1. In a short-term winter feeding trial using eight Ayrshire cows, a concentrate with a nominal starch equivalent of 68 was fed at 2 lb. and 4 lb./gal. of milk with either 16 lb. hay or 16 lb. dried grass. Silage (40 lb. per day) was also fed.

2. The weight of dry matter from grass products eaten per day was similar on all four treatments. Differences in total daily dry-matter consumption were very largely attributable to differences in the consumption of concentrate dry matter.

3. The daily dry-matter consumption as a percentage of live weight increased from 2·5 on the rations that included hay and 2 lb. concentrates per gal. to 3·2 on the rations that included dried grass and 4 lb. concentrates per gal.

4. Milk yield was significantly increased from an average of 33·5 to 37·3 lb./day when the amount of concentrates fed increased from 2 to 4 lb./gal. and also significantly increased from 33·9 to 36·9 lb./day when dried grass was fed in place of hay.

5. There was an increase of 0·9 lb. milk for each 1 lb. increase of starch equivalent plus 0·2 lb. digestible crude protein intake as a result of either increasing the rate of concentrate feeding, or of feeding dried grass instead of hay.

6. The fat content of the milk was similar on all four rations, whereas the solids-not-fat varied from 8·42% on the ration that included hay +2 lb. concentrates/gal, to 8·59% on the ration that included dried grass +4 1b. concentrates/gal. This change in S.N.F. was almost entirely in the protein fraction of the milk.

7. There was a gradation in live-weight change from -0·74 lb./day on the hay +2 lb. concentrates/gal, to +0·34 lb./day on the dried grass +4 lb. concentrates/gal.

8. After a consideration of profit margins, milk quality and live-weight changes it was concluded that both treatments B (dried grass +2 lb./gal.) and C (hay +4 lb./gal.) gave optimum and almost identical results.

1. An experiment was conducted to compare the relative value for winter milk production of artificially dried cocksfoot and a mixture of timothy and meadow-fescue each cut in the spring and autumn.

2. A double balanced 4×4 Latin square design with eight Ayrshire cows and four periods each of 3 weeks' duration was used.

3. The dried grasses were fed in equal amounts (17 lb./cow/day) and equal quantities of each grass were consumed in terms of dry matter. Approximately half the total dry matter intake was grass.

4. No differences were detected between the cocksfoot and the timothy/meadow-fescue mixtures in their value for milk production.

5. With both cocksfoot and timothy/meadow-fescue the spring-grown grass gave slightly higher milk yields than the autumn-grown grasses but the differences were not statistically significant.

6. Feeding the spring-grown dried grasses led to a slightly higher solids-not-fat content in the milk than when the autumn-grown grasses were fed.

7. The feeding of the autumn grown timothy/meadow-fescue grass mixture was associated with a significant loss of live weight in the cows. It was suggested that this loss of weight was due primarily to changes in the amount of the contents of the rumen and intestines.

1. Data from seven winter feeding experiments with dairy cows, using 3×3 latinsquare designs and from two randomized block experiments on the supplementation of summer grazing, were used in a study of some aspects of experimental technique.

2. In the 3×3 trials the use of observations from 2 weeks instead of the last week of each period reduced minimum significant differences by 0·1 lb. milk/cow/day, and 0·06 and 0·02 mean percentage unit of fat and solids-not-fat, respectively.

3. In three out of four experiments no advantage was obtained from recording live weights on more than 1 day at the end of each period.

4. It was shown that the analysis of variance of milk yields expressed as individual differences between yield in the experimental period, and the mean of pre- and post- experimental control periods was considerably more efficient than that of yields in the experimental period, or covariance with yield in the initial control period.

5. The results were discussed in relation to the sources of error involved, and the relative efficiency of various types of experimental design.

1. Two levels of steaming-up and two levels of concentrates feeding during the first 84 days of lactation have been compared in a 2 x 2 factorial experiment over three winter seasons, using fifty-two Shorthorn and thirty-six Friesian heifers. The levels of steaming up were 2 cwt. and ½ cwt. of concentrates fed over the last 21 and 14 days of pregnancy, respectively. The levels of concentrates during lactation were 5 and 3 lb. per 10 lb. milk per day. Roughages were fed for maintenance. After the 84th day of lactation all animals were given the same treatment for the remainder of the lactation. Milk yield, milk composition and live weight were measured throughout the lactation.

2. The treatments were annotated as HH, HL, LH and LL, the first letter indicating level of steaming-up and the second the level of concentrates feeding in the first 12 weeks of lactation.

3. Mean milk yields were as follows:

4. The LH treatment reqiuird 3 cwt. more concentrates than HL to produce the same amount of milk.

5. The response to additioal concentrates on the LH treatment after calving was l lb. additional milk per l lb. additional starch equivalent.

6. Butterfat percentages were higher (3·69%) in both groups on low-lactation feeding than in the two groups on high-lactation feeding (3·43%). The HL group (i.e. the group that received high steaming-up and low-lactation feeding) combined high milk yield with a higher butterfat percentage, and at current milk and feedingstuff prices and under the conditions of the trials, this group showed a greater cash return than the groups on the other three treatments.

Machine milking has been studied in three cows by cineradiography at 50 frames/sec. The findings of Pier et al.(5) have been confirmed.

The effects of variation in the pulsation rate, ratio and liner tension were investigated and the results compared with those obtained by other methods. The proportion of each milking cycle during which milk flows has been determined.

1. Four hundred and sixteen individual cow samples of milk from fourteen herds were analysed for fat, total solids, crude protein, lactose and chloride.

2. Wherever possible information concerning feeding, incidence of mastitis, age of cow, stage of lactation and successful service dates was collected.

3. The milk samples were divided into normal (with 8·50% S.N.F. or over) and substandard (with less than 8·50% S.N.F.) groups.

4. The difference in S.N.F. content between the normal and substandard groups was due almost equally to a drop in lactose and protein, but with the latter playing the major part. Thus, although inadequate feeding was probably a major factor in lowering the S.N.F. in this population the fact that the lactose content was so often low was taken as indicating that mastitis was also important.

5. Breed, age of cow and stage of lactation all affected the incidence of low S.N.F. milks.

6. On average the substandard milks showed a reduced fat content as well as a lowered S.N.F. level.

1. Seven sets of monozygous twins were kept under the same open-grazing conditions for one year. Milks were analysed for fat, solids-not-fat, lactose and total protein contents, and butterfats were isolated at fortnightly intervals for estimation of iodine value, refractive index, softening point, saponification value, Reichert value, carotene content and vitamin A content. The results are presented both graphically and statistically.

2. Consistent differences in body weights of the cows of a twin set did not render the twin set unsuitable for use in experimental work on causes of changes in milk yield, fat yield, milk composition or butterfat characteristics.

3. The advantages of monozygous twins over unrelated cows was shown: (a) by the close general correspondence of the within-twin-set results as compared with the between-twin-set results throughout the period of the trial; (b) by the similarity in the forms of the curves for the different sets of twins; (c) by the marked similarity of the reaction of the cows of a twin set to change in environmental conditions—this applied particularly for butterfat characteristics such as iodine value and refractive index; (d) by the ‘twin efficiency values’ for butterfat properties computed from the whole-lactation weighted means, viz. iodine value, 7·8; saponification value, 3·4; Reichert value, 5·5; carotene content, 7·8; vitamin A content, 1·7; and vitamin A potency, 2·5.

4. There were wide differences in the degree of approach to ‘identicality’ with different sets of twins. The cows in two of the sets showed a consistently close within-twin-set relationship for all characteristics studied.

The shape of the lactation curve in guinea-pigs has been studied. The peak was found to occur between days 2 and 3.

The oxygen consumption and lactose synthesis rates were investigated in mammary gland slices and results expressed on both DNA and DW bases.

Oxygen consumption rates on a DNA basis tended to be low at the onset of lactation and high at the peak. Lactose synthesis on a DW basis tended to decline linearly from the onset of lactation, but on a DNA basis showed a peak coinciding with that of the lactation curve.

The significance of these findings has been discussed.

1. An experiment has been carried out under commercial conditions to compare the value of dried skim-milk and white fish meal as protein supplements in a basal meal fed to fattening pigs in conjunction with unrestricted amounts of whey.

2. There were four treatments. All pigs received 3 lb./pig/day of a basal meal, reduced to 2 lb. by the 13th week of life, plus unrestricted amounts of whey. The protein supplement in the four basal meals comprising the four treatments were: (1) 10% white fish meal, (2) 10% dried skim-milk, (3) 15% dried skim-milk and (4) 20% dried skim-milk.

There were five pens of nine group-fed pigs on each treatment, involving a total of 180 pigs. The pigs were on experiment from approximately 10 weeks of age to bacon weight. Comprehensive carcass measurements were made on all the pigs.

3. The results showed that the white fish meal in the basal meal fed with unrestricted amounts of whey could be replaced by dried skim-milk on a weight for weight basis (10%) without adversely affecting the performance of the pigs. Increasing the percentage of dried skim-milk to 15% appeared to have no advantage, but increasing it to 20% resulted in a slightly better performance of the pigs.

4. The results are discussed in relation to the findings in previous experiments and it was concluded that the evidence obtained so far strongly suggests that as a protein supplement, dried skim-milk is superior to white fish meal for the fattening pig, when compared on a protein content basis.

5. The commercial grading results of the pigs on all four treatments were quite good, with no very marked differences between any of the groups.

6. An average daily peak consumption of about 4¾ gal. whey/pig was reached by all groups of pigs during about the 26th–28th weeks of life. The growth curve and pattern of food consumption of the pigs during the experiment were compared with those in two earlier trials carried out under similar conditions, and possible reasons for some of the differences observed were discussed.

1. Results are reported of an experiment carried out under commercial conditions to obtain information on the comparative value of dried skim-milk and white fish meal as protein supplements for fattening pigs.

2. There were four treatments. Control pigs (i) received a meal mixture containing 10% white fish meal, reduced to 7% from 130 lb. live weight. In the other experimental groups, the fish meal was replaced by dried skim-milk, (ii) on a weight for weight basis, (iv) on a protein basis and (iii) intermediate between (ii) and (iv). All pigs were fed the meal dry and ad lib. up to a maximum of 6½ lb./head daily. Unrestricted water was available.

There were five pens of nine group-fed pigs on each treatment, involving a total of 180 pigs. The pigs were on experiment from approximately 10 weeks of age to bacon weight. Comprehensive carcass measurements were made on all the pigs.

3. The results showed that a significantly better rate of growth was obtained when the white fish meal was replaced by dried skim-milk on a weight for weight basis. Increasing the percentage of skim-milk in the diet up to 15% resulted in a slightly better performance of the pigs, but no further improvement was obtained by raising the level of skim-milk to 20%. These results were discussed in relation to an apparent set-back that the control pigs suffered at the start of the experimental period.

4. The inclusion of dried skim-milk in the diet had an adverse effect on the carcass grading, particularly in so far as it resulted in a thicker layer of back fat over the loin as compared with that in the control pigs.

5. There was also some indication that the carcasses from the animals given dried skim-milk had slightly thicker bellies and a smaller breadth of eye muscle than the carcasses of the control animals that received fish meal.

6. The results have been compared with those of an earlier trial, which was in most details the same as that reported here, the main difference being that in the earlier experiment individual feeding and one breed of pig were used.

Certain histological factors of the mammary gland and the DNA content unit per dry weight were studied at four stages of lactation in guinea-pigs.

The porosity index, the number of alveoli per field, the number of cells per field and the DNA content, showed maximal values at parturition and minimal values at the peak of lactation, whereas the opposite trend was shown by the value for modal alveolar diameter and the number of cells per modal alveolus.

The relationships between these results are discussed.

The initial rate of growth of Str. lactis and Str. cremoris in skim-milk depends upon the proportion of CO2 present in solution. When CO2 is constantly swept from the milk medium by a stream of CO2-free gas all strains show a very prolonged lag period. They eventually emerge from this lag period after more than 8 hr. and grow well enough to coagulate the milk within 24 hr. For optimal initial growth the lactic streptococci require the presence in solution in the milk of proportions of CO2 within the range 0·2 and 2·3% by volume. Some strains have a higher CO2 requirement within this range than others. Skim-milk after sterilization contains less than the optimum proportion of CO2 for many strains of lactic streptococci.

Yeast extract in a proportion of about 0·5% can substitute for CO2 in skim-milk.

The periodic inversion of cultures in plain skim-milk reduces the initial rate of growth of some of the streptococci, particularly those which form long chains. The reason for this is unknown. It does not occur when yeast extract is present.

That there are variations in the per capita consumption of milk between families has been amply demonstrated—the reports of the National Food Survey, conducted by the Ministry of Agriculture, Fisheries and Food, record them year by year(1). Some, at least, of the reasons for these variations—such as result from differences in income and in the size and composition of families—are also known. There are, in addition, variations in the per capita consumption of milk between different towns and areas of Britain, and it is the purpose of this paper to consider them.

Bacteriophage active against a lactic streptococcus was first isolated in this laboratory in February 1935. In May of the same year some batches of phage for Streptococcus cremoris strain HP were prepared by lysing the bacteria in skim-milk culture and preparing a Seitz-filtered whey as described by Whitehead & Cox(1). At that time it was not the practice to adjust the reaction of phage filtrates and the pH of the batches prepared in 1935 was not recorded. It was probably between 5·2 and 6·1 (Hunter & Whitehead (2)). Samples of four different preparations of the phage were filled into sterile tubes and sealed off in a blow-pipe flame. Two of the samples, sealed on 24 May 1935, had titres of 107 (as determined by spotting of standard loops of dilutions on a bacterial mat) and the other two, sealed on 16 July 1935 had titres of 106. The sealed tubes were stored in the refrigerator at 4–7° C.

The determination of apparent lactic acid has been used as a basis for a rejection test for raw milk. The method consists of precipitation of milk proteins with barium chloride, sodium hydroxide and zinc sulphate, and the addition of ferric chloride to the filtrate to produce the yellow colour of ferric lactate. Lactic acid is not solely responsible for the production of the yellow colour, which, nevertheless, gives a good relationship with keeping quality of milk (measured as hours to the C.O.B. end-point at 22° C.) and the values have been expressed as ‘apparent lactic acid’.

The relationship of winter and summer milks to keeping quality has been studied, and a value of 0·03% apparent lactic acid in milk is equivalent to an average keeping quality of 5¾ and 8½ hr. for winter and summer milks, respectively. A value of 0·03% is recommended as the earliest value of apparent lactic acid at which milk could be rejected.

The apparent lactic acid in colostrum and late-lactation milk and in milk from cows suffering from mastitis has been determined, and only in late-lactation milk were the values found to be significantly higher than usual in fresh raw milk, and an inverse relationship between yield and apparent lactic acid is suggested.

Permanent glass matching disks have been prepared for use in a Lovibond comparator. This permits the intensity of the yellow colour produced with 1% ferric chloride to be determined and the apparent lactic acid in milk estimated.

Grateful acknowledgement is made to the management and staff of the Dairy Department of the Reading Co-operative Society and the Farmer's Clean Milk Dairy, Reading, and local milk producers for supplying samples for experiments; to the N.M.T.S. staff in Reading for help in finding suitable farmers, and to the Dairy Husbandry Department of the N.I.R.D. for information about and samples of abnormal and late-lactation milk. Our particular thanks are due Miss Marie Gruber for technical assistance, to Dr N. J. Berridge for the suggestion and help on the work on pH change as an indication of keeping quality (given in the appendix), and to Dr A. T. R. Mattick for the advice given in this work.

Lactobacilli occur in large numbers in cheese (1), even when the milk used for cheesemaking has been pasteurized (2) and therefore contains very few or no lactobacilli. The lactobacilli in cheese must enter during cheese-making, and a potential source of these organisms is the air of the creamery. Olson & Hammer (3), although not concerned directly with lactobacilli, showed that the numbers of bacteria in the air of dairy plants fluctuated with weather conditions. Naylor & Sharpe(4) found lactobacilli in the air of the Experimental Dairy at the National Institute for Research in Dairying: some strains were identical with others isolated from cheese, but there was frequently no correlation in type of lactobacilli from the air with those from cheese. Thus, ‘cheese types’ were found in the air some months before or after the cheese was made, which seemed to indicate contamination of the air from sources within the dairy. To obtain further information regarding the lactobacillus flora of the air the following investigation was made.

A method has been described for the routine estimation of total phospholipids and of lecithin, cephalin and sphingomyelin in butter. Butter serum was treated with ether and light petroleum in a modified Röse-Gottlieb procedure to extract the phospholipids which then were estimated in the extract.

Chloroform-methanol extracted from the butter serum slightly more phosphorus than did the ether-light petroleum mixture. Ethanol-ether, benzene-ethanol and benzene-isopropanol, when used to extract a protein-phospholipid precipitate formed by the addition of trichloroacetic acid to butter serum, also removed slightly more phosphorus than did the ether-light petroleum mixture. The higher phosphorus contents of these extracts was due partly to a slightly more efficient extraction of phospholipids and partly to extraction of non-phospholipid phosphorus compounds.

Tests over a 2-year period of milk from an Ayrshire herd showed that there was a greater incidence of low S.N.F. samples from milk with a high cell count compared with samples from that of a low cell count.

While a wide variation in S.N.F. was found at any particular cell count level, there was a broad correlation between cell count and S.N.F., the average S.N.F. for cows in the same lactation being lower for those animals with high cell counts in the milk.

It is suggested that a cell count > 100,000/ml. in bulk milk should be regarded as significant.

Bacteriological examination of the high cell count samples showed that in a large proportion of cases no typical mastitis-producing organism could be isolated.

The authors wish to thank their colleagues in the Bacteriology, Chemistry and Veterinary Investigation Departments for assistance in this work.

1. Cultural and biochemical details are given of a homofermentative species of lactic acid-producing bacteria, repeatedly isolated from typical maturing New Zealand Cheddar cheese.

2. The Gram-positive cocci, occurring mainly in pairs and tetrads in acid media, are greatly stimulated in growth by the presence of yeast or tomato extracts. The organism converts glucose into inactive lactic acid and smaller amounts of acetic acid.

3. With the exceptions that the organism does not ferment trehalose and is less tolerant to low pH and high concentrations of Teepol, all its characteristics are similar to those for the genus Pediococcus, in particular to the type species P. cerevisiae Balcke.

1. The phospholipids of skim-milk, buttermilk and butter prepared from one bulk of Friesian milk were examined, using chromatographic methods. No major differences were found between the three products.

2. The phospholipids of buttermilk contained (in moles %) phosphatidylethanolamine 29, phosphatidylserine 10, phosphatidylcholine 33, sphingomyelin 19. The content of lipid-bound inositol was 5 moles % and of plasmalogens 3 moles %. Nitrogen, associated with the kephalin fraction, equivalent to 14 moles % of the lipid P was not identified. No lysoglycerophospholipids were present.

3. Unsaturated fatty acids were present in both the α'- and β-positions of the milk glycerophospholipids, in contrast to those from other animal sources which have usually been found to contain unsaturated acids in the α'-position and predominantly saturated acids in the β-position.

The authors are indebted to Prof. H. D. Kay for providing facilities at the National Institute for Research in Dairying, Reading; to Dr R. H. Smith for advice on processing, and to Miss H. R. Chapman and the staff of the Experimental Dairy for carrying out the dairy work.

This work forms part of the programme of the Food Investigation Organization of the Department of Scientific and Industrial Research. Mr P. Whiteley contributed to the experimental work.