through Dry and Humid Air. 51 



temperature, the radiation being suspended, and the space around 

 the pile a vacuum. Suppose, in the first instance, the tempera- 

 ture of the air outside to be lower than that of the pile, that the 

 pile, in other words, is a warm body in comparison with the 

 air ; what will be the effect of admitting the air into the vessel ? 

 The reader will more easily understand me if he refers to the 

 Plate which accompanies the paper of Prof. Magnus in the 

 present Number of the Philosophical Magazine. On the upper 

 face of the pile will rest a column of air, which is heated at its 

 bottom by the surface on which it rests ; convection will imme- 

 diately set in, and heat will be continually abstracted from the 

 face of the pile. At the lower face, on the contrary, an equal 

 abstraction does not take place ; the air once warmed remains in 

 contact with the face of the pile, convection here being almost 

 nil. Thus a less amount of heat is abstracted from the lower 

 than from the upper face of the pile, and hence the instrument, 

 which before the entrance of the cool air produced no current, 

 will, in virtue of the different action of this air on its two oppo- 

 site faces, generate a current similar to what would be produced 

 by the direct heating of the lower face of the pile. 



A moment's reflection suffices to prove that precisely the same 

 deflection is obtained when the external air is hotter than the 

 pile. Supposing, as before, the temperature of both faces to be 

 the same at the commencement, the needle of the associated 

 galvanometer being at zero. When the warm air enters it is 

 chilled by the upper face of the pile, contracts, and remains in 

 contact with that face, forming in fact a pool of heavy air at the 

 bottom of the reflector. The air chilled by the opposite face of 

 the pile falls by its weight ; its place is supplied by fresh warm 

 air, which again falls and is replaced. Thus it is evident that 

 the lower face of the pile will in this case be more heated by 

 the air than the upper one ; and hence we infer that whether 

 the external air be colder than the pile, or hotter than the pile, 

 the same galvanometric effect follows its introduction into the 

 vessel. 



Instead of supposing the pile to be in the first place of uni- 

 form temperature, let us imagine it exposed to the radiation 

 from the source. This makes the upper face warmer than the 

 under one, and produces a deflection commensurate with the 

 difference of temperature of the two faces. Let air now enter : 

 it is manifest from the foregoing analysis that, whether this air 

 is colder than the pile or hotter than it, its effect will be to 

 render the lower face relatively warmer, and thus to diminish the 

 deflection. If, moreover, the air be of the exact temperature of 

 the upper face, it will warm the under one, if of the exact tem- 

 perature of the under face it will chill the upper one. If its 



E 2 



