19 
The test for nitrites was made in the usual way. A few cubic cen- 
timeters of a solution of metaphenylene-diamine in dilute sulphuric 
acid is placed in a test tube; a few drops of the water to be tested are 
allowed to fall into the test tube: if nitrites are present, a yellow color 
is produced. 
This test is said to show nitrites when present in as dilute a solution 
as 1 part in 3,000.000, but in no case was any appreciable evidence of 
nitrites obtained in the water condensed in the freezers. In only one 
case was there sufficient coloration to warrant the statement that there 
was even a trace of nitrites present. Singularly enough, while the 
respired water condensed in the freezers gave no response to the test 
for nitrites, that condensed on the cooling apparatus in the respiration 
chamber gave a very marked nitrite reaction. 
The test for nitrates was made by means of a solution of diphenyl- 
amine in rather strong sulphuric acid. Two or three cubic centi- 
meters of this solution is placed in a test tube and about one-half a 
cubic centimeter of the water to be tested is allowed to flow slowly 
down the side of the test tube so that the two liquids do not mix. If 
a deep- blue ring appears at the contact zone of the two liquids, the 
presence of nitrates is shown. The results of the test for nitrates were 
very marked, but there appeared to be little difference between the 
amount in the water of the incoming and outgoing ventilating air 
current. After making a check experiment with the alcohol lamp 
burning inside the chamber the amount of nitrates in the outgoing 
freezer water appeared to be materially increased. This is not sur- 
prising, as the high temperature of the alcohol flame might naturally 
be expected to bring about a certain amount of combustion of the 
nitrogen of the air. 
The test for nitrates and nitrites in the condensed water of respira- 
tion indicated that the actual amounts of these were so minute as to 
be ignored. It seems highly improbable that an}' measurable quantity 
of nitrogen can be excreted in this manner. 
RECORDS OF BODY WEIGHT, TEMPERATURE, AND PULSE RATE. 
The subject weighed himself on a platform scale sensitive to 10 
grams with a weight of 75 kilograms and capable of weighing 100 
kilograms. 
The observations of body temperature were made by the use of an 
ordinary clinical thermometer. They were sublingual in experiments 
previous to. and axillary in experiments subsequent to. No. 25. In 
this latter experiment temperatures were measured in both ways. All 
observations were, of course, made by the subject himself. Under 
such circumstances it is not improbable that there may have been 
occasional errors in the thermometer readings; at best, such observa- 
tions would be less satisfactory than those which would show the 
actual internal temperature of the body. In later experiments, the 
