The blood of the first six patients showed a normal PH. Only in the depth of coma a few hours before death is there a distinct fall in alkalinity, and, indeed, this fall may not be as great as in a normal person after climbing a thousand feet in twenty-five minutes, under which circumstances the PH may be 7.09 (see p. 322).

It is evident that the reaction of the blood in severe diabetes is maintained at the normal through the reduction of its carbon dioxid content. Such a reduction in carbon dioxid combining power indicates a reduction in the alkali reserve of the blood, and forms the basis of the important method of Van Slyke for investigating the intensity of acidosis.

That ammonia in the urine is an indicator of acid formation and not due to a pathologic disturbance of urea formation was shown by Muenzer,2 who gave alkali in cirrhosis of the liver and reduced the quantity of ammonia elimination to normal. Fiske and Karsner3 find that livers which have been severely damaged in the living animal by administration of chloroform, phosphorus, hemolytic immune sera, hydrazin sulphate, or phlorhizin still preserve the power of transforming perfused ammonium carbonate into urea. Janney,4 in von Muller's laboratory, gave bicarbonate of sodium to men and found that the quantity of ammonia in the urine was reduced to almost undeterminable traces; hence the ammonia in the urine has as its sole function the neutralization of acid bodies and ceases to be formed in the presence of an excess of fixed alkali.

1 Poulton: Proceedings of the Physiological Society, p. i; "Journal of Physiology," 1915, 1.

2 Muenzer: "Deutsches Archiv. fur klin. Med.," 1894, lii, 199 and 417.

3 Fiske and Karsner: "Journal of Biological Chemistry," 1914, xviii, 381.

4 Janney: "Zeitschrift fur physiologische Chemie," 1911-12, lxxvi, 99.

Howland and Marriott1 find that the administration of acid phosphates causes no increase in ammonia in the urine.

Klein and Moritz2 found that on the day following a diet which was rich in fat there was an increase in the quantity of fixed alkali in the urine and a corresponding fall in the quantity of ammonia. They interpret the results as signifying that the alkali was temporarily involved in fat metabolism (formation of soaps) and was eliminated when this need was no longer present.

The consideration of the ingestion of alkalies and bases in the food will be discussed in the chapter on A Normal Diet.

1 Howland and Marriott: Reported at the meeting of the American Physiological Society, December, 1916.

* Klein and Moritz: "Deutsches Archiv fur klin. Med.," 1910, xcix, 162.