206 
PROFESSOR TYNDALL ON THE ABSORPTION AND 
Comparing these results with those obtained with olefiant gas, we find for thick- 
nesses of 0‘05 of an inch and 2 inches respectively the following absorptions : — 
Olefiant gas. Sulphuric ether. 
Thickness of 0'05 . . . 5’37 Thickness of 0 - 05 . . . 2'07 
Thickness of 2 inches . . 32 ’80 Thickness of 2 inches . . 35 T 
Sulphuric ether vapour, therefore, commences with an absorption much lower than 
that of olefiant gas, and ends with a higher absorption. This is quite in accordance 
with the result established in my second memoir*, that in a short tube the absorption 
effected by the sparsely scattered molecules of a vapour may be less than that of a gas at 
a tension of an atmosphere, while in a long tube the gas may be exceeded by the 
vapour. The deportment of sulphuric ether indicates what mighty changes of climate 
might be brought about by the introduction into the earth’s atmosphere of an almost 
infinitesimal amount of a powerful vapour. And if aqueous vapour can be shown to be 
thus powerful, the effect of its withdrawal from our atmosphere may be inferred. 
§2- 
The experiments with the piston apparatus being completed, greater thicknesses of 
gas were obtained by means of the composite brass experimental tube already referred 
to. The arrangement adopted was, however, peculiar, being expressly intended to check 
the experiments, which were for the most part made by my assistants. The source of 
heat and the front chamber remained as usual ; a plate of rock-salt dividing, as in my 
previous investigations, the front chamber from the experimental tube. The distant end 
of the tube was also stopped by a plate of salt ; but instead of permitting the tube to 
remain continuous from beginning to end, it was divided, by a third plate of rock-salt, 
into two air-tight compartments. Thus the rays of heat from the source had to pass 
through three distinct chambers, and through three plates of salt. The first chamber 
* Philosophical Transactions, Part I. 1862 ; and Philosophical Magazine, vol. xxiv. p. 343. 
