THERMOMETERS. 633 



to Galileo, Sanctorius of Italy, Drebbel of Holland, and to Robert Fludd. It 

 seems probable that Galileo invented it — i. e. the air-thermometer — about 1602, 

 for Castelli in 1638 says that Galileo had shown him one more than thirty-five 

 years before. It was of glass, with a bulb "about as large as a hen's egg and a 

 neck two palms long, and as narrow as a straw." A rude form of the liquid 

 thermometer was invented in 1655 or 1656 by the members of the Florentine 

 Academy, which possessed the advantage over the former of being hermetically 

 sealed and not being affected by the pressure of the atmosphere. Edmund Hal- 

 ley, the English astronomer, introduced mercury in place of alcohol in 1680, 

 though some writers give the credit of this use to Reaumur. 



The scale of degrees fixed by the Florentines had for its fixed points the 

 cold of snow on ice, and the greatest heat known at Florence, both rather variable 

 points. Von Guericke, an experimental philosopher of Magdeburg, Germany, 

 was the first to propose the freezing point of water as the lower limit of the scale, 

 and Isaac Newton seems first to have discovered or made use of the facts that a 

 thermometer placed in melting ice always indicates the same temperature and 

 boiling water, when under a given atmospheric pressure, always has an invariable 

 temperature. These points have been universally adopted by thermometer-makers 

 as being most nearly invariable and the most convenient in practice. To find the 

 freezing point, the bulb and tube are immersed in melting pounded ice or snow 

 to nearly the height to which such cold will lower the column, until both mer- 

 cury and glass are brought to the freezing temperature — the contraction of both 

 thus being secured — and this point is marked on the glass. Inasmuch as water 

 boils at various temperatures, dependent upon atmospheric pressure, purity of 

 the water, the kind of vessel used and the state of the surface of the water, it has 

 been found more accurate to take the temperature of the steam arising, which 

 does not vary with these conditions. Accordingly the tube is held in the steam 

 arising from boiling water at an atmospheric pressure of 29.92 until both glass 

 and mercury are at the same temperature, when the point indicated is also 

 marked. 



Having secured these two fixed points, the graduation of the instrument is 

 easy. The principal scales in use are Centigrade, Reaumur and Fahrenheit, the 

 first having been proposed about the year 1741 by Celsus, who took the freezing 

 point as the zero and the boiling point as 100°. This scale is most commonly 

 used by scientists, especially in Europe. Reaumur established his zero at the 

 freezing point and 80° as the boiling point. His thermometer is much used in 

 Germany. Fahrenheit fixed upon 32" as his freezing point and 212° as that of 

 boiling water, placing his zero at 32° below the freezing point. Various reasons 

 are assigned for these differences in the scales. None, however, has been offered 

 for that of Reaumur. As to that of the Centigrade, it is stated that according 

 to the historical notes contained in the meteorological observations of Lafou, 

 President of the Meteorological Commission of Lyons, the first thermometer ever 

 seen there was exhibited in 1736 by Duhamel. This thermometer was con- 

 structed with alcohol according to the scale of Reaumur. Christin, a member ot 



