264 REPORTS ON THE STATE OF SCIENCE. 
APPENDIX A. 
REGNAULT’s CORRECTIONS. 
Some explanation of Regnault’s methods is necessary in order to make 
the following extract clear to those who are not familiar with his paper. 
The air was passed in a continuous stream through a pipe in a bath of 
heated oil and took the temperature of the oil. It then traversed a 
short pipe into a calorimeter, and in its passage through the calorimeter 
it took the temperature of the water therein. The rate of rise of tem- 
perature of the water in the calorimeter was observed over an interval 
varying in different experiments from five to forty minutes. It was 
assumed that the heat lost by the calorimeter to its surroundings and by 
conduction along the connecting pipe was such as to lower its tempera- 
ture at the rate A@ where: 
A§=A(0—t)+K. 
is the temperature of the calorimeter, ¢ that of the surrounding air, 
and A and K are constants independent of the rate of flow of air. The 
constant K represents the rate of flow of heat along the connecting pipe. 
A and K were determined by two observations of the change of tempera- 
ture in the calorimeter which took place when the air current was 
stopped during two periods of ten minutes which immediately preceded’ 
and followed the experiment with air flowing. 
The correction A@ to be applied at each instant during the experi- 
ment to the observed rate of rise in 6 was then calculated from the 
observed values of 6 and t. 
Regnault tested the correctness of the above assumption by making 
a series of determinations of the specific heat of air with currents of 
different velocity. He found that the apparent specific heat was practi- 
cally constant over a wide range, extending from 10 grammes per 
minute to 30 grammes per minute. If the rate of flow were outside 
this range the apparent specific heat was less. In the case of the slower 
currents this was doubtless due to an error in the correction. The 
faster currents gave wrong values because the air had not time to take 
up the temperature of the oil-bath and of the calorimeter respectively. 
There is a good illustration of Regnault’s apparatus in Haber’s 
‘Thermodynamics of Technical Gas Reactions,’ p. 212. 
Mémoires de l’Académie des Sciences de l'Institut Impérial de France, 
tome xxvi., p. 83. 
On peut conclure de ces expériences que la formule 
Ad=A(O—t) + K 
dont les constantes ont été calculées, pour chaque expérience, d’aprés 
les éléments observés pendant la premiére et la derniére période, 
peut étre employée, avec toute confiance, pour calculer les effets produits 
par les causes perturbatrices pendant le temps ot le courant gazeux 
traverse l’appareil. I] est nécessaire, néanmoins, de faire sous ce rapport 
une réserve, car il se présente ici une cause d’incertitude que j’ai vaine- 
ment cherché 4 éliminer, et dont je n’ai pas réussi 4 calculer les effets 
avec précision. Pendant la premiére et Ja derniére période de 
Vexpérience, on observe les variations de température sous l’influence 
