Since the time of the previous review in this series by the present author it has been increasingly realized that the ruminant's adaptation to its herbivorous diet is very far-reaching. The biochemical consequences of its need for coarse roughage are still being worked out and have necessitated a reassessment of the causes of bovine ketosis, of the function of lower fatty acids in its blood and of the reasons for its large requirement of cobalt. Recent work on all these three aspects of ruminant metabolism is discussed in this review. The cow's chief function is to convert into milk vegetable materials, many of which are unsuitable for human consumption. Ac-, cordingly, some of the nutritional aspects of this conversion process have been discussed in the review. The literature on the subject has been reasonably well listed, but the author cannot claim that he has not perhaps inadvertently omitted certain pertinent new work. GENERAL Sherwood & Jones(l) found that alfalfa hay plus a mixture of wheat, barley and oats gave 20% more milk than the hay alone. Milk yield was also enhanced by adding the succulents squash and Jerusalem artichoke to the hay diet. Groups of cows receiving the concentrates needed 5-6 lb. more total digestible nutrients to produce 100 lb. milk, and the authors conclude that feeding of succulents is profitable only if they produce nutrients as cheaply as does hay. Watson, Davidson, Kennedy & Sylvestre(2) compared the digestibility of rations of hay plus increasing amounts of oats. As oats increased the digestibility decreased. This effect was most marked with barley, not so marked with oats and was only slight with linseed oil meal. Feeding standards are still the subject of informed criticism in Europe(3,4,5). Nehring(3) concludes that Kellner's starch equivalents cannot be used for pigs without prior respiration calorimetry. Blaxter(4) finds large discrepancies between different methods of computing rations for livestock, and Nicholls(5) enters a plea for food recording as a complement to milk recording. The trouble with all attempts to define standards for grazing animals is the extreme variability from year to year, from field to field, from treatment to treatment and from animal to animal of the nutrients provided by grass and hay or silage. Much current research all over the world is designed to measure more accurately the nutrient intake of cattle and sheep from pasture, with a view to striking a better balance between protein and fibre in the diet, always remembering that the content of fibre biologically best for maximal production per cow may not necessarily give the maximal production per acre. This may come from more cows per acre at somewhat more fibre per cow than would give maximal production per cow. In this connexion the observations of Hancock (6) are pertinent. He points out that the maximum production per animal does not necessarily yield the greatest net returns. In practice this can be achieved by a number of cows per acre too high for maximum productivity per cow or fo...