234 Messrs. Lodge and Clark on Dusty Air in the 



temperature of the successive layers is rapidly decreasing, and 

 thus they have less tendency to rise although there is now 

 very little to prevent them, until at a distance from the solid 

 the air is at rest- 

 There is, then, in all convection-currents from a vertical 

 surface a layer whose velocity is a maximum ; on either side 

 of it the velocity is diminishing — on the cool side because the 

 temperature gets less and less; on the hot side, in spite of the 

 high temperature, because of the viscosity. The maximum- 

 velocity layer marks the position of compromise between the 

 lightening due to temperature and the resistance due to the 

 neighbourhood of the solid. 



The fact that the convective velocity increases and reaches 

 a maximum as you recede from the solid, combined with the 

 fact that the temperature constantly diminishes as you go in 

 the same direction, seems to us to have an important bearing 

 on the production of the dust-free space by molecular bom- 

 bardment in a way we will endeavour to explain. 



A mere steady fall or gradient of temperature as one recedes 

 from a body will not of itself produce the differential bombard- 

 ment necessary to keep back the dust-particles. For though 

 it is true that the air is hotter on the side next the body than 

 it is on the far side, this excess of temperature does not result 

 in an excess of pressure but in a defect of density. The air 

 near a warm body is less dense than the air outside, but it is 

 not at a greater pressure ; at least, if it is, it is a new pheno- 

 menon, and one not to be expected from any known action of 

 a solid upon gas, except, of course, within the extremely 

 minute range of the molecular forces estimated by Quincke 

 at the five-millionth of a millimetre. The warm gas there- 

 fore would produce no greater bombardment on the one side 

 of the particles than the cold gas produces on the other ; 

 for though each molecule hits harder and hits oftener, there 

 are fewer molecules to hit ; and the one effect compensates 

 the other as soon as a stationary condition as to distribution 

 of temperature has been attained. Not before ; for so long 

 as the air near a solid is getting warmer, there is a more rapid 

 diffusion of molecules outward than there is inward, and till 

 equilibrium is reached there will certainly be a pressure out- 

 wards from the body sufficient to drive the dust back. 



If the air near a solid can be kept stagnant while the body 

 is warmed, the dust in it ought to be driven back at first by 

 the outward heat-diffusion; but as soon as a stationary tem- 

 perature-condition has been attained, this bombardment ought 

 to cease, and permit the dust to slowly settle back again if 

 it likes. This is much the case of the interior of a glass tube, 



