ON MAGNETICAL AND METEOROLOGICAL OBSERVATIONS. 61h 
changes. The redaction also leaves nothing to be desired in this respect, for 
it is perfectly lucid, and has the great advantage of the tension of the vapour 
being already computed. 
But detailed observation-journals offer the further advantage of enabling 
us to trace the changes, depending chiefly on variations in the direction of 
the wind, which are mixed up with the periodical changes. If, for example, on 
every day when the direction of the wind at noon was north-west I deduct the 
preceding or 10* reading of the barometer from the noon-observation, or the 
noon-observation from that at 2 hours, I shall infer directly whether in North 
America, as in Europe, the barometer rises with the north-west wind, 2. e. 
whether north-west is the passage from a warmer lighter wind to a heavier 
colder one. Considering the large arcs through which the wind-vane moves, 
it is desirable to have observations near together, and the rare occurrence of 
several directions is an additional reason why we should have as many records 
of the direction of the wind as of the readings of the other meteorological 
instruments. With the directions the pressures also may be given, and to 
save space they might perhaps be thus recorded: SW, SW,, where the num- 
ber expresses the pressure. 
That which appears to me most desirable in the great English and Russian 
undertaking, to which Brussels, 'Munich and Prague have so meritoriously 
joined themselves, is, that the materials gathered should not only be published 
as was done with the Mannheim Ephemerides, which remained unemployed 
for more than half a century afterwards, but that results should be deduced 
from them as soon as possible. The extensive calculations connected here- 
with will require a division of labour: I would propose that the British As- 
sociation should undertake the distribution of the parts, as the Berlin Aca- 
demy did in regard to the maps of the stars. For example; in order to learn 
the variations in the pressures of the air and vapour, and in the temperature, 
dependent on the direction of the wind, we must not combine in a common 
mean the values at the several observation-hours when any particular wind 
has blown, but we must first eliminate from these values the periodical varia- 
tions by which they have been affected. I would willingly offer to undertake 
this calculation for each year for one station. If at the end of five years the 
data from the stations of Greenwich, Newfoundland, Toronto, Van Diemen 
Island, Petersburg, Nertchinsk, Pekin, Brussels, Munich and Prague were 
combined, we should obtain from them a satisfactory reply in respect to the 
modifying influence of situation, whether on the east or on the west side of the 
sea,—whether in the interior of a continent, or on the coast,—whether in the 
northern or in the southern hemisphere. 
As within the tropics the lower current of air flowing towards the equator is 
compensated by an opposite current above, so we may assume that in the tem- 
perate zone the equipoise is maintained by currents on the same level flow- 
ing in opposite directions, and thus that the air, which in the course of the 
year passes over certain stations on a given parallel towards the pole, returns 
towards the equator, passing over other stations on the same parallel. We 
should expect, that if we find (taking the intensity into account) a prevailing 
direction of the wind at some stations, we should find an opposite direction at 
_ other stations. But the air which passes over the parallel coming from the 
_ equator brings with it a higher temperature, which it gradually parts with as 
it flows over the surface of the earth, and which it cannot therefore bring 
back with it when it passes the same parallel on its return towards the equator. 
Now colder air occupies less space than warmer air, and therefore the current 
of air flowing from the pole to the equator is narrower than when it flows 
from the equator to the pole. If the beds in which these opposite currents 
