[ 3°9 ] 
touch the end of the tube, in order to fit the infiru- 
ment for a new experiment. 
If, by means of a bladder, the fea- water can be 
kept out of the glafs, this inftrument may be made 
to (hew the common degrees of heat ; but even then, 
in order to render it lefs bulky, it may be better to 
fupply the want of them as in the lafl inftrument. 
The longer kg of the fyphon may in that cafe be 
made as Ihort as you pleale ; only making the ball B 
big enough to receive all the mercury, which may 
be driven into it by heat. 
If thermometers of this kind were to be fent up 
into the air by means of a kite, they might be made 
like thole propofed for the fea j but it would not be 
necefiary to leave them open. 
As it would be definable to make them as final! as 
pofiible, they fhould be made fo as not to (hew the 
common degrees of heat : and it would alfo, on 
that account, be convenient to omit the thread of 
glafs placed within the fhorter leg of the fyphon in. 
fig. 3. and 4. This thread of glafs is placed there 
in order to make the mercury fall into the ball A. 
fig. 3. or cylinder C. fig. 4. in fmaller drops, and alfo 
to facilitate the filling the fhorter leg with mercury. 
The latter purpofe may be anfwered by having a ball 
blown towards the bottom of the fhorter leg, as. 
marked in fig. 4. at M : for as foon as the mercury 
driven out of the cylinder by heat gets to that ball, 
it will pafs by the fpirit of wine. The other pur- 
pofe may probably be anfwered by having the tube 
contracted as much as pofiible at n. 
In the foregoing inftruments the tubes made ufe 
of were of a large bore,, as mod of the errors in 
them 
