254 Proceedings of the Royal Society 
liquid in the cavities being possibly less diathermanous than the 
Iceland spar, and absorbing the beat transmitted to it by radiation 
through the crystal ? 
In order to try if the motion of a vapour bubble could be ex- 
hibited on a larger scale, I made use of a hermetically sealed tube 
containing liquefied sulphurous acid (sulphur dioxide) which I had 
some time ago prepared to show the high dilatability by heat of 
that liquid. When the tube was placed horizontally the void space, 
like the bubble of a spirit level, was about 15 inches long ; and I 
found that its extremity moved towards the point where a piece of 
heated brass was applied to the tube. I then nearly filled a tube 
with ether made from methylated alcohol ; and after heating the 
top, so as to vaporise the ether and expel the air, I hermetically 
sealed the tube. Placing the tube horizontally, the vapour bubble 
is about 0-3 of an inch long ; and when a finger is put on the tube 
about 0’25 of an inch from the bubble, in a little while the bubble 
moves towards the finger with a rapidly accelerated motion, and 
places itself in a position of stable equilibrum under the finger, 
about which it slightly oscillates even after the finger is removed 
from the tube. A piece of metal too hot to be touched acts still 
more energetically. 
I have thought it proper to note that the ether I used had been 
made from methylated alcohol, because in exhibiting as a lecture 
experiment Dalton’s method of measuring vapour tensions, I have 
found that ether made from methylated alcohol seems to show a 
higher vapour tension than that of ether as determined by Regnault. 
This is probably due to the presence of some other substance more 
volatile than common or diethyl ether, possibly to a portion of 
dimethyl ether whose boiling-point is so low as - 21° C. 
United College, St Andrews, 
nth Mai f 1873 . 
