14 



THERMOMETER AND PYROMETER. 



is much more difficult, because it is 

 affected by atmospherical pressure, and 

 even by the form of the vessel in which 

 the water is heated. 



The Committee of the Royal Society* 

 recommend that the boiling point 

 ought to be fixed under a barome- 

 trical pressure of 29.80 inches. For 

 the graduation of the thermometer they 

 recommend that the bulb should not be 

 immersed in the water; because they 

 found, that according to the depth of 

 this immersion the mercury rose to a 

 greater height in the tube. They re- 

 commend a vessel of tin plate, pro- 

 vided with a cover which fits easily on, 

 and rendered steam-tight by a ring of 

 woollen cloth between it and the vessel. 

 This cover has two apertures a chim- 

 ney, with an area not less than half a 

 square inch, and two or three inches 

 high, to carry off the steam of the boil- 

 ing water ; and a hole for a cork, 

 through which the thermometer tube 

 is inserted in such a manner, that the 

 ball does not touch the surface of the 

 water, but may be surrounded with an 

 atmosphere of steam ; while no more of 

 the tube should be above the cork than 

 is sufficient to show the height to which 

 the mercury rises when the water is 

 briskly boiling. When all things are 

 thus adjusted, a thin plate of metal is to 

 be laid over the chimney, to prevent the 

 escape of the steam as it is formed ; heat 

 is to be applied to the bottom of the 

 vessel ; and when the mercury has re- 

 mained a few minutes stationary in the 

 atmosphere of steam, its height is 

 carefully to be marked with a file on 

 the tube. 



The water may be distilled, or any soft 

 water, such as clear rain water, be used ; 

 for, if there be much saline ingredient 

 in the water, this will affect the boiling 

 point, and may lead to error. 



Various mechanical contrivances have 

 been proposed for more conveniently 

 fixing the tube in the cover, but they are 

 of little comparative importance. Some 

 prefer plunging the ball into the water 

 to the depth of two or three inches : in 

 this case there is no necessity for a plate 

 of metal on the chimney, nor for the 

 tightness of the cover; but the adjust- 

 ment of the boiling point is to be made 

 for the barometer at 29.50 inches. To 

 those unprovided with such a vessel 

 the following method is recommended. 

 Wrap several folds of linen, or flannel, 



* Phil. Trans, vol. Ixvii. part ii. 



round the tube, nearly as high as the 

 supposed boiling point, which may be 

 guessed at by previous immersion of 

 the bulb in boiling water : hold the ther- 

 mometer in an ascending current of 

 boiling rain water about two or three 

 inches below the surface ; pour boiling 

 water three or four times on the~ cohering 

 of the tube, at intervals of some seconds ; 

 and waiting a few seconds, after the last 

 affusion, to allow the water to be in 

 brisk ebullition, mark the height of the 

 mercury in the tube, which will be the 

 boiling point of the instrument. 



Having thus obtained two fixed 

 points, the freezing and boiling points 

 of water, it is easy to mark off corre- 

 sponding divisions on the scale which 

 is to be graduated. If the tube be truly 

 cylindrical, nothing more is necessary 

 than to divide the intervening space into 

 as many equal parts as it is^ intended to 

 have degrees between those points. 

 Should the tube not be of uniform bore, 

 the size of the divisions ought to be ac- 

 commodated to the inequalities of the 

 tube. This may be done by taking in- 

 termediate points in mixtures of water 

 at different temperatures ; and after 

 marking them on the tube, proportion- 

 ing the size of the degrees, at short in- 

 tervals, to the varying diameter of the 

 tube. This method of graduating from 

 intermediate points ought, in nice in- 

 struments, to be adopted, however true 

 the tube may appear ; but a tube with 

 sensible inequalities is in general to be 

 avoided. 



Although it would be advisable to fix 

 the boiling point when the barometer 

 is at the height above recommended, 

 this may be attended with serious in- 

 convenience to artists ; and philosophers 

 have therefore investigated the correc- 

 tion to be made for every ordinary varia- 

 tion of atmospheric pressure. 



The first considerable series of ex- 

 periments on this subject are those of 

 De Luc, in 1762, published in his in- 

 teresting Recherches sur les Modifica- 

 tions de I" Atmosphere,* which were 

 extended and verified by Sir George 

 Shuckburgf in 1775 and 1778. Em- 

 ploying Reaumur's scale, De Luc as- 

 certained, that if y represent the 

 height of the barometer, T the height of 

 the thermometer above the freezing 

 point, expressed in hundredths of a 

 degree of this scale, when immersed in 

 boiling water ; and a the constant number 



Vol. i. 382 ; vol. ii. 338. 

 t Phil. Trans, vol. Ixix. partii. 



