Lovering on Magnetic Observations at Cambridge. 155 
and these periods alternate with the times of maxima and minima 
of gaseous pressure, so that the maxima of the former nearly cor- 
respond to the minima of the latter. The diurnal march of the 
other three forces from maximum to minimum, and from mini- 
mum back again to maximum, are continuous, like that of the 
temperature, while the blending of the gaseous and vapor press- 
ures in the mercurial column produce the apparent double maxima 
and minima of the barometer. This might be expected from the 
common dependence of the changes in the force of the wind, 
the amount of evaporation, and the oscillations of the atmosphere 
upon the antecedent changes of temperature. The Greenwich 
observations, as published in the volume for 1841, by Mr. Airy, 
confirm the deduction drawn from the Toronto observations. 
We thank Mr. Sabine for having so distinctly brought out. this 
relation, and we allude to it here as showing the necessity here- 
after of accurate and full hygrometric observations, from which we 
may deduce the amount of vapor pressure which must be applied 
as a new correction to the indications of the barometer. 
In deriving correct and valuable mean results from the daily 
observations, there are two difficulties. 1. A single day of 
extraordinary disturbance may be sufficient to vitiate the true 
character of a whole month. I have endeavoured to diminish this 
source of error by discarding from the means those observations 
which were largely and notoriously warped by such abnormal 
influences. 2. It is important that each of the means should be 
derived from the same number of daily observations, made on the 
same days; otherwise it may happen, that a day, whose declina- 
nation curve, for example, is regularly formed, but whose average 
value indicates a great easterly or westerly tendency, will be 
allowed to impart its own peculiar character to those hours when 
