A second experiment with a dog of 5.8 kilograms, fed on meat, fat, and rice, led to essentially the same results as the preceding experiment. At the end of the first month, there appeared indications that the animal was not well, loss of appetite being marked, with disturbance of the stomach accompanied by occasional vomiting. These symptoms disappeared quickly when the animal was given for a few days large quantities of meat. On returning to the original low proteid diet, with its large content of rice, the symptoms gradually reappeared. At the end of two months the animal had again lost its appetite, and before the end of the fifth month the subject was dead. Post-mortem examination showed especially a strong fatty degeneration of the epithelial cells of the mucous membrane of the stomach and intestine. Rosenheim concludes that a diet poor in proteid is unhealthful for dogs, and that a daily ration containing even 0.32 gram of nitrogen per kilogram of body-weight, and with a fuel value of 110 calories per kilo, is not sufficient to maintain the animal in a condition of health. The next series of experiments was made by Jägerroos[70] of Finland. This investigator was evidently impressed by the unfavorable and monotonous character of the diet made use of by the preceding investigators, and sought to introduce a little variety, recognizing also that with a carnivorous animal it is difficult to reduce the proteid to a low level and maintain the necessary fuel value, without introducing foodstuffs to which the animal is wholly unaccustomed. In the first experiment, the dog had a body-weight of 5.77 kilograms, and at the beginning was fed daily 40 grams of meat and 100 grams of sugar, equal to 0.31 gram of nitrogen and 80 calories per kilo of body-weight. The experiment continued for eight months, sugar being replaced in part by butter, and occasionally bread, fat, and wheat meal being used in proper amount to yield the given nitrogen and fuel values. During the last five months, the intake of nitrogen per day averaged 0.29 gram per kilo, with a fuel value amounting to 89 calories daily per kilo of body-weight. During this period, the animal maintained a plus nitrogen balance for a large part of the time. The experiment was then continued for two months longer, with a gradual diminution in the nitrogen of the food and in the fuel value, the animal dying at the end of the tenth month. [70] B. H. Jägerroos: Ueber die Folgen einer ausreichenden, aber eiweissarmen Nahrung. Skandinavisches Archiv für Physiologie, Band 13, p. 375, 1902.