1877.] 



Thermometers at the Kew Observatory. 



89 



of the apparatus. It will be sufficient if I give a few summary tables of 

 the results. 



The mean variation of temperature during a double process of com- 

 paring each of twenty sets of thermometers, each set averaging nineteen 

 instruments, and each instrument being read four times, was as follows : — 



Temperature at which the 1 

 comparison was made. ] 



50° 



70° 



80° 



90° 



100° 



110° 



Mean variation during \ 



+0-06 



±0-07 



+0-06 



±0-07 



+0-09 



±0-15 



The extreme variation of o, 30 occurred in one case, and that of 0°*25 

 in three cases. 



It takes about four minutes to read a complete set of ordinary 

 thermometers. 



The rate of heating by gas, and of cooling after the gas has been 

 wholly turned off, is of course much affected by the temperature of the 

 air of the room ; it may be roughly taken as follows : — 



When the water in the vessel is about ... 



45° 



100° 



150° 



200° 



Eise of temperature in 5 minutes when T 



2°-80 



2°-26 



l°-95 



l°-45 



Fall of temperature in 5 minutes when \ 





0°-25 



0°-60 



1°-00 



The rate of cooling is much reduced when the process consists in first 

 raising the water to the highest required temperature, and then cooling 

 it by successive additions of cold water. The heat of the stuffing that 

 surrounds the vessel being thus much higher than the water it contains, 

 keeps it at an equable temperature. 



The temperature of the water in the vessel, after agitating it and 

 allowing it to settle, differs somewhat at different levels ; this is due to 

 the impossibility of securing perfect intermixture and to the variations of 

 the temperature of the stuffing in respect to that of the water. The 

 greatest differences observed between a thermometer whose bulb was 

 immersed 2 inches below the level of the water and one that was im- 

 mersed 19 inches was o, 68. 



