o i 2 An Attempt to make a Thermometer 
III. Th is clay, dried by the fummer heat, or in a mode- 
rately warm room, or with more heat before a fire, has not 
been obferved to differ in degree of drynefs. After being fo 
dried, it lofes about a hundredth part of its weight in the heat 
of boiling water, about as much more in that of melted lead, 
and from thence to a red-heat ten parts, in all -f A- Each of 
thel'e heats loon expels from the clay its determinate quantity 
of matter, chiefly air ; after which, the fame heat, though 
continued for many hours, has no further effect. I had fome 
hopes, that the graduation of the common thermometer might 
be continued, upon this principle, up to the red-heat at which 
the fhrinking of the clay commences, fo as to connect the two 
thermometers together by one feries of numbers ; but the lots 
of weight appears not to be fufficiently uniform or proportional 
to the degree of heat to anfwer that purpofe ; for it was found 
to go on quicker, and bladders tied to the mouths of the veflels 
in which the pieces were heated, became more rapidly diftended, 
at the commencement of rednefs than at any other time. From 
low red-heat to a ftrong one, fuch as copper melts in, the lofs 
of weight was only about two parts in a hundred ; though the 
difference between thefe two heats appears to be much greater 
than what the lame lofs correfponds to in the lower ffages. 
Alter this period, the decreafe of weight intirely cealcd. 
The vapours expelled from the clay, caught feparatcly in 
the different degrees of heat, feemed, from the few trials 
made with them, to con fill ot common air mixed with fixt 
air. They all precipitated lime-water ; that which was firfl 
extiicated, exceeding weakly ; the others more and more confi- 
aerably ; but the lalf not near lo Ifrongly as the air expelled 
£rom lime- hone in burning. None of them were inflammable. 
IV. 
