574 JAMES CLERK MAXWELL. 



The stress thus arising from variation in rate of change of 

 temperature varies inversely as the pressure of the gas, and 

 is therefore most conspicuous in high vacua. If two small 

 bodies are warmer than the air, the line joining them will 

 be a line of maximum pressure, and they will repel each 

 other, while they will attract one another if they are colder 

 than the air. If, however, a ring be placed so as to have 

 the line joining the bodies for its axis and be sufficiently 

 heated the repulsion may be changed into attraction. In the 

 case of a cup, as noticed by Stokes, the variation of the rate 

 of change of temperature is much greater on the convex side 

 than on the concave, where it is nearly uniform, like the elec- 

 tric potential within a hollow vessel, and hence the normal 

 pressure is greater on the convex surface than on the concave, 

 which will account for the motion of the cup radiometer if 

 tangential stresses are neglected. But when the tangential 

 stresses on any portion of gas are considered, it appears, 

 that they, with the normal forces, form a system which is in 

 equilibrium, so that inequality of temperature has no tend- 

 ency of itself (i.e. without the action of gravity, etc.), to 

 produce currents in the gas. Maxwell therefore concludes 

 that the above explanation is insufficient, and that the true 

 cause of the motion is to be found in the character of the 

 tangential action between the solid and the gas, allowing 

 the gas to slide over the surface of the solid, and thus 

 diminishing the tangential stresses without affecting the 

 normal stresses. In the appendix, dated May 1879, Max- 

 well determined the character of the tangential action on 

 certain hypotheses respecting the nature of the surface of 

 the solid, and the character of the collisions, and concluded 

 that the gas may slide over the surface of the solid with a 

 finite velocity, and that inequalities of temperature at the 

 surface " give rise to a force tending to make the gas slide 

 along the surface from colder to hotter places." 



Most of the more elementary theorems respecting the 

 kinetic theory of gases are given in a very concise form by 

 Maxwell in his Theory of Heat, the more recent editions 

 of which also give an account of Professor Willard Gibbs' 



