The source of the feces was further investigated by Hermann,2 whose work was later elaborated by Fritz Voit.3 The latter separated a loop of the intestine about a third of a meter long from the rest of the intestine of a starving dog. Both ends of the loop were tied and the loop remained in the abdomen in connection with its normal nerve and blood supply. The two ends of the remaining portion were reunited. After a few days food could be given and the normal excretion of feces took place. After three weeks the animal was killed. It was found that the isolated loop contained a thick, fecallike mass. It was found that the dry solids of this mass contained the same percentage of nitrogen as did the feces passed by the dog during the three weeks of the experiment. It was also calculated that the amount of nitrogen excreted through the wall of the intestinal loop was nearly the same per unit of area as the amount of nitrogen in the feces when spread over the surface of the whole of the rest of the intestine. The following table shows this:

1 von Muller: "Zeitschrift fur Biologie," 1884, xx, 340.

2 Hermann: "Pfluger's Archiv," 1890, xlvi, 93.

3 F. Voit: "Zeitschrift fur Biologie," 1892, xxix, 325.

The loop contained fat and fatty acids in greater quantity than is normally found in feces, which may indicate a usual reabsorption of these substances.

Fritz Voit has therefore shown that the excretion of substances from an isolated loop of the intestine produces a mass of a similar constitution and of nitrogen output equal to that in the normal intestine of the same animal through which meat and fat were passing. He therefore concludes that the feces are derived principally from the substances excreted through the wall of the intestine. The nitrogen so excreted is as much to be considered a product of protein metabolism as is the nitrogen of urea. It is regretable that very little is known regarding the chemistry of these nitrogenous compounds excreted into the intestine.

It has been seen that the feeding of simple food-stuffs, such as meat, fat, and sugar, scarcely influenced the composition of the feces in the dog. In herbivora we pass to another extreme. Here vast amounts of cellulose are eaten, a great part of which is never disintegrated, but even after long retention in the capacious intestinal tract is passed in the feces. After giving an ordinary feed to a cow one may find as much nitrogen in the feces as in the urine. Under such conditions as these the very voluminous feces evidently do consist largely of the undigested residues of the fodder. Armsby and Fries1 have shown that only 45 per cent, of the energy contained in hay is of actual use in cattle feeding. The waste in the feces reaches 41 per cent., in the urine 7.25, and in methane gas 6.75 per cent, of the total energy content.

Concerning the fecal production in man, it has been found that Cetti2 excreted 3.8 grams of dry fecal solids per day during a fast of ten days, Breithaupt 2 grams, and a medical student3 2.2 grams, less in reality than would a dog of similar size. Benedict4 states that he was unable to find any evidence of the formation of feces during a seven-day fast in man.

Rieder5 gave a man a diet containing starch, sugar, and lard from which a cake was baked. The food contained no nitrogen, but the fecal excretion was 0.54,0.87, and 0.78 gram of nitrogen per day, contrasting with 0.316 gram from Cetti, 0.113 from Breithaupt, and 0.13 from a medical student during fasting. The food, even though it contains no protein, stimulates the fecal production.

Wallace and Salomon6 have administered 250 grams of cane-sugar daily to normal persons and to patients suffering from intestinal diarrhea, and have determined the amount of fecal nitrogen during periods of two or three days. The sugar was given in doses of 50 grams dissolved in 300 c.c. of water and flavored with fruits, such as apple and lemon, or with wine. Their results with this diet were as follows:

1 Armsby and Fries: Bulletin 101, 1908, Bureau of Animal Industry, U. S. Dept. of Agriculture.

2 Lehmann, Muller, I. Munk, Senator, Zuntz: "Virchow's Archiv," 1893, Bd. cxxxi, Suppl. Heft.

3Johansson, Landergren, Sonden, Tigerstedt: "Skandin. Archiv fur Physiologie," 1897, vii, 29.

4 Benedict: "Influence of Inanition on Metabolism," Carnegie Institution, 1907, P. 345.

5 Rieder: "Zeitschrift fur Biologie," 1884, xx, 378.

6 Wallace and Salomon: "Medizinische Klinik," 1909, v, 579.

It is evident that the quantity of fecal nitrogen eliminated in intestinal diseases is largely increased.

It has been stated that Voit early noticed the occurrence of starch particles in the feces. A large number of experiments have been made to test the digestibility of the various vegetables and cereals. Rubner1 fed an able-bodied soldier on 3078 grams of variously cooked potatoes daily and found pieces of potatoes in the feces. He notes that an inhabitant of Ireland will eat 4500 grams of potatoes a day. Friedrich Muller2 writes that after the ingestion of a large quantity of bread the feces may have practically the same composition as bread.

The better understanding of this question of the digestibility of the carbohydrates has come through the work of Prausnitz3 and his associates, Moeller and Kermauner. Moeller found that no starch appeared in the feces after feeding well-cooked white, rye, and graham bread, rice or potatoes (even when fed in pieces), or legumes when they were prepared in the form of puree. Legumes not in the form of puree, such as string beans eaten as salad, may resist the action of the digestive juices so that the starch contents of the cell are untouched, and the vegetable cells appear in the feces. These facts explain the appearance of bread in the feces if the bread be badly cooked, or if such a "heavy" bread as pumpernickel be eaten. The imperfectly cooked bread contains starch granules whose coverings are impermeable to the digestive juices, as are also many of those in the unbolted rye of pumpernickel.

1 Rubner: "Zeitschrift fur Biologie," 1879, xv, 146.

2 Fr. Muller: Ibid., 1884, xx, 375. 'Prausnitz: Ibid., 1897, xxxv, 335.

Prausnitz finds that if a man be put on a rice diet and then meat be substituted for most of the rice, the composition of the feces does not vary with the diet. Such feces he calls normal feces. They may contain a negligible quantity of fibers of meat (Kermauner) or of cellulose from the rice.

The feces of 6 persons placed alternately on meat and rice diets yielded normal feces, the percentage composition of the dry solids of which was as follows:

Composition Of Feces On Different Diets

It is seen from this that whether the food solids contain 1.5 per cent. N, as in rice, or ten times that, as in meat, the composition of the feces remains uninfluenced. Normal feces result from the eating of any food which is completely digested and absorbed. In all such cases these feces have the same composition and are derived from the intestinal wall. It is therefore not astonishing that a vegetarian of many years' standing produced the same kind of feces when fed on rice as did the other men. The same quality of feces has been obtained after giving good bread.