1. Four species of Medicago, six species of Trifolium and fourteen grass species, all frequently found in British dairy pastures, have been tested for oestrogenic activity by the mouse uterine weight method. Activity was detected in four, four and six species respectively.

2. It is probable that the quantity of oestrogen present in British pasture plants is insufficient to affect milk composition in cows at pasture. Some further experiments with cows are, however, suggested which might elucidate this.

1. Four test organisms were used for the microbiological assay of vitamin B6. The test organisms and the forms of vitamin B6 active for them were: Saccharomyces carlsbergensis 4228 (pyridoxine, pyridoxamine and pyridoxal), Streptococcus faecalis R ATCC8043 and Str. faecium Ø51 (pyridoxamine and pyridoxal), and Lactobacillus casei ATCC7469 (pyridoxal).

2. By simultaneous assay of the test material with the four organisms the amounts of pyridoxine, pyridoxamine and pyridoxal present were determined. The differential assay technique has been improved by the use of Str. faecium, which is more specific for measuring total pyridoxamine plus pyridoxal, instead of Str. faecalis.

3. The optimum conditions for measuring vitamin B6 in milk were studied and the differential assay technique was applied in an investigation of the effect of heat on the vitamin B6 activity of milk.

4. It was found that pyridoxal accounted for about 80% of the vitamin B6 activity of raw milk, the other 20% being due to pyridoxamine.

5. In the manufacture of evaporated milk there was a marked fall in the pyridoxal and a slight increase in the pyridoxamine content. The overall loss was some 60%.

6. On storage of evaporated and of sterilized liquid milk at room temperature there was a further loss of vitamin B6 activity. Most of the residual activity in stored evaporated milk appeared to be due to a substance other than pyridoxal or pyridoxamine, possibly pyridoxine.

1. For chicks and rats pyridoxine, pyridoxal and pyridoxamine were equally active in terms of the free bases when given separately from the diet.

2. Under our experimental conditions pyridoxine mixed with the chick diet was stable, but 20% of pyridoxamine, and a variable amount of pyridoxal was lost.

3. The vitamin B6 activities measured with Saccharomyces carlsbergensis, chicks and rats respectively and expressed as μg. pyridoxine/g. freeze-dried milk were: raw milk 3·4, 3·2 and 4·9; evaporated milk 1·0, 2·1 and 2·7; stored evaporated milk 0·6, 1·4 and 2·0. For the chicks the milks were mixed with the diets; they were given separately to the rats.

4. The microbiological and biological results for raw milk agreed within the limits of experimental error. For the processed milks the differences between biological and microbiological tests were statistically significant.

5. All three methods of assay showed a 45–70% loss of vitamin B6 activity on processing and a further loss of 30% of the remainder after storage for 6 months at room temperature.

We are indebted to Mr J. Rothwell, Department of Dairying, University of Reading, for preparing the evaporated milk and to Dr B. Record, Ministry of Supply, Microbiological Research Establishment, Porton, for freeze-drying the milk. We should like to thank Dr S. K. Kon for his interest in this work.