[ 292 ] 
witli a pendulum of fo fimple conftniiSlion, is not eafily 
affedte'd by any tremulous motion of the building to 
which it is faftened. In the months of IMarch, April, and 
part of May, 1 had occahon to make alterations in the 
top of my houfe, in order to gain more rooms m it; 
and notwithftanding the great jarring heceffarily confe- 
quent upon taking off the old rafters, and laying on a 
new leaded roof, and new joiifs and floor over the obfer- 
vatory itfelf, the clock feems not to have been difor- 
dered at all by it. Between February 7 th and 20th 
there will appear an error in the calculation of gain, to 
any one who fliall take the trouble to examine it : not 
that I believe there is really any error in it; but by an 
accident in winding the clock (not having put down the 
fpring fufficiently, which is intended to keep on the mo- 
tion of the wheels,) there were 6 " loft, as appeared by 
the afliftant clock, which had been fet with it Juft before. 
Thefe 6 " being allowed for, will reduce the lofs of 
n Si 
2,1 to a gain of 3,9 in that interval of thirteen days, 
it ti n 
-2,i+ 6 = + 3,9. In the months of February, March, 
and April, I was frequently from home ; fo that the ftate 
of the thermometer and barometer, if I were to fet them 
down, would be very imperfeeft. In the other months 
they are more compleat ; yet there may, perhaps, in my 
abfence, have been fome days in them either higher or 
lower than what are here given. 
1773 - 
