Life of Count Rumford. 461 



He, however, deferred these experiments till another 

 incident, two years subsequently, freshened his curiosity. 

 While experimenting on the communication of heat, he 

 had prepared several thermometers of an uncommon 

 size, their globular bulbs being above four inches in 

 diameter. These he had filled with various kinds of 

 liquids. One of them containing spirits of wine, poured 

 in as hot as the glass tube would endure, he placed to 

 cool in a window where the sun was shining. The 

 divisions on the tube were marked by a diamond on the 

 glass. The bulb, which was of copper, having been laid 

 aside for two years, and its orifice not being filled with a 

 stopple, some fine particles of dust had found their way 

 into it. These particles, intimately mixed with the 

 spirits of wine, helped to show the whole mass of liquid 

 through the thin, transparent, colorless glass of the 

 tube, in a most rapid motion, running swiftly in two 

 opposite directions, up and down, at the same time. 

 On examining the instrument with a lens, the Count 

 observed that the ascending current occupied the axis of 

 the tube^ while the descending current followed its sides. 

 When the tube was inclined, the rising current moved 

 out of the axis and occupied the uppermost side, the 

 descending current making use of the lower side. When 

 the cooling of the spirits of wine was hastened by wet- 

 ting the tube with ice-cold water, the velocities of both 

 currents were accelerated ; and the motion ceased when 

 the instrument and its contents had acquired nearly the 

 temperature of the air of the room. The motion was 

 prolonged by wrapping the bulb of the thermometer in 

 furs, or any warm covering. The appearances were the 

 same when the experiment was tried with a similar ther- 

 mometer filled with linseed oil. The observer at once 



