Extirpation of the spleen has no influence upon the course of phlorhizin glycosuria.2 Nor has the establishment of an Eck fistula.3 An Eck fistula is one which diverts the whole of the portal circulation to the liver into the inferior vena cava, and leaves the liver supplied by the hepatic artery only. In this case the ingestion of glycocoll by the animal resulted in its complete transformation into urinary glucose, showing that the diversion of blood away from the liver in no way affected the synthetic production of sugar from this amino-acid.

Levene found that the bile contained a small amount of glucose after the administration of phlorhizin, and this has been confirmed by Woodyatt.4

Loewi5 has conceived the idea that the blood-sugar is normally in a loose combination with colloid substance. This colloid sugar cannot pass through the glomerulus. If, however, sugar accumulates in the blood above the combining power of the colloid, then the crystalloid glucose readily passes away through the kidney. This condition exists in diabetes mellitus. In phlorhizin glycosuria the kidneys break up the colloid sugar, and the sugar may then be eliminated. Stiles and Lusk, while accepting Loewi's theory, have added the hypothesis that the colloid sugar cannot be burned. Phlorhizin acting in the kidney will split the compound and permit the elimination of sugar. Any free glucose in the general circulation unites with the colloid radical and is protected from combustion, as is the case when 5 grams of glucose are administered subcutaneously, only to reappear in the urine (Stiles and Lusk). The presence of a colloid-glucose combination is denied by Rosenfeld and Asher,1 who find that the sugar of normal blood is readily diffusible.

1 Stiles and Lusk: "American Journal of Physiology," 1903, x, 67.

2 Austin and Ringer: "Journal of Biological Chemistry," 1913, xiv, 139.

3 Sweet and Ringer: Ibid., p. 135.

4 Woodyatt: Ibid., 1909-10, vii, 133.

5 Loewi: "Archiv fur exp. Path, und Pharm.," 1902, xlviii, 410.

It was discovered by Ringer2 that when a large quantity of glucose (75 grams) is given to a phlorhizinized dog it is completely eliminated in the urine, and Lusk found that the ingestion of this large quantity in no way affects the respiratory quotient (see p. 244). It is therefore evident that the completely phlorhizinized dog has lost the power of oxidizing glucose. This probably does not occur on the first day of the administration of phlorhizin and may possibly be due to the development of acidosis (see p. 261). Stanley Benedict3 reports that administration of glucose to the phlorhizinized dog causes the amount of blood-sugar to rise above the normal, which shows that sugar is present in ample concentration though it remains chemically untouched.

Phlorhizin glycosuria is only temporary in character, and subcutaneous injections of alkaline solutions of the drug three or four times daily have been employed in order to obtain constant results.

A more convenient method is that of Coolen,4 who noticed that the subcutaneous injection of 1 gram of phlorhizin suspended in 7 c.c. of olive oil caused a glycosuria of maximal intensity which lasted between five and ten days. Common laboratory practice at present calls for daily injections of this material.

The character of phlorhizin glycosuria has been dwelt upon because the protein metabolism is here identical with that observed in diabetes mellitus.

1 Rosenfeld and Asher: "Zentralblatt fur Physiologie," 1905, xix, 449. 2 Ringer: "Journal of Biological Chemistry," 1912, xii, 431. 3 Guion, C. M., and Benedict, S. R.: Paper read before the American Society of Biological Chemists, 1915.

4 Coolen: "Archives de Pharmacodynamic," 1895, i, 267.

Von Mering and Minkowski1 removed the pancreas from dogs and obtained a condition which was markedly analogous to diabetes mellitus in man. There is hyperglycemia and a large excretion of glucose in the urine; ingested glucose cannot be burned, but is completely eliminated. The dogs show a considerable acidosis with excretion of β-oxybutyric acid, and they die in coma.2 If a portion of the gland remain in the abdominal cavity there is either no diabetes or only a partial diabetes. Minkowski3 reports that if a piece of the pancreas be ingrafted under the skin of a dog and afterward the whole of the remainder of the pancreas be removed from the abdomen, the dog's urine remains free from sugar for two months, but on extirpation of the piece ingrafted under the skin an extreme diabetes sets in.

Allen4 reports that a dog which has a large part of its pancreas removed, but is free from diabetes, may gradually become diabetic by giving protein and fat, and may then manifest the spontaneous downward progress observed in human patients.

By an operation which united the blood supply of two dogs Forschbach5 established the condition of parabiosis. On the removal of the pancreas from one of the dogs neither developed diabetes. An analogous experiment is that of Carlson,6 who performed pancreatectomy upon bitches near to term and found little or no sugar in the urine. Here the embryo apparently furnished the mother with the substance essential to sugar oxidation. Murlin, however, in unpublished experiments finds that such dogs have diabetic respiratory quotients (0.69), and suggests that the absence of glucose from the urine is due to carbohydrate retention by the fetus.

1 von Mering and Minkowski: "Archiv fur exp. Path, und Pharm.," 1890, xxvi, 371.

2 Allard: Ibid., 1908, lix, 391.

3 Minkowski: Ibid., 1908, Supplement-band, p. 399. 4 Allen, F. M.: "Harvey Lectures," 1916-17. 5 Forschbach: "Archiv fur exp. Path, und Pharm.," 1909, lx, 131. 6 Carlson, Orr, and Jones, W. S.: "Journal of Biological Chemistry," 1914, xvii, 19.