THE PEIENOMEHA OF HUMAN RESPIRATION. 
15 
contact with the contents of the flask ; this precaution was found absolutely necessary 
to insure accurate results. 
5. The analyses of the alkaline solution were made the day after the experiment 
except when a Sunday intervened. I have found it necessary not to delay the analyses 
of atmospheric air, a precaution peidiaps less important when a large proportion of CO 3 
is present, as occurs in air expired from the lungs. 
6. The standard solution of oxalic acid used in Pettenkofer’s method usually 
undergoes a change after about a fortnight, when mould is formed in the fluid. 
In order to avoid any error from this cause, the solution was prepared very carefully 
by heating the dry flask in which it was to be made to a sufficient temperature to 
destroy the bacteria, and allowing it to cool with its opening plugged with cotton 
wool. The weighed oxalic acid was rapidly transferred to the flask, and the solution 
made with fresh boiled distilled water. The solution prepared in this way keeps for 
a much longer time unaltered in strength, than if these precautions are not taken. 
There is practically I believe, only one source of error in Pettenkofer’s 
method of analysis—from the uncertainty of the acidity of the oxalic acid used, as 
equal weights of different samples do not always give exactly tne same degree of 
acidity. This source of error can be easily obviated by substituting hydrochloric acid 
for oxalic acid. The analyses referred to in this paper have all been made with oxalic 
acid from the same stock kept in a glass stoppered bottle. 
In order to make cpiite certain of the strength of the solution, a standard solution 
of pure hydrochloric acid was prepared by titrating it with a weighed quantity 
of anhydrous sodium carbonate. The alkalinity of the barium solution wms then 
determined with the standard oxalic acid, and also with the standard hydrochloric 
acid, while a known volume of the same barium solution was precipitated with 
sulphuric acid, and determined by weight as barium sulphate. The results obtained 
were as follows :— 
With oxalic acid 100 c.c. barium solution w’ere found 
equivalent to.0’2683 grins. CO 3 
With hydrochloric acid, 100 c.c.0'2G7 5 ,, 
By weight, 100 c.c.0'2679 ,, 
It was therefore obvious that the oxalic acid used in these analyses was quite 
pure, and moreover that the barium hydrate was uncontaminated wdth any other 
alkali. 
It has been long known that a certain proportion of the oxygen of the air inspired 
is not returned in the expired air as carbonic acid, and Pegnault and Beiset in 
their admirable experiments'^ have shown that a portion of the oxygen inhaled is 
occluded in the body. While experimenting on dogs they found the mean relation 
between the weight of the oxygen expired and i.he total weight of oxygen consumed 
* ‘ Annales de Chimie efc de Physique,’ 1849, vol. 26. 
