[ x si 3 
height of the mercury. This, I believe, is feldom 
attended to in common barometers ; but it requires 
this exa&nefs in a barometer of this kind, as T '- inch 
rife or fall in the tube, is increafed to an inch in the 
range of a dial-plate io inches diameter. 
The axis of the index pulley, as alfo the regifters, 
are placed on friction wheels, as thofe of the ther- 
mometer 5 but it requires, that the work be made 
with greater nicety, in order to lay the lead: weight 
on the mercury. I therefore employed Mr. Vul- 
liamy, a watch-maker, and very ingenious mechanic, 
to make the machinery, which, on trial, has exceeded 
my expectation, as it requires but the weight of two 
grains to turn the regider and index freely. 
The weight c , which reds on the mercury, is 
made of ivory, in the fliape of a cone, hollow within, 
and made narrowing towards the bottom, with a 
fcrew in the middle to open ; fo that by pouring in 
a fmall quantity of mercury, you may readily adjud 
its weight, which is to be fo much heavier than the 
counterbalance d, as ferves to turn the index and re- 
gifters. The bottom of the weight c is made convex, 
in order to meet the fird rife of the mercury, which 
is obferved to fwell in the middle of the tube, before 
it can overcome the friction occadoned by the ddes 
of the glafs, and alfo to fink in that part fird; 3 by this 
means, a rife or fall of 3 or 4 degrees is often ob- 
fervable, by the index of this indrument, when the 
mercury in the common barometers feems to con- 
tinue dationary. 
The weights c and d are fufpended on filk threads, 
as wound off from the cocoons. This kind of filk, 
which is not twided, and has the natural gum on it,. 
probably 
