shaken down, and then they are stabilized, reversed, and 

 read a second time. With the thermometers still reversed, 

 the tank temperature is raised to 30°C and the increase in 

 scale reading recorded for each instrument. Finally, the 

 thermometers are removed from the tank and returned to 

 storage. 



DATA REDUCTION 



The calibration sheets are completed by determin- 

 ing main stem corrections from the average of the two 

 reversal readings. TheV^'s are computed from the average 

 of the two reversals at 0°C, the rise after the second re- 

 versal at 30°C, and the K value. 



PROCEDURE FOR DETERMINING THE Q FACTOR 



The pressure factor, Q, of unprotected reversing 

 thermometers is determined by use of a steel vessel in 

 which six instruments can be held in pressure equilibrium 

 with a sensitive deadweight balance (Crosby, Model CD-2). * 

 The pressurized vessel can be rotated 180° in a vertical 

 plane while suspended in the constant temperature bath. 



The rack holding 24 thermometers is removed 

 from the tank and the pressure vessel lowered into it. 



The outer glass jackets of six unprotected 

 thermometers are filled with the alcohol-water mixture 

 used in the tank. The mixture is retained in the ther- 

 mometer by use of a rubber finger cot of suitable length to 

 extend past the mounting holes after being rolled over the 

 open end of the glass shell. Care is taken to exclude all 

 air. The thermometers are placed in a circular rack 

 (fig. 9) in an upright position and lowered into the pressure 

 vessel, which can be opened at either end. 



The vessel is sealed and an internal pressure 

 of 100 kg/ cm' ±0. 25 percent is applied and maintained. One 

 hour is allowed at 0°C for temperature stabilization of tlie 



*A Type 2400H RUSK A deadweight gage is now available. 



24 



