^ROM THE HIGHER ATMOSPHERE. 4S9 



a train of damp straw set on fire, was, from the difficulty of 

 managing it, found to lead to no certain conclusion. 



But during winter, a much easier mode occurred for decid- 

 ing the question. In a room where a steady fire was kept up, 

 the sectoral aethrioscope was set on the inside of the window, 

 and directed to the upper part of the opposite wall ; but on 

 throwing up the window, the instrument being now surround- 

 ed by a body of cold air, which, however, did not penetrate far 

 into the room, the liquor sank 5 or 6 degrees, indicating im- 

 pressions of heat, caused evidently by the excess of tempera- 

 ture of the remote air of the room above that which was con- 

 tiguous to the aethrioscope. It need hardly be observed, that the 

 effect increased in colder, and diminished in milder, weather. 



A similar experiment is readily performed by help of the erect 

 aethrioscope. In a close apartment, where a good fire is con- 

 stantly kept up, the ceiling and the floor may be discovered by 

 the pendant differential thermometer to have exactly the same 

 temperature with its adjacent stratum of air. Yet the upper 

 portions of the confined air of the room will be found several 

 degrees warmer than the lower. Instead of being divided onlv 

 into opposite ranges, the whole mass, from the floor to the ceil- 

 ing, will, in consequence of the expansion and buoyancy of its 

 heated particles, form a series of intermediate strata, not dis- 

 tinguished, however, by any very precise boundaries. But the 

 intensity of action being proportional to the difference of tem- 

 perature, the effect on the aethrioscope must evidently be the 

 same, whether it is produced by a single set of large pulses or 

 by several sets of smaller ones. If, for example, instead of one 

 bounding surface, above which the air is seven degrees warm- 

 er than immediately below it, we suppose seven such boundaries 

 each having an excess of temperature of only a degree : the 

 pulses excited at the first of these intermediate surfaces, and 

 successively augmented as they reach the second, third, and 



Vol. VIII. P. II. 3 Q fourth, 



