} Quantity of Heat evolved from Atmospheric Air. 227 
| According to these yngicog and deductions, while the 
densities of air increased in the geometrical progression 2, 4, 8, 
the heats evolved corresponded, nearly, to the arithmetical series 
} , 5. But the ratio in the differences of temperature between 
the assigned densities followed a very different rate of progression 
from either; thus for the densities 2, 4, 8 atmospheric pressures, — 
the corresponding differences of heat evolved were in the d 
; creasing numbers, nearly, (277, 118, 80) 3:5, 1:5 
I deem it highly probable that the foregoing deductions are . 
| very near to true expositions of the relations between the con 
densation of atmospheric air, and the evolution of its latent heat: 
-The discrepancies between the actual observations and the num- 
bers which should belong to a law of progression, are more rea 
sonably to be referred to error in the former than in the latter. — : 
A slight examination of the tables will shew how very important =. 
it is in such experiments, that the instruments for making obser- 
vations should be both delicate and accurate. It has been men- — 
tioned that the thermometers used could not - relied upon within 
| alf a degree of their indications ; while an error to this extent 
would produce a difference in the deductiouié*of the first line of _ . 
Table V, of one-tenth part and, of course, materially affect the. . 
value of the experiment. Besides the necessity of proper philo- | 
sophical ee the series of observations on both densities & 
and temperature ought to have been more extended to justify, the _ 
attempt to establish a natural law from their deductions. So 
well aware was I of the importance of an extended succession of 
observations, that in drawing up the design of the machine, I 
endeavored to adjust its proportions so that it ee Pe capable 
of working to a tension of sixteen atmosphe res; b nel trial, 
many portions of it were found too weak to sustain suclta pres- . 
sure. I will, however, repeat that such exper ts ast he ma- 
chine admitted of bein ng made, have been ents 
Viction of their importance, and with a SR fet that ie 
ing should be noted which my own senses 
set down but as the result of re 
and, Gar thstanding, pee im rfectic n| 
| chiner am o tae hes 
siderable le"extent, nor can a hia recom m 
. contidence. i ae 
Apalachico 
ola, 
March 7 1850. 
