48 | oy John Garric on the Quantity Ka Heal; a tg ‘ 
upon the subject, I have fe ave the quantity” ‘at the generally te 
ceived one of 140° F'., though some observations incline me t 
consider it as several degrees less. 
With the foregoing apparatus numerous experiments have been. 
made; but as asmall number, conducted on proper scientific prin- 
ciples—more particularly if they agree with each other, and have 
been, as in these instances, checked by the observations of two 
persons—are as aalecwore as any greater number, I have de- 
taited, and given the calculations in full of only a few. These — 
will serve as examples for any one who is disposed to com- 
pute from the observed data, the resulting quantity of heat ob- 
tained, and thus prove the truth, or error of the reported result. - 
Finding that observations were made with more accuracy in pro- 
portion as experience familiarized me with them, I considered 
that the later ones were more to be trusted than the earlier, and 
fa) 
= 
te 
3 
oO 
< 
o 
ig 
oe 
= 
oD 
5 
o 
oO 
i 
“5 
fee) 
i 
= 
9 
= 
@ 
Set 
fo 
D 
a 
Qu 
fas) 
72) 
iz) 
jen 
ca 
f. 
i 
7 
and that the temperature of the atmosphere at the time, was 
very different and much lower than at any of the later ex- 
periments. 
It is proper to remark that in conducting these delicate experi- 
ments many causes of error have been either unavoidable, uncor- 
rected, or overlooked. Owing to defects of mechanical contriv- 
ance and unskillful workmanship, incidental perhaps to every new 4 
device, and a noviciate intercourse with practical mechanies, th 
machine was not capable of performing all its duties with the ac- 
curacy the natural laws involved called for. And notwithstand- — 
ing a great desire to avoid and correct errors from the leakage or 
irregular working of the machine, by compensating for them, I 
cannot in all cases rely upon the correctness of the allowances 
made therefor. 
e thermometers used in these experiments were manufac- 
tured in New Orleans, and their indications differed so much 
from each other that they could not be fully relied upon. The 
nature of the investigation required instruments that should 
mark, if possible, the tenth part of a degree of Fahrenheit; but 
none that I could obtain was of sufficient sensibility to be "read 
accurately within half a degree; and, therefore, all indications 
set down as more minute, must be regarded as conjectures, 
or interpolated calculations. As I found that no two ther- 
. mometers agreed with each other, I was aware that it was desir- 
able one or more should be verified by a comparison with the in- 
* See continuation, in next number of this Journal. 
sie 
