266 MR JOHN AITKEN ON THE 



of the hot surface seems to be to differentiate the movements of the molecules, 

 causing the greater amount of the movement of the hot ones to be outward 

 and of the cold ones inward, and the outward moving molecules, having the 

 greater kinetic energy, exert a greater pressure on the dust particles and drive 

 them outwards. In the hydrogen diffusion effect the particles of dust were 

 driven away, because a greater number of hydrogen molecules were moving one 

 way than air ones the other. In the heat effect they are driven away, because 

 the molecules moving from the hot surface have a greater kinetic energy than 

 those moving towards it, and the particles are bombarded on the one side by a 

 greater number of hot molecules than on the other. 



We have the same effect intensified when the hot surface is wet. When 

 this is the case, the vapour molecules diffusing outwards carry with them the 

 dust particles to a much greater distance than the heat alone, as there is no 

 inward current of vapour molecules to contend with the outward one, and tending 

 to drive the dust particles inward ; the result is, we get a dark plane at 

 least twice as thick with heat and vapour as with heat alone. Of the two, the 

 vapour seems to be the more powerful, as very little heat with moisture gives a 

 thicker dark plane than double the heat would do. If we carefully fix the 

 experimental test surface in a vertical position and simply wet it, the effect is 

 to cool it by evaporation, and a downward current is produced ; but, at the 

 same time, a clear space is formed, showing that in this case the outward 

 effect of the vapour is greater than the inward effect of the cold. 



There seemed to be a possibility of getting an answer by experiment as to 

 whether the radiation or the diffusion theory is the correct one. If radiation is 

 the cause of the repulsion, then we should expect that a good radiator would 

 cause the particles to be driven further away, and thus cause a thicker dark 

 plane than a bad radiator. For the purpose of testing this, another experi- 

 mental flat test-surface was prepared. This test-surface was made of silver and 

 highly polished. One-half of it was then covered with lamp-black. After the 

 test-surface was fixed in the dust-box heat was applied to it, and the thickness 

 of the clear space over the two halves of the test-surface carefully noted. To 

 do this, the dust-box was so arranged that I could look down the test-surface — 

 not across it as usual — and could thus see down the boundary line between the 

 dark plane over the polished surface and over the lamp-black. The result was, 

 not the slightest difference could be detected between the two. The boundary 

 line of the dark space in front of the plate was a straight line parallel to the 

 surface of the plate. This experiment, while it gives no support to the diffusion 

 theory, shows us that radiation is not the principal cause of the dark plane. 



If the explanation here given of the repulsion of the dust by a hot surface is 

 correct, then this effect is not produced in the same way as the repulsion of the 

 discs of a Crookes' radiometer when heat is falling on them, but is similar to 



