FORMATION OF SMALL CLEAR SPACES IN DUSTY AIR. 265 



side getting heated by contact, would rebound from it with greater velocity 

 than those on the other side, the dust particles being thus driven away in a 

 sort of rocket fashion. On examination, however, this explanation does not 

 appear satisfactory ; because the particles are so very near the hot surface that 

 they will not be heated principally by radiation, but by contact with the hot 

 gases near the heating surface, radiation having but a slight effect. 



So far as I have been able to form a mental picture of the mechanism of 

 this repulsion, it seems to be produced in the following way :— First, let us go 

 back to the diffusion experiments. We saw that when hydrogen was diffused 

 into air, a clear space was formed over the diffusing surface. Now why was 

 this ? The air molecules were moving towards the diaphragm and passing 

 through it, yet they did not carry any dust particles with them. The reason 

 seems to be this. In the air in front of the diaphragm there are two currents 

 of molecules — one of hydrogen, moving outwards from the diaphragm, and one 

 of air, moving inwards ; but as the hydrogen current is the stronger, it carries 

 the dust particles along with it, and the difference in the strength of these two 

 currents in this case gives rise to a thin clear space over the diffusing surface. 



Let us now apply the same reasoning to the heat effect. When we re- 

 member that hot and cold gases tend to diffuse into each other, the explanation 

 given does not require to be greatly altered. The molecules of air on the surface of 

 the hot body get heated by contact, and these molecules tend to diffuse themselves 

 outwards into the colder molecules. In imagination, let us look at a section of 

 the air close to the hot body. The air there is no longer homogeneous. Some 

 of the particles have more kinetic energy than others. Those molecules with 

 the greatest kinetic energy have the greatest amount of their motion in a direc- 

 tion away from the hot surface, while the cold ones have the greater amount of 

 their motion in a direction towards the hot surface. Now what will happen to 

 any particle of matter hung among these heterogeneous molecules ? The side of 

 the particle next the hot body will be bombarded by a larger proportion of hot 

 molecules than the other side, and the result will be to drive the particle away 

 from the hot body. It maybe objected that, as the air pressure is the same on 

 the front and back of the particles, therefore the total energy of the molecules on 

 the front and on the back must be the same, and therefore there will be no 

 tendency to cause the particles to move. I think, however, this does not cor- 

 rectly represent the case. Near the heating surface the hot molecules are 

 moving outwards and the cold ones inwards. If there were more cold ones 

 moving inwards than hot ones outwards, so that the total energy of the inward 

 moving ones was equal to the total energy of the outward moving ones, which 

 would be necessary in order that the pressures might be equal, then no motion 

 would be produced in the dust particles. We must, however, remember that 

 there are exactly the same number of molecules moving each way. One effect 



