r 330 ] 
going on, without any alteration of any kind : it is 
only by long and uninterrupted trials, that any judge- 
ment can be formed concerning the caufe of 
errors. 
The three firft months of the following tables 
are, perhaps, lefs accurate than they might have 
been. I was abfent from home great part of that 
time j and could only take fuch obfervations as oc- 
curred when I was occalionally in the country : 
hence the thermometer and barometer might be 
either higher or lower, in any of thofe months, 
than I have, here fet them down. I could truly 
give none but what I obferved. 
In the courfe of the fummer I received from 
Mr. Nairne a Smeaton’s hygrometer , which I had 
ordered the year before. 1 did not get it adjufted 
to my mind till the beginning of Aug-uft j but from 
that time, have added its higheft and lowed: Hate in 
each month, to thofe of the thermometer and ba- 
rometer. Its (ituation is the fame as that of the 
clock itfelf ; being fattened againft the fame wall, 
and clofe by its fide. I do not apprehend the rod 
of the pendulum to be affe&ed by fudden or fmall 
changes in the degree of humidity of the air ; 
though it feems to be fo by a long continuance of 
damp or dry weather. The hygrometer may per- 
haps (hew that. The general drynefs of fummer, 
and the thickening of the oil in winter, (as far as I 
have had opportunity for trial) I take to be the 
principal caufes of the change of rate in fuch a 
clock as mine. It now throws out rather lefs than 
it did : perhaps owing to its being lefs clean, or to 
the drying of the oih 
J772. 
