DK. T. R. ROBINSON ON THE REDUCTION OF ANEMO GRAMS. 
411 
The first thing which strikes one in this Table is the irregularity of the wind. It 
varies in each octant ; in each octant it varies with the month, and in each octant and 
month it varies with the year. As to the first of these variations, both the velocity of 
the wind and the number of hours during which it blows are, in general, a maximum in 
the first octant (S. to S.W.) ; they decrease from this to a minimum at octants N. to 
N.E., and increase to octant 1. The products of the velocity and time at the maximum 
and minimum are as 6:1. The predominance of south-westerly winds is what might 
be expected from the combination of an equatorial current with the earth’s rotation ; 
but it is not obvious why it is not absolute. Probably much of the change of 
direction arises from circumstances local to the place of observation. Por instance, 
the direction of the west coast of Ireland, which runs nearly N. and S., may occasionally 
turn the S.W. currents northward ; and the mountainous ground of Antrim may divert 
it here towards the east. It must also be remembered that our anemographs give only 
measure of the wind at the earth’s surface, where it is at once retarded and thrown into 
gigantic eddies and vortices by the effects of friction. 
The experience of aeronauts shows that at a few thousand feet elevation the velocity 
is often far greater than it is below, and that the direction is much more uniform. But 
I do not see how this error is to be remedied. The summit of a mountain is not exempt 
from it ; and though a small and lofty island, like St. Kilda, far from any extensive land, 
would be better, yet even here the friction of the sea’s surface will destroy velocity. It 
is possible that an anemograph at the top of a tall and slender “ stack ” would give a 
much larger velocity than one at its base ; the record could be easily effected below 
by telegraphy. We must remember that a current of air comports itself like one of 
water, and shall be assisted in comprehending the nature of a gale by watching the 
irregular movements of a river in flood. There must also be eddies in a vertical plane. 
On the action of these see a valuable paper by Prof. Hennessey in Phil. Trans. 1860. 
An anemograph for vertical currents might be made by a set of windmill-vanes placed 
horizontal. 
Secondly, in each octant the amount of wind varies with the month. It is a maximum 
in January; decreases from this to July, the ratio being 2^ : 1. From this it increases 
to the end of the year. There is an exception to this in March, where the daily amount 
is greater than in February in the ratio of 1T3 : 1. This might seem to countenance 
the vulgar notion of stormy weather prevailing near the equinoxes ; but there is no such 
excess in September above October ; and in March, though the yearly maxima are higher 
than in February, yet the number of hours when the velocity exceeds 25 miles is consider- 
ably less. This monthly change is an obvious consequence of the change of the sun’s decli- 
nation, for the zone where the easterly winds of low latitudes confine with the westerly 
ones of more northern regions must shift with that to which the sun is vertical. 
For the third of these irregularities, that which prevails from year to year, there can, 
in the present state of our knowledge, be no certain cause assigned. It will be seen 
that in the same octants the variation is very different in each month, and that the 
3 i 2 
