HOW TO STUDY METEOEOLOGY. 
147 
consider simple barometrical tubes, which are merely graduated 
to inches and decimals, to be barometers at all , Admiral Fitzroy’s 
notation should be adopted, if any. What this is, is pretty well 
known. 
The more expensive barometers read to thousandths of an inch, 
but it is submitted that a reading to hundredths is quite suffi- 
cient ; and that a good working barometer, answering all the 
reasonable requirements of the amateur, need not cost more 
than 61. 
The due application of the necessary corrections * must not 
be neglected. These are, for temperature, sea level, and capil- 
larity. 
Mercury, like most other substances in nature, expands under 
the influence of heat and contracts under that of cold. A read- 
ing taken when the thermometer stands at 35° may therefore 
seem to differ from one taken when the thermometer is at 70°, 
yet the pressure of the atmosphere may be really identical in 
the two cases. To make the different readings of a series legiti- 
mately comparable, they must be brought to some common 
standard of temperature, and meteorologists are agreed that that 
shall be 62° of Fahrenheit’s thermometer. 
It is well known that the barometer is employed for the 
measurement of heights. What does this imply ? Clearly that 
the elevation of the mercurial column varies as the elevation 
of the place of observation varies ; and such is really the case. 
An instrument placed 700 feet above the sea may read the same 
as one placed at the sea level, and yet the real pressure of the 
atmosphere at the two stations will differ considerably. The 
sea level is the standard of altitude ( more Hibernice) to which 
it is agreed that all barometrical observations shall be referred. 
Speaking roughly, every reading has to be augmented by about 
-j^th of an inch for every 100 feet that the place of observation 
is elevated above the level of the sea. 
The correction for capillarity depends on the well-known 
property of all liquids to adhere to solid surfaces with which 
they come in contact. A fictitious level is thus given rise to, or, 
more accurately, the surface is not level at all, but curvilinear. 
When mercury is enclosed in a cylindrical tube of small dia- 
meter, such as a barometer tube, the surface is convex, and a 
subtractive correction must be applied to every reading. The 
* For tables of these, of course in the present article I have no space. 
They will be found in various books. 
Rise for N. Ely. 
N.w.; X.E. 
Dry, or less wind. 
Except wet from. N. Ely. 
S.E.j S.W. 
Wet, or more wind. 
Except wet from N. Ely. 
Fall for S. Wly. 
