The facts just related bring forward a new aspect from which the relative nutritive values of different foods may be judged. The older criteria must frankly make room for the new or else be displaced by them. Experiments upon the utilisation of food-stuffs, in which what remains undigested is determined as well as what is absorbed into the body fluids, cannot alone be trusted to solve the question in a satisfactory manner. Suppose, for instance, that in the digestion of a given food the alimentary canal has been given a certain work to perform; if it be in health the work will be accomplished in the best possible manner - that is to say, with complete abstraction of everything nutrient. You will thus learn how much nutrient material was contained in the food, but the question of its digestibility remains as obscure as before. In your experiment you do not know how great an effort it has cost the alimentary canal to extract all the nourishment from the food. Nor can artificial digestion experiments settle the question of digestibility, for experiments in which food is normally partaken of are quite different from those in the test-tube, where we have to deal with only one juice, and not with the interaction of different juices and different food constituents.

That one must here, as a matter of fact, make a distinction is clear from the observation of Dr. Walther in our laboratory. Fibrin, which is regarded by all as the most easily digested proteid, proved, when compared with a nitrogen equivalent of milk, to be a much stronger excitant of the pancreas, although milk contains, in addition to nitrogenous substances, a good deal of other non-nitrogenous material. The digestibility and nutritive value of foods must obviously be decided by an estimation of the real work which they entail upon the digestive apparatus, both in regard to the quantity and quality of the juices poured out on a given amount of nutrient material. The energy used up in gland metabolism must be deducted from that of food taken in. The remainder will then indicate the value of the food to the organism; that is to say, will give the amount available for use by all the other organs exclusive of the digestive apparatus. From this point of view those materials must be taken as less nourishing and less digestible which are in large part used up to make good the expenditure entailed by their digestion on the part of the alimentary canal; that is to say, those food-stuffs are less useful whose nutritive value little more than covers the cost of their digestion; consequently, it is of great practical importance to compare from this aspect the same foods differently prepared - for example, boiled and roast meat, hard and soft boiled eggs, boiled and unboiled milk, etc.

A discussion of some further medical questions may here be taken up. The first concerns the therapeutic use of the neutral and alkaline salts of sodium. In clinical, pharmacological, and physiological text-books it is stated now, as ever, that these salts promote a flow of gastric juice. We may look in vain, however, for any experimental foundation to support this doctrine. The experiments brought forward cannot be regarded as conclusive. When Blondlot sprinkled sodium bicarbonate upon flesh, or Braun and Griitzner introduced sodium chloride solutions directly into the blood, they began with methods either false in themselves or far removed from normal conditions. In this case, however, the gaps in the experiment were happily made good by the clinician, for the experiment appeared to be confirmatory of clinical experience. That sodium salts (the chloride and bicarbonate) are useful in disorders of the digestive apparatus there can be no doubt. How do they act, however? It appears to me that here, as in some other cases, medical science has fallen into error.

When we know that an effect takes place it does not by any means imply that we know the mechanism by which it occurs; and although medicine is broad enough and comprehensive enough to make free use of empiricism in practice, yet it often thinks in narrow grooves when it turns to the explanation of facts. It frequently tries to explain complicated healing processes in the simplest way, on supposed physiological data. And this is true in the present case, which affords an example of prevalent medical reasoning; the alkalies work favourably in digestive disturbances - therefore they are succagogues. Naturally the stomach, under the influence of alkalies, sometimes begins to secrete a greater quantity of juice. This means, however, that it has recovered from a disordered state and has returned to normal conditions. Consequently, the effect is due to the fact of recovery, and not to a direct influence of the alkalies. This latter, however, must be specially proved. The assistance afforded by the alkalies to the organism might be capable of another explanation; for example, that which is ordinarily given. In this case, however, I venture to offer a reason for the effects of sodium chloride, and of the alkaline salts of sodium, which is exactly the opposite of that generally accepted.

We were unable to convince ourselves of any succagogue influence on the part of these salts. Indeed, both on the stomach and pancreas they proved in our hands to have an inhibitory effect. In addition to the experiments which I previously brought forward concerning the relation of alkalies to gastric and pancreatic juice, I may relate the following observation. A dog which fortunately had survived the performance, one after the other, of a gastric fistula, a pancreatic fistula, and an cesophagotomy, received daily during the course of several weeks an addition of soda to its food. The animal enjoyed good health and had an excellent appetite. When the first sham feeding experiment was carried out, the relatively small effect of this otherwise very active juice-exciting procedure at once struck us. At the same time we observed that the pieces of flesh which fell from the upper end of the oesophagus, contrary to the ordinary rule, were hardly at all insalivated. In this dog, therefore, a greatly lowered activity of several digestive glands - viz., of the gastric, pancreatic, and salivary glands - simultaneously existed. With regard to the salivary glands the circumstance was naturally submitted to closer investigation.