[ exeg J) 
4. Lastiy, from September 27th, 1790, take May 26th, 1786, 
there remain four years and four months; and from 27 de- 
grees 15 minutes take 26 degrees 21 minutes, the remainder, 
divided as before, quotes 12 minutes 27 feconds. ‘The mean 
deduced from thefe and feveral others (not here inferted for 
brevity’s fake) is about 12 minutes 20 feconds. Hence I con- 
clude that the uniform change is 12 minutes 20 feconds annually. 
If this be fo, then the year, when there was no variation at Dub- 
lin, may be difcovered, by faying as 12 minutes 20 feconds is to 
one year, fo is 27 degrees 15 minutes (the variation found in 
1790) to the number of years elapfed, thus F 
E 
, Yr. i : 
12° 20 ry Se hh ip bangs : 132% years. 
From 1790 
Take 1324 
Remains 1657+, the precife time when there was no variation here. 
Nexr let it be required to find what the variation was in the 
year 1751, when Mr. Gibfon publifhed an account of it, 
From 1790 
Take 1751 
39 remains. 
od 
Multiply 12’ 20” or 740 feconds by 39, the produét is 28860 
feconds, which reduced is eight degrees one minute. This taken 
from 27° 15, the remainder is 19°14’, one-fourth of a degree 
more than Mr. Gibfon found it; from which it is evident 
the 
