480 
DR. T. R, ROBINSON ON THE REDUCTION OE ANEMOGRAMS. 
caused by the solar action in the vicinity of the place of observation and varying with 
the sun’s declination. Supposing qy to be that part of it due to the altitude, its mean 
9 
annual value would be V=2-721, about 06 of V' (page 412), and its D = 207° 8 ,- 7. 
Other periodical causes, such as the length of air traversed by the sun’s rays at different 
altitudes, the difference of the earth’s daily and nightly radiations, and the amount of 
watery vapour in the air, might be similarly taken into account. 
I have already stated that I thought it useless to deal with the observations of single 
days ; I, however, tried two experiments in this direction, which may be of some 
interest, though the first of them was unsuccessful. 
1. In many instances, even when the wind is moderate, there are variations in its 
direction which suggest the notion that they are due to aerial whirlpools on so small a 
scale that they are not likely to reach any other meteorological station. 
I thought it might be possible to determine the constants of such a motion in the 
following way. The curve described on such a supposition by the thread of wind which 
passes the anemometer at a given station is that which would be traced by a pencil 
fixed there on a plane revolving with an angular hourly velocity u round a centre which is 
carried in a line inclined at the angle a to the axis of x with the hourly velocity V, | and n 
being the coordinates of that centre at the origin of the time, and a the angular motion 
there. It is obvious that we have 
dx=dt\V cosa(l dy=dt { V sina(l 
Then at successive hours equating ^ to tangD, and dx 2 -\-dy\ to s, I would be able 
to get values of the unknown quantities. But against this is my ignorance of the rela- 
tion between cd and this distance from the centre of the circle, which is not given in any 
book to which I can refer. Newton, in the vortex which he considers, gives it inversely 
as the distance. It is probably nearer the inverse square. Either of these suppositions 
would make direct integration impossible, so I gave up the project. 
2. The other was an attempt to determine from these observations the existence of 
an atmospheric tidal current. As in the case of the ocean, so in the atmosphere, the 
air must be heaped up in the meridian passing through the moon, or a little to the east 
of it ; and this elevation must be accompanied by a horizontal current. 
Laplace (Mec. Cel. ii.) has shown that the maximum air-tidal current is 0 , 07532 metre 
in a centesimal second *, which in English measure and time is 0T95 mile in an hour. 
He, however, gives no indication of the phase of this maximum, or in what stratum of 
the atmosphere it occurs. At the earth’s surface, owing to friction and other causes, it 
must be considerably less than the above value, and the analogy of sea-tides is too slight 
to give much assistance in the research. It may, however, authorize us to assume that 
on opposite sides of the lunar meridian the directions of this current will be opposite. 
* It is to be regretted that in this noble work Laplace used the centesimal division of the quadrant, and the 
decimal and centesimal divisions of the day. Whatever be the fate of the metric system, it is very unlikely that 
either of the others will be generally adopted. 
