498 LECTURE LII. 



difference of nine tenths of an inch either way requiring an alteration 

 amounting to T^- of the interval between freezing and boiling. This 

 interval is subdivided, in Fahrenheit's thermometer, into 180 degrees ; in 

 Reaumur's, into 80, and in the centigrade thermometer of Celsius and of the 

 French, into 100 ; and in making the subdivision, care must be taken to 

 examine the equality of the bore throughout, by observing the length occu- 

 pied by a detached portion of mercury, and to allow for any irregularities 

 which may have been thus detected. The scales of Reaumur and of Celsius 

 begin at the freezing point of water ; but in that of Fahrenheit the freezing 

 point stands at 32, the scale beginning from the cold produced by a 

 freezing mixture, which was supposed by Fahrenheit to be the greatest 

 that would ever occur in nature. 



The expansion, which is observed in a mercurial thermometer, is in 

 reality only the difference of the expansions of mercury and of glass ; but 

 this circumstance produces no difference in the accuracy of the results. The 

 separate effects of the expansion of glass are, however, sometimes per- 

 ceptible ; thus, when a thermometer is plunged suddenly into hot water, 

 the glass, being first heated, expands more rapidly than the mercury, and, 

 for a moment, the thermometer falls. This circumstance would perhaps 

 be still more observable in a thermometer of spirit or of water ; for an 

 equal bulk of these liquids would be much longer in acquiring the tempe- 

 rature of the surrounding medium than a mercurial thermometer. 



The expansion of elastic fluids affords in some cases a test of heat, which 

 is very convenient from its great delicacy, and because a very small quan- 

 tity of heat is sufficient to raise their temperature very considerably. The 

 thermometer first invented by Drebel was an air thermometer ;* but instru- 

 ments of this kind, when they are subject to the variations of the pressure 

 of the atmosphere as well as to those of its temperature, are properly called 

 manometers, and require, for enabling us to employ them as thermometers, 

 a comparison with the barometer ; while on the other hand, they may be 

 used as barometers, if the temperature be otherwise ascertained. They are 

 however, very useful even without this comparison, in delicate experiments 

 of short duration, since the changes of the barometer are seldom very 

 rapid ; and they may also be wholly freed from the effects of the pressure 

 of the atmosphere, in various ways. Bernoulli's method t consists in closing 

 the bulb of a common barometer, so as to leave the column of mercury in 

 equilibrium with the air contained in the bulb at its actual temperature, 

 and capable of indicating, by the changes of its height and of its pressure, 

 any subsequent changes in the temperature of the air, which must affect 

 both its bulk and its elasticity. Mr. Leslie's photometer,;}: or differential 

 thermometer, has some advantages over this instrument, but it can only be 

 employed where the changes of temperature can be confined to a part only 



* The invention is claimed for Drebel, by Boerhaave (Elem. Chimise, 2 vols. 4to, 

 Lugd. 1732, i. 152), and by Musschenbroek (Elem. Phil. Nat. 780) ; whereas 

 Santorio claims it as his own (Comm. in Avicennam, 1626), and his claim is sup- 

 ported by others. See Martine's Essays, Edin. 1787, and Dr. Traill's Thermome- 

 ter and Pyrometer, Lib. of Useful Knowledge. 



t Segner, De ^Equandis Thermometris Aeris, 4to, Gott. 1739. 



t On Heat ; and Nich. Jour. iii. 461, 518. 



