Chap. 3-] Beft^Form of thermometers. 113 



fome other ftopper, to prevent the air from rufhing in. 

 When by thefe means a quantity of mercury is in- 

 cluded in a glafs tube with a fmall bulb, the glafs may- 

 be eafily clofed by applying an ignited 'charcoal and a 

 blow-pipe, fuch as the jewellers make ufe of, which 

 melts the glafs, and enables us to twift it round, in 

 fuch a manner as completely to clofe it againft the ad- 

 miiFion of air; and this operation is called hermetically 

 fealing it ?. In order to graduate the thermometer, 

 it muft be firft immerfed in a mixture of pounded ice 

 or fnow and fal ammoniac, and the point at which the 

 mercury fettles muft be marked as the commence- - 

 ment of the fcale, or o. It is next to be immerfed in 

 boiling water, and that point is to be marked 212, 

 and the intermediate part of the thermometer muft be 

 divided into this number of degrees. 



Thermometers with final! bulbs, and proportionable 

 cylinders, are moft ufeful, fmce a large volume of mer- 

 cury requires a confiderable time to heat and cool. It 

 is alfo accounted a favourable circumftance when that 

 part of the bulb which is adjoining to the tube is rather 

 of a conical~form. 



That the thermometer is a true meafure of heat, is 

 proved by fome very fatisfactory experiments. If 

 equal quantities of hot and cold water are mixed toge- 

 ther, the heat of the mixture will be nearly as the mean 

 heat of the two component parts f. This fact was 



* It is eafy to prove whether the tube is a perfeft vacuum or 

 no)!, after it is hermetically fealed, by merely inverting it, and ob- 

 ferving whether any bubble of air remains to refill the mercury's, 

 falling to the botcpm of the tube. 



f This experiment was originally made by Dr. Brook Taylor. 

 It was afterwards repeated by M. de Luc, and by Dr. Crawford ; 

 who, from the impofllbility of conducting the experiment without 

 lofs of time, found the heat of the mixture always below the arith- 

 metical mean. Crawford on Heat, p. 1 8, e t feq. zJ edit. 



VOL. I, I alcertained 



