1. During the first 2 weeks and last 2 weeks of a 9-week period nine cows, milked by machine, were prepared for milking by hosing and washing the udders with cold water 15 min. to 1 hr. before milking, while in the intervening 5 weeks the udders only were washed with water at 115 to 120° F. immediately before milking.

2. The hot-water routine had no pronounced effect on the yield and quality of the milk from the majority of the cows, although one reacted unfavourably to a change back to the cold water routine and had to be stripped by hand.

3. The gradual restriction, over an interval of 4–6 weeks, in the milking time from 100 to 60% of the previous normal flow period for three cows in mid-lactation had no great influence on the maximum rate of milk flow from the udder, even when hot-water washing of the udder immediately before milking was introduced late in the experiment. The maximum rate of flow was, however, reached slightly sooner after putting on the teat cups.

4. The restriction in milking time caused only a small reduction in yield as the flow during the latter part of the flow period had in any case been small.

5. The fat content of the milk of the cows whose milking time was restricted fell somewhat while the solids-not-fat content remained steady.

6. Washing the udder with hot water or restricting milking time showed no clear-cut effect on the incidence of mastitis.

It has recently been claimed that in modern high temperature-short time pasteurization plant fluctuations in temperature of the order of 1° F./sec. are unusual and probably artefacts, and that an instantaneous drop is certainly fictitious.

It has, nevertheless, been shown that such rapid drops in temperature do in fact occur frequently in high temperature-short time plants of the type investigated. The plant investigated was not of the latest design incorporating devices for speeding up the control by injecting steam directly into the hot-water pipe.

Temperature changes of the order of those in question were first observed by Mattick & Hiscox(1) of the National Institute for Research in Dairying, who carried out tests on pasteurization plant using a small mirror galvanometer with a very short time constant. The maximum rates of change were, however, not observed.

Tinned trays were artificially infected with Staphylococcus aureus in diluted milk and the dried surfaces disinfected by rinsing with sodium hypochlorite solutions.

Under the conditions described, which we believe approximate those on many dairy utensils which have been imperfectly cleaned, the pH of hypochlorite solutions was found not to be a very important factor in determining their germicidal efficiency. At pH 10 and 11, the solutions were equal to or slightly more efficient germicides than solutions at pH 7.

In the destruction of 99% of the organisms on the metal trays, a solution at pH 10, containing 200 p.p.m. available chlorine was about 7 times as rapid as one with 25 p.p.m. available chlorine.

The application of the Schibsted fat aldehyde test for the routine examination of butter and butterfat is facilitated by the following modifications:

(1) Storage of the reagent at cool temperatures.

(2) Use of sodium metabisulphite instead of sulphur dioxide.

(3) Preparation of the standard colour at a lower pH

The value of the phosphatase reaction for controlling the heat treatment of milk is unquestioned, but the accepted method of estimating phosphatase activity, namely, the colorimetric method of Kay using the phenol reagent of Folin, has three disadvantages:

(1) The number of estimations that can be performed in a given time is limited.

(2) Errors in matching the tints may occur.

(3) Samples of milk may be accidentally contaminated with traces of phenol, and this will lead to too high a Lovibond value.

The addition of 10% of sodium chloride to milk caused an average drop in pH of 0·42 unit. Different workers have found very variable relations between the pH of cream and the pH of salted butter churned from it. When 10% of sodium chloride is added to 0·2% solutions of sodium citrate or sodium phosphate there is a drop in pH of the order of 1 unit. When an alkaline solid phase is present as in the serum of neutralized cream and butter, and in milk, this change in pH may be diminished or reversed, possibly by increased solubility of the alkaline precipitate in sodium chloride solution. Factors are listed which will cause the pH of salted butter to differ from that of the cream from which it was made. In the churning of alkaline creams there occurs in the butter an increase in the proportion to the serum water of calcium phosphate and of other solids-not-fat.

1. A method is described based on wet-ashing in micro-Kjeldahl flasks for the accurate estimation of iron and copper in 10g. samples of butter or dry butterfat or in 5g. samples of cream.

2. An acid extraction method for the routine estimation of copper and iron in dry butterfat is described. This method is shown to give a reliable estimate of the quantities of the metals present.

3. A simplified method for the estimation of iron in cream and butter is described in which the iron is first freed from its protein complex by the addition of thioglycollic acid. Copper is estimated by a similar method, but without the addition of a reducing agent. Both methods give results agreeing reasonably well with the wet-ashing method and are suitable for the routine estimation of these metals in cream and butter.

1. Chemical estimations of carotene and vitamin A in New Zealand butters from two of the principal butter-making districts of the North Island during the seasons 1935–6 show that the total vitamin A potency was fairly high.

2. There were seasonal variations apparently due to nutritional rather than physiological causes. The minimum values (33–37 i.u./g. butterfat) for total vitamin A were found in late summer (February) at the time when the pasture normally tends to dry up, while the peak values (42–53 i.u./g. butterfat) occurred in late winter and spring (July-October). The variations in vitamin A potency with season were in the opposite direction to the variations recorded in the literature for Europe and America. The difference is no doubt due to the practice of stall-feeding in these countries in contrast with the all-the-year-round grazing commonly practised on dairy farms in New Zealand. The spring flush of grass is also later in the season in Europe than in New Zealand.

3. The more deeply-coloured Jersey butterfat was only slightly richer in total vitamin A potency than Friesian butterfat.

4. Contents of carotene and vitamin A in the fat of colostrum were very high, but reached normal values within 4 or 5 days after parturition.

A Gram-positive, aroma bacterium, the so-called ‘malt coccus’, has been isolated from malty-flavoured starters. It coagulates milk most readily at 35–37° C. The ‘malt coccus’ differs from the aroma bacteria mainly in that it forms acetaldehyde from pyruvic acid, evidently through carboxylase. Acetaldehyde is the chief factor in the formation of malty flavour. The ‘malt coccus’ sours milk feebly. If starter is heavily contaminated with the ‘malt coccus’, the acidity (SH°) is reduced but the aroma reaction (V.-P.) increases. The malty flavour persists in the starter during transfers and becomes more marked with rising temperature. The ‘malt coccus’ retains its aroma-forming ability for long periods in the laboratory.

1. A simple method of determining the acid degree of cheese fat has been described.

2. An average value of fat acidity on a series of thirty-eight samples of experimental Canadian first-grade Cheddar cheese, covering a period of about a year, has been eatablished at 1·0 ml. N/10 NaOH per 10 g. for cheese 2–4 weeks old.

3. An average acid degree for four experimental cheese from pasteurized milk was found to be 0·6.

4. A high acid degree was obtained on cheese fat made from milk in which lipase action was induced by vigorous agitation; the longer the period of agitation the higher was the acid degree. Extensive lipolysis took place at some stage before the cheese was removed from the press. Subsequent increase in acid degree was slight.

5. Fat acidities on a series of commercial Cheddar cheese, selected so as to include a high proportion of cheese degraded for unclean and rancid flavour, ranged from 1·2 to 4·6 with two exceptionally high values of 24·8 and 26·5. The incidence of undergrade cheese increased as the acid degree increased. All cheese with an acid degree of 2·8 or higher were below first grade on flavour.

6. Cheese made from milk, in which acid was developed by the addition of a small amount of starter the night before and which was partly homogenized, showed a high acid degree although the flavour was quite good—often first grade. Similar cheese, but with 3% starter added in the morning, showed only slightly higher fat acidity and scored lower for flavour. These findings cannot be explained as a simple inhibition of the action of milk lipase by acid development.

1. Three experimental starters were prepared by mixing three unrelated phaging strains of Str. cremoris. The corresponding phages were used to determine whether changes occurred in the relative numbers of these strains on daily subculture.

The results indicated that (a) in a multiple-strain starter one strain may overgrow the others and become responsible for most of the acid produced by that starter about 1 week after mixing the strains. The starter may soon become virtually a single-strain starter. (b) Even during the day's making, one of three strains, mixed together at the time of introduction into the cheese vat, may partly overgrow the other two strains, (c) The dominance of a particular strain in a starter mixture may be affected by the temperature of propagation.

Bacteriophage infection of the starter streptococci, which commonly causes ‘slowness’ in cheese manufacture, may arise from whey infection of milk cans used for the cartage of cheese whey from factory to farm. Thus ‘slowness’ can occur even when the starter itself and all the factory utensils are free from infection. The pasteurizing treatment which the cheese milk receives is not sufficient to destroy the traces of phage derived from infection of the milk cans.

Three methods for eliminating this trouble were investigated:

(i) Pasteurization of the whey at 190° F. for a few seconds immediately after removal from the cheese vat.

(ii) Treatment of the whey with chloride of lime.

(iii) The use of a series of single-strain starters in daily rotation in the cheese factory.

All three methods proved practicable, although some mechanical difficulties were encountered in the particular pasteurization system which was tried. The third method is at present undergoing an extensive trial on a commercial scale. It is preferred in the first place because it is costless, whereas the first two methods would involve a regular additional cost in steam of chloride of lime.

An apparatus has been developed for the routine estimation of oxygen in the headspace gas of tins of dried milk. The sampling device is an integral part of the apparatus, which is constructed mainly of metal and is simple to operate. As many as twenty samples may be taken and analysed in an hour.

Our thanks are due to the Drawing Office and Workshop staffs of the Division of Industrial Chemistry, Council for Scientific and Industrial Research, who designed and manufactured the equipment, and to the Maffra Co-operative Milk Products Co. Ltd., who permitted the apparatus to be tested in their laboratories under factory conditions.

1. Storage tests have been made independently in two laboratories on four samples of spray-dried whole-milk powder prepared on a Gray-Jensen plant from one batch of milk using pre-heating temperatures of 160 and 190° F. with and without the addition of 0·06–0·08% of ethyl gallate as antioxidant. The powders were packed in lacquered and in plain tinplate containers and stored at 47, 37 and 15° C, and at room temperature. Deterioration was followed by tasting tests, by determination of the amount of oxygen absorbed by the powders and by estimation of peroxide in the fat.

2. Raising the pre-heating temperature from 160 to 190° F. or the addition of ethyl gallate to the milk improved the keeping quality of the resulting powder, as measured by taste, by a factor of the order of 1½–2½ at 47° C, of 2–3 at 37° C. and of 3–4 at 15° C. and room temperature.

3. Increasing the pre-heating temperature and incorporating ethyl gallate in the milk extended the storage life of the powder by a factor of the order of 3 at 47° C., 4 at 37° C. and 8 at 15° C. and room temperature.

4. For an equivalent loss of palatability powder from high-temperature pre-heated milk absorbed appreciably more oxygen than the corresponding powder from low-temperature pre-heated milk, and the gallate-treated samples slightly more than the corresponding control powders.

5. The temperature coefficients for deterioration were slightly higher for the high-temperature powders than for the corresponding low-temperaturepowders.

6. With the two control powders storage in lacquered tinplate resulted in a small increase, of the order of 10%, in keeping properties as comparedwith storage in plain tinplate. With the ethyl gallate-treated powders no advantage resulted from the use of lacquered tinplate.

Previous work from this laboratory has indicated that the biological value of the proteins of milk is not affected by pasteurization (1), by evaporation and condensation (2) or by spray or roller-drying (3). Sterilization (4), on the other hand, caused a statistically significant decrease of 6% in the biological value; with this process heating at a high temperature is more prolonged than in the other forms of heat treatment.

It was of interest to find out whether the difference of 30°F. in the pre-heating temperature of the spray-dried whole milks described by White, Smith & Lea (5) in the preceding paper would have any effect on the biological value of the proteins.