THE PHENOMENA OF HUMAN RESPIRATION. 
11 
commenced rising a stopwatch or chronograph was started, which registered minutes 
and seconds only. After about 35 litres of air had been collected the watch was 
stopped at the end of an expiration, and the stop-cock, turned during the next 
inspiration, the air expired being diverted into the external air. The volume of air 
collected was read off on the scale, exactly under atmospheric pressure. In some few 
instances the air was analysed at once (in which case care was taken to rinse out 
the bell-jar previously with expired air), but as a rule the bell-jar was filled twice 
or three times so as to obtain two experiments yielding as near as possible the same 
volume of air expired in a given time. In many cases the difference did not exceed 
two or tliree per cent., but it occasionally happened to be much greater. These 
occurrences were perplexing, and led to careful discussions before the figures were 
accepted. The air collected in the bell-jar was next submitted to analysis, but first 
of all its temperature was observed with a thermometer introduced in situ through the 
bell-jar, then the barometer was read.'"' 
The determination of carbonic acid in expired air and atmospheric air has taken 
up my attention for many years, and after trying various forms of volumetric methods, 
in which volumes varying from 1 to 7 litres of air expired were submitted to analysis, 
I finally fell back upon Pettenkofer’s method with which I was well accjuainted and 
which was known to yield reliable results, while simple and speedy in its manipulafions. 
It was necessary to collect the air from the bell-jar into a vessel of the proper shape and 
size, and agitate this air with a solution of barium hydrate of known strength. The 
vessel is a cylinder made of strong glass of a capacity of 1 litre, it is closed at 
both ends by brass caps, screwed over a brass flange cemented to the ends of the 
cylinder, wmshers are interposed at the points of contact so as to make the cylinder 
absolutely air-tight, a brass tube with a stop-cock is soldered to each cap. One 
of the tubes is widetied out and fitted with a thread, so as to screw to a brass 
piece fixed to a bottle of a capacity of a little over 100 c.c.; the connection being- 
secured air-tight with a washer (a drawing of the instrument is given on Plate 2), 
the bottle is closed by a wide brass cap screwing upon it perfectly air-tight, so that 
the bottle can be readily washed out and dried with a towel. The air is drawn from 
the bell-jar into the dry glass cylinder by means of a pump, the cylinder being first 
exhausted and then filled with the air from the bell-jar. The operation is repeated 
eight or ten times, when a weight being placed on the bell-jar 10 or 12 litres more of 
expired air are driven through the cylinder, which is closed by turning the tap 
while the air is in transit. Next, the cylinder still attached by india-rubber tubing 
to the pipe leading from the bell-jar is immersed in water held in a cylindrical tin vessel 
* An experiment is qnoted here in the abstract of this paper (‘ Proceedings,’ vol. 46, p. 344), towards 
testing the accuracy of the method of breathing in the bell-jars. It was found subsequently, on 
repeating it a large number of times, that the results obtained varied to such an extent, from some cause 
connected exclusively with the experiment, as to make it useless with reference to the object in view. 
Hence it is omitted on the present occasion. 
c 2 
