The temperature was taken in the mouth every morning while the men were lying on the couch preparatory to respiration observations at Springfield. These temperatures will not be reported, as they are physiologically of no significance other than to demonstrate that the temperatures were normal. Toward the latter part of the research it was made a general rule to obtain the rectal temperatures when the men were inside the large respiration chamber in Boston, just prior to their rising in the morning. These were made with 12 sensitive clinical thermometers, calibrated with a standard thermometer made by Rich-ter of Berlin.

During the last few weeks of the research an apparatus for recording the surface temperatures of the skin was installed in the library at the Nutrition Laboratory, with which the temperature of the forehead and of the backs of the hands was taken with a thermo-couple. This thermo-couple consisted of a copper-constantan junction, one end of which was kept in a Dewar flask at approximately 34° C., the temperature being read to hundredths. The other end of the junction, which was in the shape of a hairpin and protected with a thin layer of cotton wadding to prevent the influence of external temperature, could be applied to the surface of the skin. The thermo-junctions were connected with a sensitive Deprez-d'Arson val galvanometer and the readings taken. It was soon found that the deflections of the galvanometer pursued a clearly characteristic course and a point could easily be found which represented in all probability the true skin temperature. These observations were made for all the subjects at various times and quite frequently were controlled by measurements obtained on several members of the Laboratory staff. Sometimes, also, visitors were measured to assist in the interpretation of the results.