apparatus was used, represented in figure 1. The hydrogen sulphide 
was prepared from a solution of natrium sulphide, which was free 
from carbonic acid and diluted sulphuric acid; the former solution 
was obtained by saturating natron, which had been freed from 
its earbonic acid by the addition of barium hydroxide, with hydrogen 
sulphide. The gas prepared in this way was dried by phosphoric 
anhydride, and led into the exhausted vessel 5 through the cock 
A (fig. 1); it was then freed from air by condensation in O by 
means of the two vessels ?, and /, which were filled with carbon 
of cocoa-nut; OY, P,, and P, had been placed in liquid air for this 
purpose. If now there were no discharges at all in the GeIssLEr 
tube .V, the hydrogen sulphide was again conveyed from © into 
the reservoir 4 by evaporation. Then the gas which had remained 
in the tubes of the apparatus between J/ and £ could be sucked 
off by means of a water-jet-pump, when the cock C was opened. 
The Cailletet tube, which was made of common glass for this 
experiment, where the pressures did not exceed 25 atmospheres, and 
which could have a comparatively large bore (5 mm.), was sealed 
to a glass spring “4, and thus connected with the apparatus. 
Now a quantity of mercury freed from air by boiling which 
was sufficient to fill the entire test-tube ), was brought into the 
reservoir H through G. After G had been sealed, /” was provided 
with a little distilled water, and the upper end of this tube was 
also sealed. Now the water in # was frozen, and the test-tube 
was evacuaied by means of the water-jet-pump (cock C°) and carbon 
(Ff, and P,). After the cock 4 had been closed, the water could 
now be distilled over from # to the upper end of the test-tube, 
which had been cooled with liquid air (vessel S). 
A quantity of hydrogen sulphide was admitted from Z into the 
apparatus between J/ and / with open vessel 0; by means of the 
manometer M, whose rightside leg had been evacuated, this quantity 
could be roughly estimated. This gas too was solidified in the upper 
end of the test-tube when / was opened. 
When I had then convinced myself that no air was present in 
the test-tube (discharge in \), the Cailletet tube was tilted, in con- 
sequence of which the mercury flowed from H into the U-shaped 
lower end of the test-tube, and was forced up to the top of the test- 
tube by the air in consequence of the opening of the cocks C' and 
L. Now the Cailletet tube was separated from the apparatus at K, 
and placed in the pressure cylinder which was filled with mercury 
purified and freed from air by boiling. 
When we work in this way the presence of air is practically 
54 
Proceedings Royal Acad. Amsterdam. Vol. XIII. 
