Stoney.— (9« the Reduction of Daily Weather Reports. 255 
the intermediate seasons, and in other directions at various seasons, 
owing to the proximity of mountains, height above the sea, and other 
local peculiarities. And in the fourth place they depend on the passing 
phase of weather. 
A similar complication of causes affects the records of the barometer, 
but there is this material difference between the cases of the ther- 
mometer and of the barometer : — The passing phase of weather is the 
cause which most affects the barometer, and the other operative causes* 
produce effects which are relatively small ; whereas in the case of the 
thermometer, the passing phase of weather produces an effect which is 
so disguised as to be almost hidden under the much larger effects of 
the other three causes enumerated above. 
My proposition is, then, to remove by one simple reduction the 
portion of the result due to all the other causes, and to tabulate and 
exhibit in the chart that residual portion alone which is intimately 
dependent on the character of the passing phase of weather. This 
could be effected after first forming the annual curve of temperature 
for each of the stations in telegraphic communication with the Meteoro- 
logical Office. The materials for doing this, I believe, exist at the 
Meteorological Office, from the accumulation of daily records of past 
years. This curve for each station will tell what the average tempera- 
ture at that station is on any specified day of the year, and the 
difference between this and the observed temperature sent to the 
Meteorological Office by telegraph will tell how much warmer or 
colder than the average the day is at that station. It is these excesses 
and defects of temperature which I propose should be represented on 
the map. 
The curves to which the proposed records would give rise would 
doubtless be found to have as intimate a connexion with the direction 
of the wind as the curves of equal pressure have. During the recent 
extensive but feeble cyclone, the centre of which passed over the British 
Isles on the 4th of this month (May 4, 1872), the curves of equal 
pressure have, as usually happens in cyclones, been nearly coincident 
with the direction of the wind. But the directions of the isothermal 
lines have had no obvious relationship with the cyclone. And in fact 
the crude information given in the words ''somewhat warmer," 
''somewhat colder," and so on, which are also entered on the maps, 
throws more light upon the phenomena of the cyclone than the iso- 
thermal lines do. "Whereas if the simple reduction I have pointed out 
were applied, and the differences of the temperature from its mean 
amounts on the successive days tabulated and mapped, we should 
certainly become possessed of a fresh body of usefnl information. The 
lines of the new chart would probably be found arranged somewhat 
in radial lines. 
* Except height above the sea,, from the eflFect of which barometric records aro 
always freed. 
