§ 23d. On passing to the metabolism characterizing animal life, which, as already indicated, is in the main a process of decomposition undoing the process of composition characterizing vegetal life, we may fitly note at the outset that it must have wide limits of variation, alike in different classes of animals and even in the same animal. If we take, on the one hand, a carnivore living on muscular tissue (for wild carnivores preying upon herbivores which can rarely become fat obtain scarcely any carbo-hydrates) and observe that its food is almost exclusively nitrogenous; and if, on the other hand, we take a graminivorous animal the food of which (save when it eats seeds) contains comparatively little nitrogenous matter; we seem obliged to suppose that the parts played in the organic processes by the proteids and the carbo-hydrates can in considerable measures replace one another. It is true that the quantity of food and the required alimentary system in the last case, are very much greater than in the first case. But this difference is mainly due to the circumstance that the food of the graminivorous animal consists chiefly of waste-matter--ligneous fibre, cellulose, chlorophyll--and that could the starch, sugar, and protoplasm be obtained without the waste-matter, the required bulks of the two kinds of food would be by no means so strongly contrasted. This becomes manifest on comparing flesh-eating and grain-eating birds--say a hawk and a pigeon. In powers of flight these do not greatly differ, nor is the size of the alimentary system conspicuously greater in the last than in the first; though probably the amount of food consumed is greater. Still it seems clear that the supply of energy obtained by a pigeon from carbo-hydrates with a moderate proportion of proteids is not widely unlike that obtained by a hawk from proteids alone. Even from the traits of men differently fed a like inference may be drawn. On the one hand we have the Masai who, during their warrior-days, eat flesh exclusively; and on the other hand we have the Hindus, feeding almost wholly on vegetable food. Doubtless the quantities required in these cases differ much; but the difference between the rations of the flesh-eater and the grain-eater is not so immense as it would be were there no substitution in the physiological uses of the materials.