It is a well-established fact that during the summer months, particularly July and August, the chemical quality of milk tends to be poor. The samephenomenon has been reported in Germany (i) and in the United States of America(2, 3). The individual factor or factors responsible have not been thoroughly established, although the work of Ragsdale and Turner (4) andHays (5) shows clearly that the fat content of milk tends to be reduced as the environmental temperature increases. The extent of the summer depression in milk quality in this country is much greater during hot dry summers. Cranfield(6) and Golding et al. (7) found exceptionally poor-quality milk during the summers of 1929 and 1933 respectively. In both these years most cows were subjected to a higher temperature than usual and also to a variety of other unusual factors such as the changes in nutrition occasioned by dried-up pastures. The work reported here was planned in order to throw some light on the influence of temperature, and whether it may be considered solely responsible for low-quality summer milk or whether other factors such as nutritional ones may be partially responsible.

The results of the present experiments may be summarised as follows:

(1) When minimal quantities of protein are fed in the maintenance rationof steers, a deficiency of lysine will lead to a marked increase in the urinary N output. A deficiency of cystine has no effect.

(2) The catabolism of body tissue is shown by the high creatine excretionand the S: N ratio of the excess S and N excreted. This is probably the resultof excessive deamination with the production of a mild acidosis.

(3) As regards urinary partition of N there is some indication that alysine-deficient diet causes a decreased percentage excretion of urea N and anincrease of ammonia N.

(4) The biological values have been calculated, the only marked difference occurring with wheat gluten, a protein very deficient in lysine

(5) An approximate value for the lysine requirements for maintenance has been calculated, i.e. 0·80 g. lysine N per 1000 lb. live weight.

1. The feeding of butterfat in the form of butter or cream had no significant influence on the yield or composition of the milk.

2. At the beginning and end of the fat-feeding periods there was atendency for some of the cows to go off feed.

Figures are given for the yield, specific gravity and composition of samples of milk collected at varying intervals throughout lactation from eight Cheviotewes kept under experimental conditions and of odd samples taken from ewesunder natural grazing conditions in different parts of Scotland.

The figures recorded do not show any marked differences in compositionas a result of wide differences in the calcium content of the rations.

1. With the cows and machines tested there was a great range in the duration of milking, the coefficient of correlation between yield and duration of milking being 0·429.

2. With some individual cows the correlation between yield and duration was high, and lines representing the regression of the duration of milking relative to the amount of milk drawn have been constructed.

3. Average composite curves representing the rate of milking at various stages during the complete operation have also been set out for varying yield groups. The main conclusions to be drawn from these curves are:

(a) With some cows the rate of flow was noticeably checked about –1 min. after milking commenced. This check was of temporary duration.

(b) Maximum rate of flow in all yield groups was reached about 2 min. after milking commenced.

(c) A very much more rapid rate of flow was recorded with high yields than with low yields.

(d) The decrease in rate of flow near the end of milking was rapid, especially with high yields.

1. The ratio of the weight of fat in the milk required to make 1 lb. of butter has been taken to indicate the size of the fat globule in the milk, and so the type of “ cream line ” produced.

2. The Dairy Show records have been analysed from this standpoint, and variations in the above ratio studied together with variations in milk yield, fat yield and fat percentage. The effects considered are those between breeds and within breeds, the latter including milk yield, period of lactation and ageof cow.

3. With a constant milk yield the fat percentage does vary between breeds, but not to the same extent as it does when no account is taken of the differences in average yield of the different breeds; the same applies tothe fat: butter ratio, i.e. size of fat globule. The effect of equal increases of milk yield on the butter:fat ratio is much greater between breeds than it is within any one breed.

4. It is suggested that the size of the fat globule is determined by two factors: (a) the rate of butterfat formation by the cell, which varies with breed and stage of lactation, and (b) the rate of milk secretion which affects the size of the globules by the rate at which they are washed out from the cell.

1. The colours of the butter for individual cows at the London Dairy Show have been measured on a colour scale of yellow shades on cellulose strips, and the results analysed statistically.

2. The mean values and variability curves of the ranges of butter colour in the different breeds of dairy cattle are given. Circumstances which may affect these values are indicated.

3. The two most important factors affecting butter colour are the genetic character of the cow and the method of feeding.

4. A slight increase in the shade of colour was found to occur under most of the various conditions which give rise to increased milk yield. It is suggested that this may be due to the greater intake of colouring matter (associated with the greater food intake) in proportion to butterfat produced, since increase in yield is usually associated with decrease in fat percentage.

1. Some account has been given of the cattle kept on the Government Farm at Samaru, Zaria, and of their management in the wet and dry seasons.

2. Figures extending over four seasons have been presented to show thatthe butterfat content of the milk produced on this farm is consistently higher in the morning than in the evening during the wet season, and that the reverseis true during the dry season.

3. The relative quantity of milk produced in the evening during the dry season is slightly less than during the wet season, but this difference is quiteinadequate to explain the change in the percentages of butterfat.

4. Some evidence is offered that this change in butterfat content is directly due to the marked climatic differences between the two seasons, and that it cannot be controlled by altering the diet of the cattle.

The weekly samples of typical milk from a small herd of Shorthorn cows have been analysed for chlorine content and nitrogen distribution (ninety-six samples). Lactose has been determined in seventy-five samples and copper and iron in twenty-four samples. The results are tabulated and discussed.

The author is indebted to Mr A. Wagstaff for carrying out the determinations of lactose.

1. A technique for examining bulk milk, by cultural and microscopical means, for evidence of mastitis among the cows is described. To demonstratea light contamination with haemolytic streptococci, cultural examination of gravity cream is advocated, whilst, if much significance is to be attached to the actual count, whole milk should be cultured.

2. Forty–two samples of Certified milk from sixteen producers wereexamined, and twenty–four of these contained beta-haemolytic streptococci when whole milk was cultured. Twelve of these herds were examined on three occasions during fifteen months, and samples from only two were free frombeta-haemolytic streptococci at every test.

1. In an examination of the coliform flora of 114 specimens of bovine faeces, 342 cultures of coliform bacteria were isolated by the ordinary methods.Of the cultures obtained, 330 (or 96·4 per cent.) were typical Bact. coli andonly twelve (or 3·5 per cent.) were of Bact. aerogenes, Bact. cloacae and other Koser-positive types. The most prevalent types of Bact. coli were MacConkey's B, No. 71 or Bact. coli communior (185 cultures or 56 per cent.), MacConkey's B, No. 34, or Bact. coli communis Escherich (sixty-nine cultures or 20·9 per cent.), and MacConkey's B, No. 1 (twenty-three cultures or 6·9 per cent.).

1. Acetyl methyl carbinol and diacetyl are present in New Zealand butters made from slightly ripened cream.

2. Most of the carbinol + diacetyl found in such butters would appear to have been added with the starter to the cream, rather than developed during the customary mild ripening process.

3. Little change in the natural carbinol + diacetyl content of butter occurs during cold storage at 14—17° F. for more than 6 months.

4. Larger amounts of diacetyl (4 p.p.m.) added to butters made from cream with and without the use of starter, show considerable losses after several months' cold storage. The losses are of the same order, but, with the butters, made with starter, part of the diacetyl has merely been reduced toits precursor (carbinol) form.

The coagulating enzymes employed in the manufacture of Cheddar cheese were varied in the following ways:

(1) Pepsin was substituted for rennet; two samples of somewhat different proteolytic action were used.

(2) The proportion of rennet normally employed for coagulating the milkwas doubled.

The use of pepsin resulted in cheeses in which the extent of protein degradationwas considerably smaller than in rennet control cheeses. The type of nitrogen partition observed was influenced to some extent by the proteolyticaction characteristic of the particular enzyme employed. The use of an increase dproportion of rennet led to the anticipated greater extent of protein breakdown, the type of nitrogen partition not being appreciably affected. Nodirect correlation was observed between the commercial quality of these cheeses and the extent or type of protein degradation which occurred.