we quantitatively recovered from the diet (Fig. I ) are not subject to this criticism. Fig. 2, which records graphically information from the paper of Owen, Smith & Wright (1943), who demonstrated that urea can be substituted for protein, shows the output of nitrogen by cows as a function of their N intake. I n these experiments the diet was of concentrate meal and hay which required separate 2-day analyses, and the output was urine plus faeces plus milk. Refusals of meal or hay were measured. I n Fig. 2 the heights of some of the ordinates above the line of equilibrium, indicate retentions of N which might at first sight seem unacceptably large. Two questions, therefore, arise: ( I ) how large a positive balance is physiologically admissible, and (2) if the balances are inadmissibly large, from what technical imperfections could they have arisen?
Question ( I )T h e 1943 experiment was done during the war when cattle food was rationed, and on a farm where the ration coupons were returned in order to find out whether wartime self-sufficiency was feasible. Before the experiment, therefore, these cows had been undernourished. Moreover, as a result of many years of N balance experiments on milch cows, the Institute had recommended a lower N allowance than was customary (Owen, 1941). Fig. 3 is a similar graph from a later experiment https://www.cambridge.org/core/terms. https://doi