certain new forms of Thermo-Barometers. 3 
is settled, and when the mercury in the barometer indicates mean 
pressure, or nearly mean pressure, that is, about (p), or 29°5 
inches. The liquid must be made to stand about the middle of 
the stem at mean temperature (¢), or 62°. The flask is then 
placed in a water-bath, which is brought to different degrees of 
temperature (¢,), say to 42°, 62°, and 82°, and marks are made 
on the paper mn coincident with the level of the liquid in the 
stem at these different temperatures; the spaces between these 
marks are respectively divided into twenty equal parts, thereby 
forming intervals of 1 degree. These divisions may be extended, 
if desirable, as the intervals are very nearly uniform (see for- 
mula 6). 
The lower part of the barometrical scale DC must be gra- 
duated at a time when the weather is settled, and when the mer- 
cury in the barometer stands at a high column (p,), the tempe- 
rature of the air being (¢) at, or not far from, the mean tempe- 
rature. Having placed a mark (p) about the middle of the 
scale D C, bring this mark coincident with the temperature ¢ on 
the scale mn, and make a mark on D C coincident with the level 
of the liquid in the tube; then the space, —g,, between these 
two marks will indicate a change of atmospheric pressure mea- 
sured by the column of mercury equal to p—p,. Substitute the 
values —q, and p—pz,, thus obtained, in equation (4), and deter- 
mine the value of the constant ee the values of s and h having 
been previously found by experiment and observation. Substi- 
tute the value of tes thus determined, in equation (5), and caleu- 
late the values of g, for p.=30, 30-5, and 31; these values of 
go being marked off from the middle point p on the scale DC, 
will give the points corresponding to these pressures; these 
spaces may then be subdivided into equal parts so as to read off 
tenths and hundredths. In precisely the same manner the 
upper part of the scale must be graduated, when the mercury in 
the barometer stands at a low column. 
The observations of atmospheric pressure are made as fol- 
lows :—The temperature of a delicate thermometer being first 
noted, the middle point p of the barometrical scale DC is 
moved until it coincides with this temperature as marked on the 
thermometrical scale mn; then the mark on the scale D C coin- 
cident with the level of the liquid in the tube, will give the atmo- 
spheric pressure as measured by a column of mercury. 
Constructed in this manner, the errors of the indications can- 
not exceed 03 of an inch of a column of mercury. Thus, by any 
person possessing ordinary skill in manipulation, an instrument 
may be made for a few shillings, which will be quite as accurate 
B2 
