116 Professor Piazzi Smyth's Experiments on the 



But before reaching this, the air had to pass the bulb of a 

 thermometer in the middle of the pipe, and pass the opening 

 into a mercury gauge close by : thus giving its temperature 

 under compression, and the degree of that compression. 



On opening the 2d stopcock the air rushed out into the 

 atmosphere, striking, in so doing, on the 2d thermometer, 

 placed just within the exit mouth of the tube ; a comparison 

 of the 2d thermometer with the 1st should therefore 

 shew the lowering of temperature on air escaping from 

 xjompression, and by determining the temperature of the air 

 entering the valves of the pump below, and comparing that 

 with thermometer 1, or the temperature in the reservoir, the 

 increase of heat on air undergoing a certain compression 

 tjould be determined. 



These temperatures and pressures were not however to be 

 determined so very easily, by the mere reading off the ther- 

 mometers in any way ; for 



\st. With regard to the temperature of the air entering the 

 pumps, if the thermometer was held too far from the valves, 

 it was not sufficiently identified with the stream of entering 

 air; if too close, some of the compressed air would escape 

 out of the cylinder on the closing of the valves, and so in- 

 fluence the height of the thermometer. But after holding 

 the thermometer in various positions, and taking readings 

 through a considerable length of time, it was concluded that 

 63° would represent the temperature without a greater error 

 than +r. 



2d, "With regard to the temperature of the air, on having 

 passed the force valve of the pump, or on having entered the 

 reservoir. This might have been best determined by a long 

 thermometer placed in the middle of the reservoir ; whereas 

 our thermometer No. 1 intended for this purpose was in a 

 tin tube some two feet long and only one inch in diameter ; 

 the consequence was that when the 2d stopcock was closed, 

 and there was no stream of condensed air through the pipe, — 

 but it was merely then in a quiescent state, — it was rapidly 

 acted on by the temperature of the room generally into which 

 the pipe protruded out of the reservoir. The real tempera- 

 ture of the reservoir was therefore best obtained by opening 



