SLIDING-SCALE FOR CORRECTING BAROMETER READINGS. 49 
into prominence one of the little troubles of a sa ict com- 
puter. Itis this. Suppose this morning Mount Victoria sends 
in a reading of his barometer 26°742, and air temperature 50°, 
— at Sydney the barometer read 30 ‘021, and air temperature 
57°; now the question arises for which of these temperatures, 
or fage what other, shall I compute the altitude correction. 
retically, in computing altitude corrections it is assumed that the 
air is cooler in proportion to the elevation, and therefore the 
mean of the two should be taken ; but in our practice this is foun 
to be incorrect, and all the barometers will agree better if they are 
corrected for altitude at the temperature of Sydney, and the 
corrections are always therefore computed at the Sydney tem- 
perature. You will naturally ask why is theory wrong in this 
instance ; probably the answer would be found in the — con- 
dition of our atmosphere, which is that of having a warm win 
QR 
using the sliding-scale, we follow the old rule of using the Sydney 
temperature, but when the readings are taken on this line, H eek & 
they are affected by the right altitude correction, they are 
by the wrong correction for temperature, for the line is plotted as 
stated before, so that the temperature and altitude correction for 
the thermometer a the station in question. 
It is found impossible to pore for this difficulty in the sliding- 
scale ; but it is met by adding to the reading found, 2} times the 
difference between the thermometers when the u upper thermo- 
meter is lower, and subtracting it when it is higher. It is evident 
that such a sliding-scale will not give the correct to 
0-001 in, but it vill to 0-01, which is more than sufficiently 
accurate for the purpose of daily weather maps; and if the third 
place of decimals is taken by estimation, the readings are found 
sie eat to coincide with the computed readings. 
I should mention here that as Kiandra, one of the meteoro- 
—_ ae is at an altitude of 4,640 feet, it — be necessary 
he scale at 25,000 inches, so ae sliding-scale 
ontesiéioen of 1, 2, or 3 inches were made to the seer of the 
barometers, and the scale plotted accordingly. 
