WEIGHT & MORRIS (i) have .briefly reviewed the energy requirements of dairy cattle along with those of other classes of stock. They draw attention to the calculations of Brody and of Kleiber, which indicate that the estimation of the maintenance requirement on the basis of the surface area of the body is not well founded. Both these latter authors have abandoned considerations of surface area, and assess the maintenance requirement on the basis of its proportionality to W°' n. Brody (2) suggests the formula <2 = 39-5PF 073 , where the basal metabolism (Q) is measured in calories and the live weight (W) in pounds. Gaines(3) has suggested an analogous method of calculating the energy requirement for milk production, his formula, derived from an analysis of American data, being D = 0-275.F + 0-009 W, where D, F and W represent lb. of digestible nutrients, lb. of milk of 4 % fat content, and lb. of live weight respectively. He incidentally concludes in a further paper (4) that the yield of milk, measured in terms of energy, is independent of age and that the so-called 'age correction' is merely an indirect way of allowing for live weight. Wright & Morris (l), while drawing attention to the proposals for a single basis of calculating energy requirements which were made at the Pennsylvanian Conference on Energy Metabolism and which were largely based on Brody and Kleiber's findings, have concluded that the general principles of rationing adopted in Great Britain, as laid down by Wood & Woodman (5), do not warrant revision meantime. They suggest, however, that the maintenance requirement could be safely reduced from 6 to 5£ lb. per 1000 lb. live weight (though in practice this would form an almost negligible reduction in the total ration), but state that the requirement for milk production does not need any modification. As regards protein requirements, reviews have been published by McCandlish(6) and by Wright (7). Both these authors point out that during the past 25 years there has been a marked general tendency to reduce the earlier standards. From his own experiments McCandlish concluded, however, that there was a real practical difficulty in estimating the protein requirements by ordinary feeding trials, for the protein intake might, he found, be changed considerably without producing any proportionate alteration in the milk yield. This view does not receive support from the experiments of Borland (8), who, in a two-year trial involving twenty-four cows, employed rations of grain at the Haecker standard and at 10 and 20 % below and 10, 20 and 30 % above that standard. Returns in milk production were found to be proportional to the grain fed, even at the highest level. Bartlett et al. (9) and Mackintosh (10) in experiments designed to test the effect of lowering the protein standard for milk production from 0-6 to 0-4 lb. protein equivalent per 10 lb. of milk, failed to obtain any reduction in the yield of milk or any loss of condition in the animals, though the rise in milk yield when the animals were put out to gr...