670 
Proceedings of the Roycd Society 
ments were now made by testing the readings of a thermometer 
when placed in different positions. The results with which we are 
at present concerned were got when the thermometer bulb was 
pushed up inside the tube, and when it was pulled down clear of 
it. When the bulb was pulled down and clear of the tube, it 
always showed about a degree lower than when in the tube. Now, 
why was this ? The thermometer and the tube were both subjected 
to the same conditions, and we might have expected that they 
would have the same temperature, that the thermometer would 
read the same inside the tube as outside, or we might have ex- 
pected it to be cooler inside the tube, as it would there have 
the advantage of the upward current ; yet it showed a higher 
temperature inside than outside. I now stopped the upward 
current in the tube to see the effect on the thermometer. When- 
ever the current was stopped, the temperature began to rise 
till the thermometer was more than 2° higher than when placed 
outside the tube. 
It was thought that the high temperature inside might be partly 
due to the heat radiated from the hot upper part of the tube. 
The apparatus was therefore taken within doors, to get a constant 
temperature, and the draught tube heated by means of a gas flame. 
No increase of temperature was, however, observed when the upper 
part of the tube was heated. We may therefore conclude that but 
little heat is radiated downwards, or at least that all the heat that 
is radiated to the thermometer is easily carried away by the upward 
current of air. Another possible explanation is, the difference in 
the absorbing powers of the surface of the tube and of the thermo- 
meter bulb. As the tube, however, was polished tin, their absorb- 
ing powers would not differ greatly. Neither of these explanations, 
therefore, seem satisfactory. The other explanation which suggested 
itself was that the difference in temperature was due to the difference 
in size of the two bodies, the thermometer bulb being about J of an 
inch in diameter, while the tube was nearly 3 inches. 
When we consider the conditions under which the thermometer 
and the tube were placed, we can easily see that the comparative 
size of these objects would have a most important influence in deter- 
mining their temperatures. When the sun is shining, all objects 
exposed to its rays get heated, and they in turn become sources of 
