Defcription of a new Thermomeier. 63 
the mercury in the fmaller tube; to which float is joined a 
wire, termed the float-wire, reaching to H, having a knee 
bent at a right angle, which raifes one index, and deprefles 
another, according as the mercury rifes or falls. ‘This part 
of the apparatus is defended from wind or rain by means of 
aglafs tube 7+ inches long, clofed at top and open at bottom, 
fo wide as to {lide eafily over the feale, and, by means of a 
brafs rim cemented to it, made to fit exa¢tly to the circular 
bafe of the feale. This cover need not be taken off, except 
when the inftrument is to be prepared for an obfervation. 
[The operation of the float and indexes will be better un- 
derftood from fig. 4, where a fimilar apparatus, but belong- 
ing to a newly invented barometer, is reprefented en a larger 
feale.] 
F G is the {eale fixed to a circular piece of wood or brafs 
D, through which the fimall tube is made to pafs. From 
- Gto Fis a piece of the fmalleft gold wire ftretched along 
the feale, fixed at the ends by two pins. LL are two in- 
dexes, formed of thin black oiled filk, pierced by the fmall 
wire in fuch a manner as to flide upwards and downwards 
with a very fmall force, not more than two grains. H, the 
knee of the float-wire before defcribed, is made to encom- 
pafs the fmall wire between the two indexes, fo that, whem 
the float rifes, the upper index is moved upwards, and, 
when it defcends, it leaves the upper index ftationary, and 
pufhes down the lower one, which is alfo left ftationary 
when the float rifes. When the inftrument is to be prepared 
for an obfervation, the one index is to be pulled down, and 
the other raifed, by means of a bit of wire, bent for the 
purpofe, until both of them touch the knee of the float- 
wire: and when it is again obferved, the upper index will 
point out the greateft degree of heat, and the lower the 
greateft degree of cold, fince the time they were fet. 
If this thermometer is to be adapted to a piece of clock- 
work, in order to record the degrees of heat at each hour 
and minute of time, it ought to be made of larger dimen- 
' fions. 
