'1- 



[ 105 1 



REMARK IV. 



On the Mercury. 



The mercury Ihould be pcrfedly well cleaned ; this is beft done 

 by repeatedly waftiing it in a vial with fuccefltvc frelh waters, 

 and when dried pouring it through a pin-hole made in a white 

 paper cone. 



The mercury may (if required) be boiled in the tube ; but 

 I apprehend there are fome confiderable objcdions againft this 

 mode. 



R E M A R K V. 



On the Thermometers. 



The attention to the flate of the attached thermometer, fo 

 neceffary in the ufe of any barometer for meafuring heights, 

 is particularly fo in this ; as from the fize and material of the 

 ciftern or cylinder its dimenfions are liable to pofTible changes. 

 To obviate this objedion the fides of the cylinder ftiould be 

 made pretty thick and ftrong, as otherwife they might be aded 

 upon in fome degree like Mr. De Luc's hygrometers, and a 



Vol. V. O change 



