SOME REMARKS ON 
209 
the vertical spindle. The way this is done is shewn on a larger 
scale in Fig. 4, and a section is given in Fig. 5 of the way in 
which the steel spheres (nine in number) roll, having contact at 
four points. The straight lines passing respectively through the 
two points of contact above and the two below meet in the axis 
of rotation of the spindle — as suggested by Professor Stokes^ — 
in order to ensure rolling contact. 
None of these instruments shew the direction and velocity, or 
direction and quantity, of wind by one curve, and it occurred 
to the Rev. J. M. Wilson, M.A., Head Master of Clifton 
College, to obtain these latter quantities by one continuous 
curve, in which the length drawn in a given time should 
be proportional to the quantity of wind passing in that time, 
and its direction should always coincide with that of the 
wind. How this idea has been carried into practice was 
described by Mr. Wilson and the author at the recent 
meeting of the British Association at York, and Fig. 6 shews 
the experimental ‘ form of instrument which acts thus : — the 
cups give motion to a train of wheels at the lower end of a 
central spindle and turn a roller which presses on the recording 
surface and moves it with a velocity approximately proportional 
to that of the wind ; at the same time the windmill head is so 
arranged as to keep the plane of rotation of the roller always 
coincident with the direction of the wind. Fig. 7 is a curve 
formed by combining several days’ observations, and agrees very 
well with the indications of other anemometers. A careful 
examination of the weather chart of The Times shews that the 
wind cannot be said always to move in parallel planes, at any 
rate over an area such as the United Kingdom. For instance, 
while in J une its motion was fairly uniform, in April of this year 
it was on thirteen days (not consecutive) out of the thirty 
I Proceedings of the Royal Society, No. 213, 1881. 
