l64 PHILOSOPHICAL TRANSACTIONS. [aNN0 1781. 



ments, disencumbered from words. Here the decimals are not computed to a 

 very large number, that being unnecessary for this purpose. 



Weight of the glass 79-25 grs. 



Weight of so much quicksilver as filled 4.3 inches length of the tube, 0.24 

 Weight, of the whole quantity of quicksilver contained in the therm. 32.05 



Length of the tube equal to 20° 1 .33 inch. 



4.3 : 0.24 :: 1 .33 : 0.0742 = 20° 



20°: 0.0742:: 1 -.0.00371 



32.05 : 0.00371 :: 1 : 0.0001 1 + = to the expansion occasioned by 1° of heat. 



Having repeated this experiment with other thermometers, and by similar 

 calculations, each process gave a result little different from the others, which 

 irregularity is certainly owing to the imperfection of my scales, which are not of 

 the nicest sort: but taking a mean of various experiments it appears, that 1° of 

 heat, according to Fahrenheit's thermometer, increases the bulk of a quantity of 

 quicksilver by T o 9 u ir parts, viz. if the bulk of a quantity of quicksilver in the 

 temperature of 50° is equal to 100,000 cubic inches, the bulk of the same 

 quantity of quicksilver in the temperature of 51° will be equal to 100,009 

 cubic inches. 



From these observations the method of graduating, or of determining the 

 length of a degree in a new thermometer, is easily deduced, the only requisites 

 for the calculation being the weight of a quantity of quicksilver, which fills a 

 known length of the tube, and the weight of the whole quantity of quicksilver 

 contained in the thermometer when filled. Suppose, for instance, that in 

 making a new thermometer it be found, that the weight of so much quicksilver 

 as fills 5 inches length of the tube is equal to 10 grs., and that the weight of 

 the whole quantity of quicksilver contained in the thermometer weighs 300 grs. 

 It is plain, that if the whole quantity of quicksilver weighs 300 grs., then 

 ___9___ parts of it must weigh 0.027 grs. But the weight of so much mercury 

 as fills 5 inches of the tube is equal to 10 grs.; therefore, 0.027 grs. weight of 

 quicksilver must fill 0.0133 inch of the tube, and this is equal to the length of 

 1°, or the double, treble, &c. of it is equal to 2, 3, &c. degrees. 



By this means the scale may be made; that is, it may be divided into degrees, 

 but the numbers cannot be added to them without finding which of those degrees 

 corresponds with the freezing point or boiling point. Either the point of boiling 

 or freezing may be found by experiment, or any other point may be ascertained 

 by comparison with another thermometer, and then the other degrees are nomi- 

 nated accordingly. 



3. Description of a thermometrical barometer. — The determination of the 

 various degrees of heat shown by boiling water, under different pressures of the 

 atmosphere, has been attempted by various persons, but it was lately completed 



