396 Of the Propagation of Heat 



As the windows in the room in which this experiment was 

 made are double (as are all those both in summer and 

 winter in the apartment I inhabit), and as the apparatus 

 above described occupied the place of a pane of glass 

 belonging to the inside window, it was in my power, by 

 opening either the inside window or the outside window, 

 to cause the Heat on the two opposite sides of the box 

 to be either equal or unequal at pleasure; and by varia- 

 tions which that arrangement enabled me to make in the 

 experiments, I produced several interesting appearances. 



There was one very striking appearance indeed which 

 never failed to present itself regularly every day during 

 the three weeks that the experiment was continued.* 

 The clouds, after having been driven about all day by 

 the different currents in the liquid (of which there were 

 sometimes as many as six or seven running in opposite 

 directions at the same time), never failed to collect them- 

 selves together in the evening into large masses ; some- 

 times forming only one, and sometimes two or three 

 strata at different heights, where they remained to all ap- 

 pearance perfectly motionless during the night. 



There can be no question with respect to the proximate 

 cause of this phenomenon, for it was undoubtedly owing 

 to a diminution or total cessation of the operation of 

 that cause of those causes, or of some of them 

 by which an inequality of temperature in the liquid 

 was produced and continued; but it would be highly 

 curious to investigate the more remote causes of this ap- 

 pearance, and see how far light ^ or rather the absence of 

 it, was concerned in producing it : but that discussion 

 would lead me into a very abstruse inquiry, that re- 



* An end was put to the experiment by an accident ; the box being broken by the 

 carelessness of a servant in shutting the window-shutter. 



