COMPARISON 



THE BAROMETRICAL SCALES. 



THE following tables are intended for converting into each other the four most 

 important Barometrical Scales. They are sufficiently detailed to save the labor of 

 any calculation or even of interpolation for the ordinary wants of Meteorology. But 

 before making use of them, for comparing the observations taken with barometers of 

 different scales, it is necessary to reduce the observed heights to the temperature 

 of the freezing point, or to any other temperature, provided it be the same for all, 

 by means of the tables calculated for this purpose, and which will be found below. 

 The reason of it may be readily understood. 



The length of the bars of metal, or of other substances, which represent the stand- 

 ard measures of length which obtain among different nations, varying with the tem- 

 perature, it was necessary to determine a fixed point of temperature at which they 

 really ought to have the length adopted as the standard unit of measure. This 

 temperature is the normal temperature of the standard, and the length of the stand- 

 ard-bar, at this temperature, is the true length of it. 



If the normal temperature of the various standards used for dividing Barometrical 

 Scales were the same, the heights of the barometrical column, taken with these 

 scales, could be compared directly, provided the scales be made of the same sub- 

 stance, brass, for instance, because their variations above or below this normal tem- 

 perature would remain parallel with each other. But unfortunately it is not so. 

 The English Yard is a standard at the temperature of 62 Fahrenheit ; the Old 

 French Toise, at 13 Reaumur ; the Metre, at the freezing point, or zero Centigrade. 

 Thus metallic rods intended to represent these various units of measure give the true 

 standard length only when at these respective temperatures ; at any other tem- 

 perature they are longer or shorter than the standard, and their subdivisions, inches, 

 ines, or millimetres, partake of the error. 



It is obvious, therefore, that the barometrical heights, taken with different scales, 



cannot be compared directly by means of the following tables, which give the re- 



ation between these scales at their respective normal temperatures. For suppose 



he temperature of the three barometers to be the freezing point, or 32 Fahrenheit, 



C 7 



