PHYSICAL AMD CHEMICAL NOTES. 97 
atmospheric pressure contained in a vessel, the outer surface of 
which is glass, but the inner surface is a film of water con- 
tinually renewed by steam condensing on it. So long as the full 
supply of steam is kept up, this continually renewed film of water, 
in contact with saturated steam, is a perfect protection against varia- 
tion of temperature in the interior of the steam tube. When the 
steam first passes through the tube there is rapid condensation on 
the sides, and after that, on the thermometer, from which the water 
falls in a series of drops, which follow each other first slowly, then 
rapidly, then again more slowly, until, finally, a drop remains hang- 
ing from the lower extremity of the thermometer, and never falls. 
By that time the protecting film of water has established itself on 
the sides of the tube, and effectually guards the thermometer from 
external influence. If it were possible for the temperature of 
the thermometer to fall ever so little below that of the condensing 
temperature of the saturated steam in the tube, its temperature 
would be immediately restored by the condensation of some of the 
steam on the water which moistens its surface. If its temperature 
were to rise ever so little above the condensing temperature it would 
be immediately brought back again by the evaporation of some of 
the moisture on it. Consequently, when the boiling is in full 
operation, the thermometer is exactly at the temperature when the 
smallest possible increase of heat will cause evaporation, and the 
smallest possible decrease of heat will cause condensation. But the 
boiling-point of a substance is the temperature at which tt, as a vapour, 
condenses on itself as a liquid, and as a liquid evaporates into itself as 
a vapour. Therefore, the temperature of the thermometer is exactly 
the boiling temperature of the water. Further, the whole enclosure 
is guarded by a surface of water in contact with saturated steam, so 
that its walls are necessarily also at the boiling temperature of the 
liquid, and it is impossible that the thermometer can be at any other 
temperature than that of the boiling liquid and condensing vapour. 
It follows, therefore, that for this purpose we may graduate our 
thermometer into thousandths of a degree, if we choose, and it is 
quite certain that the temperature of the thermometer will not differ 
by that amount from that of the medium. During an operation the 
steam tube is the ideal enclosure at constant temperature. When 
the thermometer has taken the temperature of the steam, it remains 
perfectly steady so long as the barometric pressure remains the same, 
As evidence of the efficiency of the steam tube, the rate of gene- 
ration of steam was varied from the highest to the lowest which was 
possible with the lamp. The thermometer divided into fiftieths of a 
Hl 
