McCollum1 has observed that pigs may be nourished for long periods of time when fed with a mixture of starch and inorganic salts in sufficient quantity to maintain their weights and energy requirements. After twenty-one to thirty-six days of this diet the relation of creatinin N to total N in the urine is a constant, or 18.5 : 100. Since the creatinin N has this as an invariable constant (when the diet is free from this constituent) it .follows that the true minimal endogenous level of protein metabolism may be calculated at any stage of the experiment by multiplying the quantity of creatinin N by 5.5 Nitrogen in excess of this is supposed to be derived from the destruction of "deposit protein" (see p. 287).

1 McCollum: "American Journal of Physiology," 1911-12, xxix, 210.

The table just given, which shows Folin's urinary figures for the seventh day of a starch-cream diet in man, shows a relationship of 17.2 : 100 between creatinin N and total N, thus indicating that the principle may be applicable to man.

The daily elimination of a gram or more of creatinin in the urine is certainly of moment.

Creatinin continues to be eliminated after an Eck fistula (p. 451) has been established in a dog,1 indicating that the liver cannot be all important in its production. If creatinin be administered with the food it is almost completely eliminated in the urine.2 The creatinin elimination is not influenced by muscular work,3 nor by the increased metabolism of body proteins which follows phosphorus-poisoning in fasting dogs.4

All of these facts seem to indicate that there is a mechanism in the muscles which is constantly active preparing a definite quantity of creatinin. Whether this comes from arginin or not is unknown (see p. 204).

Shaffer6 has proposed the term creatinin coefficient to represent the number of milligrams of creatinin nitrogen eliminated per kilogram of body weight in twenty-four hours. He believes this to be an index of muscular development. In 37 normal men the creatinin coefficient was between 8 and 11. In full accord with this theory Tracy and Clark6 find the creatinin coefficient of 26 normal women students in a professional school to average 5.8. In two athletic women, with unusual muscular development and control through gymnastic exercise, the creatinin coefficients were 9 and 9.8 or the same as in men.

1 London and Boljarski: "Zeitschr. fur physiol. Chemie," 1909, lxii, 465. 2 Folin: "Hammarsten's Festschrift," 1906.

3 Van Hoogenhuyze and Verploegh: "Zeitschrift fur physiologische Chemie," 1905, xlvi, 415.

4 Lusk: "American Journal of Physiology," 1907, xix, 461.

5Shaffer: Ibid., 1908, xxiii, 1.

6Tracy and Clark: "Journal of Biological Chemistry," 1914, xix, 115.

Arguing from the premise that the creatinin metabolism is an index of the quantity of active protoplasm of muscle tissue, Palmer, Means, and Gamble1 have compared the basal metabolism of 8 men and 9 women in relation to their output of creatinin. The group of men produced 0.98 calories of heat per milligram of excreted creatinin, and the group of women 1.26 calories for the same unit. If the premise is correct, then the mass of active protoplasm is not a factor in the measurement of the intensity of the basal metabolism (see p. 130). .