INTRODUCTIONTHERE are two fairly well-defined methods for measuring the thermic energy of a feeding-stuff. In the first, the animal is maintained at a steady level of metabolism either by starving it or by feeding it for a period at a maintenance level, and a single feed of the ration to be investigated is then given. The total extra heat produced above the steady level is measured and taken as the thermic energy. Alternatively, in the second method, the feeding-stuff under investigation is fed steadily over a definite period in addition to the basal ration, if any, and the rise in heat production per day over the basal period is determined.The older work on Specific Dynamic Action is mostly based on the first method, and it has been used more recently by Hamilton (1939). For the comparison of rations for feeding purposes there are three objections to the first method. First, the experiments of the present writers described in paper II of this series confirm the observations of Barott (1938) and other workers, that there is a diurnal rhythm of metabolism causing a variation in its level at different times of day, and that there is therefore no steady level of metabolism in fowls to which the observations may be referred. In the second place, for application to feeding practice it is necessary to find the effect when the ration is being fed to an animal over a period, which should, strictly speaking, be comparable in length with that for which it is proposed to feed the ration in practice. In some experiments by the first method these conditions are fulfilled by measuring the heat produced by a single feed following a period of starvation after the animal has been fed for a considerable time beforehand on the experimental ration (Hamilton,
464Studies on the Metabolism of Fowls 1939). In this connexion it is permissible to refer to the recently published work of Mollgaard & Thorbek (1938), who find that the acid in A.I.V. silage, if not neutralized, has a powerful depressant effect on the NK t value of this silage for cows, most probably due to cumulative acidosis. This effect would presumably not appear if a single feed only of the silage were given. A third objection to the method is that the value for the nett energy obtained cannot be related to any particular plane of nutrition. In view of what has been said we have adopted the second and more usual method in this work. It may be added that the possibility of a variation in the nett energy value with time envisaged in the experiments mentioned may possibly be concerned in the cases -where nett energy varies as between animals of the same species, which have led one of the authors (T. D.) to conclude that it is an average value of a variable rather than a constant.The object of the experiments described in this paper was to gain information as to whether the nett energy value of cereals for fowls depended only on their content of digestible or metabolizable nutrients, or if definite differences exist in the quantity of thermic energy produced by equal quantit...