124 Dulong and Petit on the Measure of Temperatures, &c. [Fex. 
In this respect our apparatus is greatly superior to those 
employed to determine the dilatation of solids. In them the 
smallest derangement of the fixed point during the long duration 
of the experiment, does not merely affect the total length of the 
tale ; the dilatation itself is augmented or diminished, which 
occasions the most serious errors. We see, on the contrary, 
that when, m our experiments, the heights, / and A’, are affected 
by the cause of which we spoke, the difference h — h’ which 
measures the dilatation is not so. For it is absurd to suppose 
that the instrument becomes deranged during the very short 
interval which elapses between the successive observation of the 
hot and cold column. 
We have collected in the following table the mean results of 
a great number of observations made in the way just described. 
The first column contains the temperatures such as they are 
deduced from the dilatation of air; the second contains the mean 
absolute dilatations of mercury between freezing water and each of 
the temperatures indicated in the first column; the third column - 
exhibits the temperatures which we should obtain, on the suppo- 
sition that the dilatation of mercury is uniform, or, in other words, 
those which should be indicated by a thermometer formed of 
that fluid inclosed in a vessel, whose expansion followed the 
same law as its own. 
TABLE Il. 
Temperatures indicated by 
the dilatations ef mercury 
supposed uniform, 
euhsistice vais. dblovoerth oncuresihictventeaee 
VQ hae stiryssinlaen Ioausgrelils skrauoidnaenin cc Galle 
BO act «heat» iecehteeed akeningheh- rete SER 
SSE in: ytd guts bare epmrbaachaves ductile Nb IAM 
(To be continued.) 
Temperatures deduced A ; 
a E Mean absolute dilatations 
from the dilatation of mercury.* 
of air. 
* Each of the results contained in this column is the mean of a great number of 
measures, which it would have been too tedious to have given in detail; we shall 
satisfy ourselves with giving the extreme values for each of the three temperatures, 
Maximum. Minimum, * 
100° eee et eres e ccc reece sore Creer eset torereee 5 v3 z 
ZOD) oc cisloviciecis w'v'e|ie'e apie saty oda vwscee seccevoes S49T 
1 — 
