1. After a preliminary trial with Shorthorn cows in Berkshire, some 461 Friesians were measured in Minnesota, U.S.A., and an attempt was made to correlate 19 measurements with milk yields.

2. Significant coefficients of correlations were obtained for:

Length from withers to hooks and milk yields;

Height at hooks and milk yield; Height at pins and milk yield;

Circumference of chest and milk yield;

Circumference of barrel and milk yield;

Area of milk wells and milk yield.

3. Correlation coefficients that were almost significant were obtained for:

Width at barrel and milk yield;

Width at hooks and milk yield.

4. The best producing cows had the largest milk wells. This was not because they were the biggest cows, for there was practically no loss of coefficient of correlation for milk yield and size of milk wells when the partial correlation for area of milk wells and yield was calculated with height at hooks held constant.

5. Evidence is presented which shows that the areas of milk wells do not increase with mere skeletal growth of cows over 4½ years old, but with increased milk production. This fact makes the area of milk wells a still more reliable indicator of milk-producing power.

6. The results obtained confirm the value of some of the points which stockmen have considered in judging dairy cows. They emphasise the necessity for considering all the points together, but indicate that certain points deserve special consideration.

1. A study has been made of the degree of inbreeding in the Ayrshire breed of cattle.

2. The coefficient of inbreeding for the whole breed, as calculated by Wright's “Approximate (or Short) Method,” showed a progressive increase from nil in 1877 to a mean value of 5·3 in 1927.

3. No difference was found in the coefficient of inbreeding as between registered bulls and cows.

4. A large proportion of the inbreeding was traced to two foundation sires, Burnhouses (8) and Hover-a-Blink of Drumjoan (892).

5. Using Wright's “Long Method,” it was found that high milk-yielding cows showed a markedly lower coefficient of inbreeding than the breed average. On the other hand, when the coefficient of inbreeding of high milk-yielding cows (over 1000 gallons) was compared with that of cows giving average yields (under 800 gallons) it appeared that inbreeding itself had no detrimental effect on milk yield.

6. In 1923, 39 per cent, of the inbreeding ties of high milk-yielding cows were traced to the Drumjoan herd, as against the next highest herd which had only 6·37 per cent, of ties. Of the total coefficient of inbreeding the Drumjoan herd was responsible for 18·36 per cent., whereas the next highest herd was only responsible for 8·35 per cent. The 1928 results also demonstrated the marked influence of the Drumjoan herd. Practically 50 percent, of these ties were traced to animals related to the foundation sire, Hover-a-Blink of Drumjoan. These results demonstrate the profound influence which one herd, and indeed one bull, may exert in the genetic construction of a recently established breed.

7. It is pointed out that Wright's “Long Method,” when applied to breeds of relatively recent origin, gives far more reliable results than his “Short Method.”

The work herein reported has shown that more milk and fat were produced per hour during 9-, 12- and 15-hour night intervals than during intervals of the same length during the day.

The milk increment was, however, greater than the fat increment.

With midday and midnight milking hours more milk and fat were produced per hour from midnight to midday than from midday to midnight.

In this case the fat increment was the greater.

An analysis of the data collected at the Agricultural Research Institute for Northern Ireland, Hillsborough, County Down, was undertaken to find if the system of milk recording practised in Northern Ireland is satisfactory.

The standard deviations of the errors in the calculated milk yields, on a weekly and 2-weekly basis, of 24 cows were found. It is shown that weekly weighings give results which approximate closely to the true milk yield.

By basing the calculated butterfat yields on tests made at intervals of from 1 to 8 weeks, the standard deviations of the errors in the calculated butterfat percentages for each interval, as well as the maxmium errors, were found. It is shown that as the interval between the tests is lengthened, the errors increase. With tests made at intervals of 6 weeks, most of the errors in the calculated butterfat percentages will be less than 8-5 per cent., but in some cases the calculated butterfat percentage may differ from the true butterfat percentage by over 12 per cent.

In order to ensure that the errors in calculated butterfat percentages do not exceed 10 per cent., the interval between butterfat tests should not exceed 1 month.

An investigation of the milk of a typical herd of Shorthorn cows kept under south of England conditions has been carried out.

The results of the first three years of the investigation are now recorded. The plan of the experiment, the history of the cows during the period, the management, feeding, milking, and milk yields are described and illustrated in Tables I and II.

1. Colon bacilli grown in milk for from 3 to 16 hours partially inhibit the subsequent formation of acid by lactic streptococci.

2. The inhibition is not entirely accounted for by the increased hydrogenion concentration nor by the lactic acid produced by the colon bacilli. Other substances which have a retarding influence on the streptococci are evidently formed in milk cultures of colon bacilli. The action of these substances is not significantly modified by heating at 75°C. for one hour.

3. The possible practical significance of the phenomenon is discussed.

1. The true lactic acid bacteria cannot grow in a medium in which the sole source of nitrogen is an amino acid mixture or an ammonium salt.

2. The presence of growth factors (yeast extract, marmite and peptones) does not permit the utilisation of this “simple” nitrogen.

3. The streptococci grow fastest with those peptones containing the greatest amount of the higher fractions, Bactoproteose being superior to the other peptones tested; the lactobacilli, however, show a considerably greater fermentation rate with those peptones containing most amino nitrogen, Fairchild's being the best. Similarly, the streptococci thrive best upon a peptic casein digest which is about 15 days old, further digestion impoverishing the growth; the lactobacilli, however, grow better the longer the digestion. The significance of these observations is discussed from the aspects of bacterial metabolism, the sequence of flora in ripening cheese, and the preparation of casein digest media.

The plating method of estimating the bacterial content of milk is the most important means of determining the number of living organisms present, and is the one most frequently employed. It is well recognised, however, that accurate counts cannot be obtained by this method, since considerable errors may be involved through the difficulty of ensuring the separation of chains and clusters of organisms so that each organism is represented by a separate colony, through the inability of some classes of bacteria to grow on the medium employed, and through the fact that the presence of one group of organisms may inhibit the growth of others—particularly on crowded plates. It appears that the difficulties in obtaining uniformity of counts for duplicate tests may be increased in the case of milk because this fluid is heterogeneous in nature, and the bacterial contaminants are liable to occur not only as groups or individuals freely dispersed in the liquid but also in association with matter in suspension, especially the clusters of fat globules. Consequently the bacterial counts are likely to vary with the extent to which the clusters of fat globules are broken up or aggregated by the manipulations.

It has been shown that neither lactose, nor other sugars tested, casein peptone, nor yeast extract can satisfactorily replace milk as a means of correcting discrepancies which are frequently observed between the plate counts of the 1/100 c.c. and 1/1000 c.c. dilutions of milir on standard agar. Experiments are in progress by which it is hoped that the objections to the addition of milk to standard agar either at the time of making or at the time of plating will be removed, and a medium be evolved which will combine the ease of making and the transparency of standard agar with the efficiency of the medium enriched with milk.

No reliable information is at present available with regard to the effect of process of manufacture on the solubility of the protein of milk powders. Hunziker(1), it is true, gives a very well-balanced discussion of the subject, but it is apparent from his summary of the views of other workers that these are seldom based on adequate experimental evidence. In fact the only points on which there is general agreement among workers in this field are that the proteins of spray-dried powders are, on the whole, extremely soluble, while those of roller-dried powders are relatively insoluble. Moreover, even where experimental data are available, there remain many unexplained points. For example, Hunziker shows that the solubility of the proteins of a milk powder may vary according to the temperature at which the powder is reconstituted—a higher temperature giving a greater solubility. Such an observation cannot at present be brought into line with current theories regarding the properties of denatured proteins.

The nitrogen distributions in 62 samples of whole morning milk covering a range of solids-not-fat content from 6·5 to 9·0 per cent, are discussed.

Samples high in solids-not-fat showed greater constancy in nitrogen distribution over a consecutive range of samples than those low in solids-not-fat.

Rich samples were characterised by high total nitrogen, low non-protein nitrogen and high casein nitrogen. Poor samples were characterised by low total nitrogen, high and variable non-protein nitrogen, low casein and consequently high albumin and globulin nitrogen. In samples low in solids-not-fat the albumin and globulin distribution was variable.

The casein/albumin/globulin ratio was roughly 14:2·5:1 for samples high in solids-not-fat, but the casein dropped in its proportion to globulin as the solids-not-fat decreased and the albumin showed also a slight parallel decrease.

The samples low in solids-not-fat which were studied appeared to be abnormal secretions, since they showed increased albumin, globulin and chloride contents.

The thanks of the author are due to C. E. Lesser and A. WagstafE for the figures for fat and solids-not-fat.

1. Figures are presented which show that there is a very appreciable variation in the salt and moisture contents of small plug samples taken from different cheeses from the one vat.

2. The fact that a cheese plug has to be ground to a paste before good duplicate results for salt content can be obtained is a further illustration of the variation in salt content within a single cheese, and of the slowness of diffusion of salt in Cheddar cheese.

3. The extent of diffusion of salt into unsalted Cheddar cheese has been measured in two different experiments. The results show that diffusion is very much slower than in Limburger cheese, probably owing to differences in moisture content and texture

4. It is well known that salt has an inhibiting action on the growth of bacteria in cheese. With variation in salt content from place to place in a cheese it is therefore to be expected that local variation will be met in the bacterial content. This may explain in part at any rate, the difficulties encountered in obtaining concordant results for estimations of number of bacteria in cheese.