METHODS OF INVESTIGATION. 79 
allowances for the hydrogen gained or lost in protein and fat, to 
compute the gain or loss of water by the body. 
With the earlier forms of respiration apparatus, great diffi- 
culty was experienced in obtaining satisfactory results for the 
water,* and Stohmann { has traced the difficulty to an invisible 
condensation of water on the walls of the chamber and connections. 
More recently Rubner { has been able to make satisfactory deter- 
minations of water with a Pettenkofer apparatus by avoiding as 
much as possible differences of temperature between different parts 
of the apparatus and by taking the sample of the outcoming air for 
analysis as close to the respiration chamber as possible. Atwater 
& Rosa have shown that their form of Pettenkofer apparatus 
(p. 72) permits of very accurate determinations of water. 
OxycEN Batancr.—Owing to the technical difficulties already 
indicated in considering the different types of respiration apparatus, 
direct determinations of the oxygen balance have rarely been made. 
_ This is the more to be regretted since such a determination would 
~erve to check those of nitrogen, carbon, and hydrogen, and would 
be a test of the accuracy of our deductions from those determina- 
tions as to the nature of the material gained or lost by the body. 
Asu INGREDIENTS.—The gain or loss of ash ingredients can of 
course be readily determined, but the subject as yet has hardly 
received the attention which it deserves. 
SULPHUR AND PHosPHoRus.—Sulphur forms an essential con- 
stituent of the proteids, while phosphorus enters into the composi- 
tion of the nucleins and also of lecithin. The determination of the 
income and outgo of these two elements is often of value in rela- 
tion to special physiological questions, but from the somewhat. 
general standpoint of this work may be considered as of rather 
minor importance. 
* Zeit. f. Biol., 11, 126. 
+Landw. Vers. Stat., 19, 81. 
t Arch. f. Hygiene, 11, 160. 
