Most of those engaged in the discussion which followed the reading of Melville's paper at the annual meeting of the British Medical Association in London, 1910, agreed that the quantities were none too much, but emphasis was laid on the fact that quality of protein was of profound importance. Dr. E. P. Cathcart mentioned that Michaud had demonstrated this in his experimental feeding of dogs on dogs' flesh, and from this point of view, in human feeding, cannibals possessed really an overwhelming advantage. Carbohydrates were stated to be valuable not only in the rapid supply of energy they furnished, but because they played an intimate part in the metabolism of protein tissue. As during the march the lumbar muscles increased in bulk, protein was necessarily stored up, although their water-content was clearly responsible for four-fifths of this increase in size.

It was pointed out by others that too much stress could not be placed upon habit in the selection of food as a justification for the quantity consumed, and it was suggested that man's natural diet contained the following proportions of alimentary principles: .8 of proteins, 2 of fat, 3 of carbohydrates and .5 of mineral salts. Such a proportionate estimate, however, is purely empirical, and must necessarily vary with every alteration in the expenditure of energy, but it is evident that from such a short experiment, without careful training by a cautious and gradual reduction of the ingredients of the diet, it is hardly safe to draw conclusions as to the nutritive requirements of the body. It is fairly clear, however, that severe exertion instituted for a short period in men accustomed to a liberal dietary demands an increase in their daily rations. It is equally certain, from observations other than those of Chittenden, that if the body is kept in good training it is quite possible to lower the diet to two-thirds of Voit's standard, and yet find that the muscular strength and the power of carrying out work are actually increased.

In their efforts to uproot beri-beri, the Medical Bureau of the Japanese Navy commissioned a battleship to make a long voyage, the men receiving the usual naval ration of 91 grams of protein per day. The contrast between the enormous number of cases of beri-beri amongst the marines and the better-fed officers led to another battleship being sent on precisely the same route with a protein ration of 155 grams per day, with the result that this latter ship had infinitely less sickness than the former. This large quantity is now established by law as the Japanese sailor's allowance.

In a conversation with Chittenden on this subject, he contended that it does not necessarily follow that the improvement in the health of the crew was due to an increase in the amount of protein per se. He thought it was much more likely that other elements were introduced into the food capable of accounting for the disappearance of the disease. Protein pure and simple is unlikely to be utilised as a food in the body. It is much more likely that it requires to become a salt of lime, potash, or soda before it can be available for dietetic purposes. He thinks that the mineral salts introduced with the protein prevented beri-beri much in the same way that lemon-juice prevents the appearance of scurvy. He also combated the view that a diet rich in protein was likely to increase the resistance to disease. Dr. Reid Hunt, of Washington, fed two groups of dogs on a high and low protein diet respectively. Acetonitrile was then administered to each group, and it was found that the low-protein dogs could withstand a three times larger dose than the others, hence proving that they had a greater resistance to morbific influence.

Although Benedict very generously agrees with regard to these two observations that probably some factor was in operation to obviate the occurrence of the beri-beri other than the protein, he quotes some other experiments pointing to a different conclusion than that arrived at by Reid Hunt. Whilst he is inclined to modify his views slightly as regards some of the minor arguments against the low-protein diet, he believes that "the following observations militate against the view that a material reduction of protein in diet is desirable: -

"(1) From the results of the digestion experiments with the reduced diet on the soldiers, it is seen that abnormally low protein may affect the absorption of nitrogenous material from the alimentary tract.

"(2) Animals fed on diets low in protein do not thrive so well as on normal quantities.

"(3) Dietary studies all over the world show that in communities where productive power, enterprise, and civilisation are at their highest, man has instinctively and independently selected liberal rather than small quantities of protein."

"Energy Requirement." - Benedict has, however, contributed a much more valuable argument in favour of a liberal supply of protein from direct observation of subjects at rest and at work in the large respiration calorimeters he has installed in his laboratory. He believes that it is quite possible to live on a low-protein diet, but not that a sufficient supply of energy can be manufactured without a very much larger supply of food than is maintained by either Kellogg or Chittenden. He states that it is only necessary to know the daily energy output to determine the energy requirement, for the energy output is practically the energy requirement. He finds that a man of average size, weighing 66 kilograms and at rest within the calorimeter, has an energy requirement for twenty-four hours of not far from 2,270 calories. Naturally the average individual outside the calorimeter, pursuing his ordinary course of life but not engaged in any occupation, would actually require more than this, because with the greater liberty to move about, the movements of the body would call for the utilisation of more energy.

It is interesting to know that he subjected Mr. Horace Fletcher, who was the first man upon whom Chittenden experimented, to a calorimetric test. From the study of Mr. Fletcher's diet Chittenden came to the conclusion that Mr. Fletcher was able to hold his own with 1,700 calories of energy. During this period he was undergoing an unusual amount of muscular exercise with the Yale University crew, and yet on a small quantity of food not only was nitrogen equilibrium maintained, but his body-weight also remained practically unchanged. During the three days of residence in the respiration calorimeter, his average output of energy was 1,896 calories, whereas the amount of energy actually derived from food averaged only 1,375 calories. To make good this discrepancy, therefore, Mr. Fletcher was compelled to draw upon his body material for 541 calories, and the data of the experiment from which the gain or loss of body material is determined show that he actually lost body material sufficient to supply this amount of energy. Benedict, therefore, concludes that during the Yale experiment Fletcher's energy output could hardly have been less than 3,000 calories. The 1,300 calories representing the difference between the intake of energy as estimated in the food and the probable output during the excessive muscular exercise could be accounted for by the combustion of 150 grams, or about one-third of a pound of body fat. In the course of the short experiment of six or seven days there may have easily been a loss of one or two pounds of body fat, compensated by a gain of one or two pounds of water in the body, so that the body-weight would remain unchanged.