78 MR. T. HOPKINS ON IMPROVEMENT 
have expanded, and consequently pressing there with less 
weight on the earth and on any barometer near it. 
Change in the temperature of the gases is therefore 
recognised as a cause of alteration of atmospheric pres- 
sure, as it is measured by the barometer, and, taken in 
conjunction with changes of vapour tension, is presumed 
by meteorologists in general, to cause all the fluctuations 
of the barometer, no other cause being acknowledged to 
influence atmospheric pressure. 
Yet the barometer occasionally sinks without either 
temperature or vapour tension showing alterations to 
account for it. Indeed sometimes, tension of vapour 
varies but little, and temperature declines; but the baro- 
meter, so far from rising, as might be expected from the 
recognised influences of these two forces, actually falls, 
often to a considerable extent. From these circumstances 
we may infer that neither reduction of vapour tension, as 
indicated by the dew-point, nor fall of the temperature of the 
gases, as shown by the thermometer near the surface, nor 
both together, are the real causes of the sinkings of the 
barometer that sometimes take place. 
The sun is undoubtedly the great scurce of the heat 
that disturbs the atmosphere, but its heating effect in 
different latitudes is unequal, and its influence on dif- 
ferent parts varies also with its passage across the tropics ; 
yet any partial effect thus produced on the weight 
of the atmosphere is, in no long time, counteracted 
and corrected by the operation of gravity, which imme- 
diately begins to act on the aerial ocean to restore the 
equilibrium of atmospheric pressure. Therefore no change 
of temperature as shown by the thermometer has furnished 
reasonable ground for believing that the great falls of the 
barometer that often occur can have been caused by such 
changes as are indicated by the thermometer. 
The sun has a daily, as well as an annual influence, but 
