308 Relation ietweefi the Specific Graviftes aM 



in spirituous compouncls of all strengths, a gallon of any 

 such liquor would at any degree of heat contain a certaift 

 proportion of its bulk of alcohol, and its per-centage would' 

 therefore not vary, whatever was its temperature, though a 

 correction would be necessary with reypect to the aclual 

 measure of each, We have already said, however, that this 

 is not the case (§ 9); and it is also true, that not only the 

 quantity of this variation is diflerent in respect of their 

 strengths, but the law of its progression varies still more. 

 If the expansion of mercury be considered as uniform, as it 

 appears to be very nearly between the freezing and boiling 

 water points, the expansion of alcohol is progressively in- 

 creasing, but in a very small degree, between 30° and 80** 

 of Fahrenheit's thermometer J whereas water actually con- 

 tracts by elevation of its temperature till it reaches 40°, 

 after which it again begins to expand in a very increasing 

 progression, its expansion of bulk between 70" and 80° 

 bemg more than 3 times as great as that which takes place 

 between 40° and 50°. The expansion of mixtures of these 

 two fluids will in both respects approach nearest to that of 

 the predominant ingredient ; but the exact quantity of it, 

 or the law of its progression in any such compound, can 

 only be ascertained by experiment. 



We miist therefore apply such corrections in our process 

 as may give us the quantity of proof spirit by weight; or, 

 •which amounts to the sanie thing, by measure when re- 

 duced to 60°, which is equivalent to 100 parts by measure 

 ot any spirit at any given temperature at which it may be 

 measured and proved. 



§ 26. The ultimate conclusion from the premises laid 

 down in this chapter with respect to the mode of denomi- 

 Jiating the strengths of spirits differing from proof is, that 

 it will be convenient that the scale should be graduated in 

 -per cents indicating the relative values of each compound 

 with respect to the common standard of proof ; or the 

 number of parts of proof spirit by measure at 60°, which 

 woiild produce or be producible from 100 parts by measure 

 -at any given temperature of any given spirituous liquor. 

 Taking, therefore, the strength or value (for, cceteris pari- 

 In/s, they are in the same ratio) of proof spirit as denoted 

 by the number 1 00, some other number between unity and 

 170 u ill represent that of every other spirituous compound. 

 If, tor example, we mean to express a kind of spirit of 

 fluch a strength, that, on 100 parts of it by measure at the 

 existing temperature being reduced to G0° of Fahrenheit's 

 theruiometer, and then made up to 134 with water, it 

 4 should 



