456 H. A. Hazen—Reduction of Air-pressure to Sea-level. 
January and the highest in July. If. the temperature term 
only were at fault, we should expect to find a regularity in the 
variation of the difference of elevation as computed from the 
mean of each of the months, also the same variation at all places ; 
that is, if we should find the highest result in July and the 
lowest in January in any one place, any theory sihieh could be 
applied to the variation of the temperature at different alti- 
tudes must necessarily give in general the same results at all 
isolated peaks throughout the world; but neither of these 
suppositions is according to fact, as has been shown in the 
puted from Laplace, was in March, and the lowest in August, 
whereas in the elevation of Pike’s Peak above Dodge City, the 
highest value was in July and the Jowest in December: 
We may then consider that the temperature term will remain 
nearly constant so far as it is affected by the heat of the sun ; 
this hypothesis will be strongly dane by a_ practical 
demonstration in the course of this discussio 
On the other hand, the pressure term is affected by a great 
variety of causes. In summer there is an upward motion 
of the atmosphere, the ocean relatively to the land is cool and 
the air moves toward the that is coolest, while in winter 
exactly the reverse takes place. This may explain the facts 
just mentioned. Mt. ase near the coast, is very little 
affected by this motion, and the results differ ‘but little in 
January and July. 
There seems to be, also, a constant motion of the atmosphere 
toward the east, slower in summer than in winter. It has 
April obaderikioiy’ | is Seauently greater than that from ae 
months. This may be due to a general diminution of t 
pressure at the lower station, which does not seem to sien ; 
an elevation of 6,000’; such a diminution has been found at 
over 500 stations in the middle latitudes, but only at the lower 
stations. It would certainly seem, then, that the pressure term 
is more liable to change than the temperature term 
In order to separate the various elements which enter into 
the formula of Laplace, the following method is proposed : 
f tables be constructed from Laplace according to the plan 
of che table on page 370, this Journal, last number, for each 
thousand feet of elevation, it will be possible to obtain the laws 
of variation of the reductions to sea-level and represent them 
by a formula which nee dispense with logarithms and sepa- 
rate the various term 
The following ferdambe are founded on such a plan: 
