494 JOURNAL OF THE PLYMOUTH INSTITUTION. 
depends on the quantity of air above the instrument as well as on 
the density of the atmosphere. It is clearly seen that the situation 
of the barometer must influence the height of the mercury, 
because if situate on the summit of a high mountain, it is manifest 
there cannot be as much of the atmosphere above it as if it were 
placed in a deep valley at its foot. It is calculated that at a height 
of three miles; 7.e. just at the summit of Mount Blanc, one-half 
the whole atmosphere lies below it ; and at six miles three-quarters 
lies below; whilst at eight miles all animal life, as we know it, 
must become extinct. Hence for comparison between different 
observers it is needful to reduce all readings from places differently 
situated to one standard height, and the standard selected is the 
mean sea-level. 
Again, to any one who has thought on the subject, it is manifest 
the air must be much denser; 7.e. heavier bulk for bulk, when it 
is cold than when it is warm; and hence comes the necessity for 
a second standard ; viz., one for temperature, and the one selected 
is the freezing-point of fresh water. 
Thus, after correcting the readings for capillarity and index 
error, other corrections must be made for the height of the 
barometer, and for the temperature of the atmosphere. In my 
case I found the height of my instrument above our standard line 
by first obtaining from the office of the Royal Engineers the height 
of the nearest bench mark made during the Trigonometrical Survey, 
and levelling from thence to the position my instruments are 
placed in. I thus found the cistern of my barometer to be 69 
feet above mean sea-level. In making reductions for altitude and 
temperature, the rules given in form 14, issued from the Meteoro- 
logical Office, are near enough for all practical purposes, They are 
as follows : 
(1) Each ten feet of vertical height causes one-hundredth of an 
inch diminution ; z.e. for each foot above mean sea-level add one- 
thousandth of an inch to the barometrical reading. 
(2) Each 10° Fahrenheit above 32° causes very nearly three- 
hundredths of an inch increase, or for each degree Fahrenheit 
above 32° subtract three-thousandths of an inch from the pressure 
shown by the barometer. 
The atmospheric pressure in Plymouth has fluctuated between 
30°822, the maximum on 15th December, 1865, and 28°316, the 
minimum in the morning of the 24th January, 1872, a difference 
