616 
PROFESSOR GRAHAM ON THE MOTION OF GASES. 
Repetition of the last experiments. Barom. 29 6 16 — 29*666. Temp. 7 2 °. 
Gauge barometer in inches. 
Air. 
Oxygen. 
Carbonic acid. 
I. 
II. 
I. 
II. 
I. 
II. 
n 
// 
// 
// 
/7 
// 
28-5 
0 
0 
0 
0 
0 
0 
26-5 
237 
236 
262 
262 
196 
197 
24-5 
239 
239 
266 
266 
199 
202 
23-5 
129 
129 
141 
141 
107 
106 
From 28-5 to 23*5 inches 
605 
604 
669 
669 
502 
505 
Mean Results. 
Gauge barometer. 
Air. 
Carbonic acid. 
From 28-5 to 23*5 inches { T " '*'** 
[ lime oi oxygen =1... 
604-5 
0-9036 
503-5 
0-7526 
The three results for carbonic acid by this capillary are therefore 0*7597, 0*7550 
and 0*7526 ; of which the mean is 0*7558. The coefficient of capillary E for car- 
bonic acid is therefore not so widely different from 0*75 as it at first appeared. 
With these large aspirator jars the rise of temperature within the jar during the 
experiment becomes very small. In the six-pint jar it did not amount to one-quarter 
of a degree Fahr., and in the nine-pint jar it was altogether insensible. 
I have now detailed the results of the transpiration of gases by all the capillaries 
used except one, — a tube with which a few preliminary experiments in the inquiry were 
made, not sufficiently precise to merit being recorded. All of these tubes have given 
the same coefficient of transpiration to each gas, or coefficients closely approxi- 
mating, although the tubes themselves have varied considerably in their respective 
dimensions ; namely, in diameter, from ^-th to y^-gth °f an inch 5 in length from 2 
inches to 22 feet ; and from 12 to 70 seconds in the time of transmission of one cubic 
inch of air into a vacuum. It can be said that no selection of the tubes was made ; 
and their dimensions are in a great measure accidental. 
From the agreement in results obtained with tubes so different in dimensions, 
I consider that a glass tube of any diameter whatever will be found suitable for 
observations on transpiration, provided a certain length is given to it. If the tube 
is extremely short, a mere ring, then we know that gases will be transmitted by it 
into a vacuum according to the law of effusion ; oxygen, hydrogen and carbonic acid 
in times expressed by 1, 0*25 and 1*176 respectively. With the slightest elongation 
of the tube these ratios are disturbed ; the number for hydrogen socm increasing to 
0*35 or 0*40, and that for carbonic acid falling to 1, or 0*80, referring both to the 
time of oxygen always taken as unity. The change with the increase of length is 
very great at first, but soon falls off, and with a certain length of tube seems to cease 
altogether. The coefficient of hydrogen is then found to have risen to some number 
