[ 226 ] 
defined, juft touched a fouth and north line, which 
I had marked upon the ground ; at that inftant I began 
my reckoning, and followed the fhadow with my 
eye in its progrefs, counting feconds all the while by 
the clock, until I had reckoned up 1 5 feconds ; then 
I obferved exactly where the forefaid edge of the 
fhadow was. This experiment I repeated ten times 
in half an hour* and feldom found the difference of a 
fecond, in the time which, different clouds took to 
move over the fame fpace. On the 5th of May cur- 
rent, I repeated the trial four different times, the fun 
being alfo near the meridian, the wind in the weft, 
with light clouds floating in a clear fky as formerly ; 
and found that the fhadows of different clouds took 
fome of them 44, and others 4 5 feconds, to pafs over 
the fame fpace which they had moved over in 15 
feconds, in the former trials. 
Feet 
This fpace meafures exaftly 1384=: fpace palTed over in 1 5 feconds, 
which multiplied by 4 
gives 5536=; fpace paffed over in one minute, 
which multiplied by 60 
gives 332,160=: fpace paffed over in one hour. 
Which fpace is =62.9 Englifh miles per hour, the 
velocity of the wind in March 1763. 
One third of this (or 21 miles nearly) (hews the 
velocity of the wind on May the 6th, when it blew a 
frefh gale. 
This day, May 12, there was a fmall wefterly breeze, 
the velocity of which I meafured upon the fame line, 
the fun being 10 minutes paft the meridian, and found 
that the fhadow took 95 feconds to pafs over the 
above 
