378 T. Russell—Calibration of Thermometers. 
_ voir again. Then take the thermometer in the right hand, tap 
the index finger of the left hand with the reservoir-end, hold- 
ing the instrument about horizontal. This will detach a 
greater or less portion of the mercury, depending on the inten- 
sity of the shock. If what is left is not of the required length 
bring back the mercury to the entrance of the reservoir ; a few 
taps on the finger, as before, will cause the mercury to reunite. 
This process can be done over and over until a column of the 
it does.get into the reservoir, or there is not enough in the 
An alcohol lamp is preferable for this purpose. The expan- 
sion of what little air there is in the reservoir will almost 
always drive down the mercury, but sometimes it requires the 
volatilization of the mercury itself. This difficulty of getting 
the mercury to run down the tube is in the way of obtaining 
short columns, especially in very capillary tubes. 
Short columns can best be obtained from the bulb, in case 
mometer in the right hand and hold it inclined, with the bulb 
down. A few light shocks by the sudden stopping of a rapid 
bubble to the entrance of the tube. The expansion of the mer- 
cury as the temperature rises will drive the short column before 
it. By this process also the mercury may be perfectly reunited 
after the calibration is completed, by cooling until all the mer- 
cury runs into the bulb, when a little jarring of the thermom- 
eter, as before, will cause the bubble to disappear. ; 
Another method of getting columns of definite length 1s 
described in Fischer's Geodésie, credited to Hansteen. The 
pinciple consists in getting a small bubble of air in the column 
