-1 which resulted in a 15·5% increase in biogas production and a 6·7% increase in CH4 production. A decrease in the quantity of Na2C03l-1 added led to signs of instability and a drop in CH4 production. The stepwise addition of urea restored the stability, and a further increase of urea, to a level of about 20 mequiv. l-1, made it possible to replace Na2C03 completely. When all pH control measures were omitted, complete failure resulted within 10 d. Methane yields, under stable conditions, varied between 0·321 and 0·272 m3 kg-1 chemical oxygen demand (COD) removed which corresponds to 92 and 78% of the theoretical yields. Carbon balances indicated that between 60·8 and 68·5% of the carbon fed was converted to biogas. No toxicity was observed at a carbon: nitrogen (C/N) ratio of 7·7. This was in contrast to previous results where toxicity had occurred at the same C/N ratio but at a higher concentration of Na.