The entire floor under the dogs was composed of metal, the joints all soldered, the floor being sloped to a metal gutter in front so that all the compartments could be flushed out each morning and kept sweet and clean. In pleasant weather, immediately after their first meal, the dogs were taken out of doors to a large enclosure near by, where they were allowed perfect freedom until about four o’clock, when they were taken in for their second meal (between four and five o’clock in the afternoon). The outdoor enclosure was inaccessible to every one except the holder of the key, and the dogs while there were wholly free from annoyance. Once every month, during a period of ten consecutive days, each dog was confined in the metabolism cage so as to admit of the collection of all excreta, in order to make a determination of the nitrogen balance. Practically, therefore, each dog was in close confinement only one-third of the month, the remaining two-thirds being spent in much more congenial surroundings. I have entered thus fully into a description of the conditions prevailing, because I deem them exceedingly important, and because therein undoubtedly lies the explanation of the striking contrast between our results and those of the earlier investigators of this subject. In considering the outcome of our experiments, it may be wise to enter into some detail concerning the first case to be presented. The animal employed in this experiment was designated as No. 5, and weighed on July 27, 1905, 17.2 kilograms; it was apparently full grown, but was thin and had the appearance of being underfed. At first, it was given daily 172 grams of meat, 124 grams of cracker dust, and 72 grams of lard, the day’s ration containing 8.66 grams of nitrogen and having a fuel value of 1389 calories.[72] These figures are equivalent to 80 calories, and 0.50 gram of nitrogen, per kilogram of body-weight. The animal took kindly to the diet, but on August 3 it refused to eat and seemed to have a little fever. The next day it was better, but for the three following days its appetite was poor, and only a portion of the daily food was eaten. Body-weight began to fall off, and was soon at 15.5 kilograms. On the 7th of August, a dose of vermifuge was given, after which the appetite returned and the animal appeared in good spirits. From this time forward it seemed in perfect health, with good appetite, and showed the usual vivacity and playfulness of dog-kind. The diet as specified was continued unchanged until August 25, a balance experiment covering a period of ten days, from the 15th to the 24th of August inclusive, being carried out, in which the nitrogen of the intake was compared with the output for each day. From the accompanying table, where are given the average values of all the balance periods of the experiment, it is to be seen that during this first period the animal was laying on or gaining an average of 2 grams of nitrogen per day.