VOL. L.] PHILOSOPHICAL TRANSACTIONS. JSQ 



sists of a cylinder of glass joined to a tube, and differs from common thermome- 

 ters only in having the top of the stem drawn out into a capillary tube, which 

 enters into a glass ball c, joined on to the stem at the place where it begins to 

 be contracted. The cylinder, and part of the tube, are filled with mercury ; the 

 top of which shows the common degrees of heat as usual. The upper part of 

 the tube above the mercury is filled with spirit of wine, and some of the same 

 liquor is left in the ball c, so as to fill it almost up to the top of the capillary 

 tube. Now when the thermometer rises, the spirit of wine will be driven out of 

 the tube, and will fall into the ball c. When the thermometer sinks again, as 

 the spirit cannot return back from the ball, the top of the tube will remain 

 empty, and the length of the empty part will be proportional to the fall of the 

 thermometer. Therefore, by means of a proper scale, the top of the spirit of 

 wine will show how many degrees it has been higher than when observed ; 

 which being added to the present height, will give the greatest degree of heat it 

 has been at. 



To fit this thermometer for a new observation, it is necessary to fill the upper 

 part of the tube with spirits ; which may be done by inclining the instrument 

 till the spirits in the ball c cover the end of the capillary tube. For if the cylin- 

 der is then heated, by applying the hand to it, or by the flame of a lamp held at 

 some distance, till the spirits rise to the top of the tube and run over into the 

 ball c, and is then suffered to cool in the same position, the tube will remain 

 full of spirits, and the thermometer will be fitted for a new experiment. 



The top of the capillary tube is made to stand pretty near to one side of the 

 ball, and also to the top of it, that a less inclination of the instrument may be 

 sufficient to make the spirit of wine in the ball cover the end of the tube. The 

 ball c is joined on as high as possible, so as to hide no part of the tube, except 

 that where the bore is contracted. By this means the top of the spirit of wine 

 begins to appear before the thermometer has sunk one degree. It is convenient 

 to leave some mercury in the ball c, which may be made to cover the end of the 

 capillary tube, by inclining the thermometer more than what is necessary to make 

 the spirit of wine cover it. By this means some mercury may be got back into 

 the tube, in case any of it should happen to be driven into the ball by the ther* 

 mometer's being exposed to too great a heat. 



The scale of degrees at top, which shows the descent of the therniometer 

 from the highest point it has arrived at, ought not in strictness to be the same at 

 all times of the year ; for those degrees exceed the common degrees of heat 

 pointed out by the top of the mercury, as much as the column of spirit of wine 

 expands, and therefore are greatest when that column is so ; that is, when the 

 greatest heat to which the instrument has been exposed is least. A difference 

 of 30 degrees of Fahrenheit's scale, in the greatest rise of the thermometer, 



t2 



