BAROMETER. 
pulley, is stretched by a weight that is nearly a 
counterpoise. By this means the motion of the 
mercury is communicated to an index, which 
turns round a graduated circle, and thus the ver- 
tical range is enlarged at pleasure. 
It is extremely desirable, for meteorological 
purposes, to have a regular and successive series 
of the changes which take place in the pressure 
of the atmosphere during any given period; but 
as this would require constant attendance on the 
part of the observer, mechanical contrivances 
have been adopted for registering the indications 
of the barometer, and retaining them in a con- 
nected form. When the instrument is fitted up 
| in this manner, it is called a self-registering baro- 
meter. The most simple kinds of self-registering 
barometers are such as indicate the greatest rise 
and fall of the mercury, or its extreme range, 
during any stated period; and when this only is 
required, the object is easily accomplished. Of 
this description is the self-registering barometer, 
invented by Alexander Keith, Esq., F.R.S., Hdin- 
burgh. It consists of a bent tube, such as ABD, 
fig. 7, hermetically sealed at A. The mercury in 
the shorter leg supports. float, to which is affixed 
| aslender wire terminating in a bend or knee. This 
knee embraces a very small wire stretched along 
| the scale, and pushes upwards or downwards 
two bits of glazed silk which slide along the wire 
very easily, yet so as to retain the position to 
| which they are moved by the ascent and descent 
of the mercury. ‘The instrument is prepared for 
experiment by bringing the two bits of silk in 
contact with the bent knee of the float wire ; the 
| points to which they may afterwards be removed, 
indicate the extreme range of the mercury during 
the interval of any two observations. When not 
only the greatest and least altitude of the mer- 
cury is sought for any given time, but also its 
precise height at every intermediate moment, 
more complicated contrivances must be em- 
ployed; the instrument must then consist of a 
barometer connected with a time-piece, and a 
crayon or pencil affixed to a float obeying the 
motions of the mercury. The greater number of 
self-registering barometers of this nature are so 
constructed, that the crayon is made to describe 
a continuous line on a vertical cylinder, turning 
on its axis by means of clock-work, and making 
‘a certain number of revolutions in some stated 
time. The cylinders are divided longitudinally 
by parallel lines into equal spaces, corresponding 
to some particular portion of time; and thus the 
line described by the crayon in that time, indi- 
cates the successive heights of the mercury dur- 
ing its continuance. 
Oscillations of the Barometer.—Professor Loomis, 
in a paper on the storm of December 20th, 1836, 
which was read at a Philosophical society in New 
York, March 20th, 1840, observes that “ the phe- 
nomenon, probably the most difficult of all to be 
explained, is the oscillation of the barometer.” 
And he enumerates nine causes, which have been 
361 
given, all of which he rejects, and offers a tenth, 
which is perhaps equally insufficient. 
“1. The oscillations of the barometer,” he says, 
“ have been ascribed to the destruction of large 
masses of air in the higher regions by electricity. 
The supposition is too gratuitous to deserve seri- 
ous consideration. 
“9. They have been ascribed to the diminished 
pressure resulting from the loss of rain. But the 
amount of rain which fell in the case under con- 
sideration would be balanced by a column of 
mercury about one fifteenth of an inch in 
height. 
“3. Heat, by expanding a column of air, causes 
it to ascend to a greater height, and thus changes 
its centrifugal force, arising from the earth’s ro- 
tation. This cause is too insignificant to produce 
the effect in question. 
“4, They have been ascribed to the attrac- 
tions of the sun and moon. Laplace estimates 
the greatest oscillation of the barometer due to 
this cause to be at the equator, 0°025 inch. 
“5, Leslie ascribes them to the centrifugal 
force arising from violent winds. But, in the 
case of a hurricane, this would not produce an 
oscillation of the barometer amounting to the 
thousandth part of an inch. 
“6, The opposition of winds. 
duce a small movement of the barometer. 
“7. Thé barometer has frequently been ob- 
served to fall under the influence of a-whirl- 
wind. But in the present case there was no 
whirlwind. 
“ 8. These oscillations have been ascribed to 
sudden changes in temperature, and in the 
amount of aqueous vapour. An elevation of 
temperature of the entire atmosphere could not 
directly affect its pressure, for in proportion as | 
its density is diminished, its height will be in- 
creased. But if, by any means, a portion of hot 
air can be made to displace an equal bulk of 
cold air, the weight of the column must be 
diminished. ' It is obvious that this cannot be a 
state of permanent equilibrium ; yet it is worthy 
of inquiry whether it may not temporarily exist 
under the influence of winds. On the 20th of 
December, 1836, the air over nearly the whole 
of the United States became unusually heated, 
and its specific gravity was, of course, diminished. 
If, then, the height of the atmosphere remain 
invariable, a diminution of pressure ought to be | 
But, although a fall of the | 
the consequence. 
barometer is usually accompanied by an eleva- 
tion of temperature, the reverse is sometimes the 
ease. Thus the fall of the barometer in Europe, 
which, at most places, amounted to more than 
an inch, was accompanied by a steady fall of the 
thermometer. The barometer in this case fell in 
spite of the increased specific gravity. of the air. 
We may naturally presume, then, that a change 
in the specific gravity of the air produces only a 
secondary effect on the oscillations of the baro- 
meter. 
This might pro- | 
(Cae eee ene 
1p, 
