Elastic Force, Density, and Temperature in Gases. 59 



was 57 0, 6 in the receiver, the barometer 29*95 inches ; after one 

 stroke of the piston completed in one second, in nine seconds 

 the thermometer attained its minimum, on the average of six 

 good consecutive experiments, of 6 0, 73 below the original tempe- 

 rature, and in the succeeding eight seconds rose, on the average, 

 1°*25. The rarefaction was sensibly the same, T yth, as with 

 common air. 



The two methods above detailed for finding the temperature 

 directly being objectionable, I laid aside Breguet's thermometer, 

 and have only resumed its use lately from having found a 

 method in which it reaches its maximum of temperature for 

 condensations in one second, as will be found further on in the 

 paper. 



On the want of success by the direct methods, in the beginning 

 of 1854 I recurred to the indirect method of observing the 

 barometer-gauge of the Newman's air-pump in the case of 

 sudden rarefactions, which is in principle the same as the 

 method of MM. Desormes and Clement, and of MM. Gay- 

 Lussac and Welter, also of Mr. Meikle, and as adopted by M. 

 Poisson in his Traite de Mecanique, vol. ii. p. 643 ; that is, by 

 finding the temperatures through the pressures, and considering 

 them connected by the law of Amontons. In this manner the 

 objection to using a light fluid having a long space to move 

 through as a barometer-gauge was avoided, so that the effect 

 of the momentum acquired in moving through a long space, but 

 opposed by the capillary attraction of watery fluids for glass, was 

 not incurred; and by noting the extremes of the first two oscil- 

 lations of the mercury in the gauge, the mean might be taken 

 as the correct result ; but still the attraction of aggregation of 

 the mercury and its adhesion to the glass are defects, preventing 

 the mercury ascending so high as it might have done. 



Assuming the accuracy of Amontons's law, let p, p, t° be the 

 pressure, density, and degrees of temperature above the freezing- 

 point of water, k and a known constants, then 



p = tcp(l + *f); 



similarly, for another state of the same gas we have 



and 



t - ^± a l°) 

 p- p (l + *f)> 



and in the experiments p and p ! are so nearly equal, that we may 



take - = 1, since the volume of gas changes only by the small 



r 



space of the barometer-gauge through which the mercury moves. 



