438 
passing in, is, adopting the Newtonian law, proportional to 
the rise of temperature at such instant. But the gas having 
been always introduced in my experiments at a uniform rate, 
the rise of temperature is proportional to the time. Hence 
the velocity of cooling at any instant is proportional to the 
time. Such being the case, the well-known theorems, which 
relate to the motion of a material point actuated by a constant 
force, are here strictly applicable; and, amongst the rest, that 
the space (number of degrees) through which the cylinder 
cools in the time m, is equal to half the rectangle under the 
time and the last acquired velocity. This theorem, in fact, 
immediately gives the correction in question, not, I may ob- 
serve, in an approximate, but in a complete manner, and, in 
practice, 1 have every reason to be satisfied with it. 
‘¢ In what precedes it will be seen, that I have employed 
the Newtonian law of cooling, which the researches of Dulong 
and Petit have shown not to represent observations with rigour, 
except when the excesses of temperature are small. My results, 
however, are not on this account appreciably less accurate, for 
the thermometer which I employed only read to tenths, and 
the divergence of the Newtonian law from the truth, within 
the range of my experiments, is only observable in the second 
decimal place. 
‘* Having explained every thing necessary to enable the 
Academy to judge of the accuracy of my results, I shall now 
state the numbers at which I have arrived: 
: i Equal weights. An atom. 
Ammoniacal gas passed into water, 940° 940° 
Muriatic acid gas passed into water, 885° 1900° 
Weight for weight, then, ammonia gives out more heat than mu- 
riatic acid; but an atom of the latter gives out almost exactly 
the double of the heat evolved by an atom of the former. 
‘¢ The number for ammonia, it will have been observed, 
does not materially differ from that for aqueous vapour of maxi- 
mum density at 212°, the latter having been fixed, by the 
ped ie 
