It is well known that the tissues themselves contain specific enzymes, some of which have been isolated, and which are responsible for autolysis in appropriate conditions. It is also known that the blood proteins are not food for cells, because during starvation they undergo practically no diminution, and in any case are not sufficient to account for the continued energy put out by the tissues. Voit has pointed out that during starvation the "noble tissues," the brain, central nervous system, and the heart, literally live on the less "noble tissues," fat, muscle, and connective tissue, and considers that there is therefore a strong presumption that nutrition in starvation is by amides and amino-acids derived from the organs autolysed and carried to organs which can synthesise these katabolites into tissue proteins. Miescher showed that the Rhine salmon, which does not feed in fresh water, nourishes its reproductive organs at the expense of its muscles. There is thus evidence for the existence of two forms of enzyme in the tissues, one protein-disintegrative, the other protein-integrative in action, although it is possible that it might be the same enzyme, and thus be an example of "reversible zymolysis." Loewi fed dogs on the abiuret products of the autolysis of the pancreas, i.e., on soluble crystalline substances which give no biuret reaction at all, and, with of course the addition of fat and carbohydrates, found that they remained in nitrogenous equilibrium. Henriques and Hansen demonstrated that this would not take place if hydrochloric acid were used instead of the autolytic or tryptic enzymes, and they suggested that the acid uncouples some linkages which the ferment does not.

Besides this, there is a denitrifying enzyme at work in the intestinal mucosa which splits off the amido group as NH3, and this, sent to the liver by the portal blood, is transformed into urea and so excreted by the kidneys. There is a great increase of NH3 in the portal blood during the absorption of protein food. When, therefore, the nitrogen molecule has been dissociated from the non-nitrogenous or carbonaceous molecule, this latter is, in the shape of fatty acids, oxy-acids', or thio-acids, passed to the tissues and is used for the production of energy. The relation of nitrogen to carbon in the protein molecule is as 16: 52, i.e., 1: 325. Leathes believes that quite 90 per cent. of the carbon of the protein is metabolised in unions which are non-nitrogenous. Protein, therefore, is dealt with in two distinct ways according as its nitrogenous molecule or its carbonaceous molecule is required. In the former case the nitrogen goes to the tissues for building up and repair purposes; in the latter, the nitrogen is excreted as urea.