L 147 J 



XXV. On the Freezing Point of Tin, and the BoUhig Point 

 of Merc7iry, <with a Defcriplion of a Self-regiftering Ther- 

 mometer iwjented by Mr. James Crichton, of Glaf- 

 gozu *. 



oOME time ago, having made fome high ranged thermo- 

 meters, i wiflicd ro try their correfpondence at high tempe- 

 ratures. I procured for this purpofe 1^ pound of the beft 

 effayed tin (block), and, after fufpending two thevmomelers 

 upon a Aiding fupport, I mehed the tin in a heat of from 20° 

 to 30° above the mehing point : I then lowered my thermo- 

 meters down into the metal, and obferved the mercury to fink 

 graduallym both till it arrived at 442°; then it injlantly fuiik 

 to 439'', and as inftantaneoufly rofe to 442°. At this tem- 

 perature the mercury remained perfeftlv ftationary for five 

 minutes, at which time the metal became folid to the centre 

 of the crucible. After obferving this fa£t, which to me was 

 new, I waited upon Dr. Mickleham, and requefled that he 

 would be prefent when I fnould repeat the experiment, to 

 aflift me in difcovering the fallacy, if there was anv. We 

 then procured, from another work, the fame quantity of tin. 

 as before, and repeated the experiment feveral limes with the 

 fame refult. 



Another accurate ftandard point, that of 442% for adjufl- 

 ing thermometers, has thus been afcertained; a circumltance 

 which, judging by the ufe it will be of to myfelf in conflruft- 

 ing high ranged thermometers, I think too important to be 

 withheld from chemifts and others who may wifli to avail 

 themfelves of it. 



Before I quit the fubjecl of mercurial thermometers T beg 

 lo ftate another faft which may alfo be of ufe to the philofo- 

 phical world. Quickfilver does not boil at 6co°, but at 6^^"^ 

 at the lowcft. The precife point I have not yet found leifure 

 enough to afccrtain. 



I (hall now defcribe a regiflering tliermomefer fit for me- 

 teorological purpofcs, which I have lately invented and con- 

 ftruo^ed, and which fome of mv friends have requefted mc to 

 make public. Its a6lion depends on the different degrees of 

 expanfibility of diflerent rnetals when expofed to a change of 

 temperature. (See Plate III.) 



The length of the inUrument is about 13 inches. Fig. i. 

 rcprefcnts a front view of it, and fig. 2. a fide view : the Ict- 



* Comniunii-stecl by ihe Author. 



L % tera 



