[ 22 5 ] 
the frequent interruptions it meets with from hills, 
trees, and houfes ; and even in open plains, the furface 
of the earth, though much fmoother than it commonly 
is, mud refled:, and interrupt fucli a fluid as the air, 
and occafion great irregularity in the velocity of its 
current : this is the reafon, when a feather is let fly 
with the wind, why it feldom, if ever, defcribes a 
ftrait line, but moves fometimes in a kind of fpiral, 
now high, and then low, fometimes to the right, and 
then again to the left ; and why two feathers let fly 
at once, feldom, if ever, keep together, or defcribe 
fimilar lines. 
But, at fome confiderable diftance from the earth, 
the velocity of the wind feems to be regular and 
fteady : nothing can be more uniform, than the velocity 
of a cloud in the fky appears to be, even in the 
greateft ftorm : it is like afhip carried away infenfibly 
by a fmooth and gentle current, palling over equal 
fpaces in equal times. This fuggefted the thought, 
that the motion of a cloud, or its fhadow, over the 
furface of the earth, would be a much more proper 
meafure of the velocity of the wind. 
In the end of March 1763, I had as favourable an 
opportunity of putting this method into pradice, as I 
could have wifhed for ; the ftorm was exceeding high, 
and moved with vaft velocity 5 the fun was bright, the 
fky clear, except where it was fpotted with light 
floating clouds i 1 took my ftation in the north 
window of my dining room, near the clock, from 
which I had a free profped of the fields ; the fun 
was in the meridian, the wind due weft interfeding 
his rays at right angles ; I waited until the fore-part of 
the fhadow of a cloud, that was diftind, and well 
Vol. LVI. G g defined, 
