Cremer,1 in a series of excellent experiments, has shown that a vegetable pentose, such as rhamnose, may be burned in a rabbit and spare an isodynamic equivalent of fat. In one rabbit, on a fasting day, the metabolism amounted to 129.1 calories (protein, 22.5, and fat, 106.6), and on the day when rhamnose was given to 128.4 calories (protein, 21.36; fat, 32.9, and rhamnose, 74.11).

Lindemann and May2 found that 90 grams of rhamnose could be used by a normal man. When, however, rhamnose was given to a diabetic individual whose urine had been sugar free, sugar appeared in the urine. In cases of severe diabetes reported by von Jaksch3 it was found that rhamnose, arabi-nose, and xylose tended to increase the protein metabolism, and hence the sugar output, and also brought about diarrhea. The use of pentoses in diabetes has, therefore, not been successful. The pentoses - rhamnose, arabinose, and xylose - are not convertible into glucose in the organism.4

Opie5 was the first to establish a connection between changes in the islands of Langerhans of the pancreas and the cause of diabetes. Janeway and Oertel,6 von Noorden, and others, have reported autopsies on cases of severe diabetes in which the pancreas appeared perfectly normal. It is not always possible to observe with the microscope the cause of pathologic change in function.

Allen7 found degeneration of the islands, which was accompanied by diabetes, after the removal of nine-tenths of the pancreatic tissue in the dog, and Homans8 reports that the removal of three-fourths of the pancreatic tissue in the cat produces one of two results, either a disappearance of secretory granules in the islands of Langerhans with suggestive evidence of overactivity and without diabetes; or, occasionally, a degeneration of the islands of Langerhans without disturbance of the remaining acinous tissue, but accompanied by fatal diabetes.

1 Cremer: "Zeitschrift fur Biologie," 1901, xlii, 428.

2 Lindemann and May: "Deutsches Archiv fur klin. Med.," 1896, lvi, 283.

3von Jaksch: Ibid., 1899, lxiii, 612.

4 Brasch: "Zeitschrift fur Biologie," 1908, 1, 113.

5 Opie: "Journal of Experimental Medicine," 1901, v, 397-6 Janeway and Oertel: "Virchow's Archiv," 1903, clxxi, 547.

7 Allen: "Glycosuria and Diabetes," Boston, 1913. 8 Homans: "Journal of Medical Research," 1914, xxx, 49.

On autopsy in diabetes large quantities of fat are frequently found in the liver and muscles. The same is observed in chloroform narcosis when sugar appears in the urine, in anemia, and after respiration of rarefied air, where lactic acid is eliminated in the urine (p. 423), and in phosphorus- and arsenic-poisoning, in acute yellow atrophy, in pernicious vomiting of pregnancy, in eclampsia and in cyclic vomiting in children, which are similarly accompanied by an elimination of lactic acid. These phenomena are always associated with an increased protein metabolism and an increased ammonia and amino-acid output in the urine.1 Fat likewise appears in the mammary glands during lactation (see p. 396).

Virchow assumed a fatty degeneration of protein in which the tissue protein was converted into fat, as distinguished from a fatty infiltration in which body fat passed into the cells. Much of the earlier writing of Voit is pervaded with the theory of a considerable origin of fat from protein (p. 228). The idea of a fatty degeneration of protein in the old sense has been largely overturned by the work of Rosenfeld.2 He finds that if a dog be starved and then given sheep's fat, and again starved, the ingested fat will be found deposited as sheep's fat in his adipose tissue, while the liver will contain about 10 per cent, of fat, and this characteristic dog fat. If now phosphorus- or phlorhizin-poisoning be induced and the liver be examined, 40 per cent, of fat may be found therein, and this in the form of sheep's fat. Hence, in these cases the fat is simply transported to the liver from the fat deposits of the body. The fat in the blood is largely increased. The fat becomes normal in quantity in the liver twenty-four hours after the cessation of the phlorhizin action. It is retransported to the places of fat deposit.

1 For literature consult Ewing: "Archives of Internal Medicine," 1908, ii, 476.

2 Rosenfeld: "Ergebnisse der Physiologie," 1903, ii, I, p. 50.

B. Fischer1 reports a case of coma diabeticum in which the blood-serum contained 23 per cent, of fat. Klemperer and Umber2 state that of 9 diabetics with acidosis 7 had lipemia. Adler3 and Irnbrie4 report cases in which the blood-serum contained respectively 29 and 14 per cent, of fat plus cholesterol; of the latter there were 3.1 and 1.5 per cent, in the serum in the 2 cases. Lecithin was absent, and Imbrie found that the fatty acids entering into the composition of the blood fat had an iodin number similar to that of the fatty acids entering into the composition of subcutaneous fat. Hence, the lipemia was due to the mobilization of tissue fat.

A supposed production of fat from protein has long been believed to occur in the ripening of cheese. However, Kondo5 finds that in the process of ripening cheddar cheese 9 per cent, of the fat content disappears after thirty days and 12 per cent, after forty days.

If a fatty "degeneration" were to be found anywhere, it would certainly be looked for in the dying cells of the liver in phosphorus-poisoning, or in the analogous condition of acute yellow atrophy of the liver. But another explanation avails. Mandel and Lusk6 have shown that lactic acid disappears from the blood and urine of a phosphorized dog if phlorhizin glycosuria be induced. The writer believes that the lactic acid which occurs is derived from the sugar formed in protein metabolism. In the above case the sugar is removed without conversion into lactic acid. In phlorhizin diabetes, glucose does not burn; in phosphorus-poisoning lactic acid derived from glucose does not burn. In both cases a sugar-hungry cell, or one where carbohydrate is not oxidized, is found, and under these circumstances fat is attracted to the cell, and in larger quantities than can be useful. Wherever sugar freely burns, this fatty infiltration is impossible (p. 249). A reduced local circulation in a portion of the heart may produce anemia of the part, an imperfect local oxidation of lactic acid normally formed, and a fatty infiltration of the locality. The writer offers this hypothesis as his explanation of fatty changes in tissue in general.

1 Fischer, B.: "Virchow's Archiv," 1903, clxxii, 30 and 218.

2 Klemperer and Umber: "Zeitschrift fur klinische Medicin," 1908, Lxv, 340.

3 Adler: "Berliner klin. Wochenschrift," 1909, xlvi, 1453.

4 Imbrie: "Journal of Biological Chemistry," 1915, xx, 87. 5 Kondo: "Biochemische Zeitschrift," 1913-14, lix, 113.

6 Mandel and Lusk: "American Journal of Physiology," 1906, xvi, 129.