OF ARTS AND SCIENCES. 41 



IV. 



CONTRIBUTIONS FROM THE PHYSICAL LABORATORY OF 

 THE MASSACHUSETTS INSTITUTE OF TECHNOLOGY. 



XL ON THE EFFECT OF TEMPERATURE ON THE VISCOSITY 



OF AIR. 



By Silas W. Holman. 



Read, June 14, 1876. 



The developments of the "kinetic theory" of gases made within the 

 last ten years have enabled it to account satisfactorily for many of the 

 laws of gases. The mathematical deductions of Clausius, Maxwell 

 and others, based upon the hypothesis of a gas composed of molecules 

 acting upon each other at impact like perfectly elastic spheres, have 

 furnished expressions for the laws of its elasticity, viscosity, conduc- 

 tivity for heat, ditfusive power and other properties. For some of 

 these laws we have experimental data of value in testing the validity of 

 these deductions and assumptions. Next to the elasticity, perhaps the 

 phenomena of the viscosity of gases are best adapted to investigation. 



According to the kinetic theory, the molecules of the gas are con- 

 stantly in rectilinear motion. In virtue of their mass and velocity, 

 these molecules have a certain momentum. Hence, if we have two 

 layers of air moving over each other, we shall have a mutual inter- 

 change of momentum from the transference of molecules from one 

 layer to another, the result being a tendency toward an equalization 

 of the velocities of the two layers. This produces the effect of friction 

 between the two layers, and its amount determines the viscosity of the 

 gas in any particular case. From analytical considerations Maxwell 

 has deduced * an expression which, as corrected by Clausius,t should 



read, 



_ Mu 



where j/ is the coefficient of viscosity of any gas ; Mis the mass of a mole- 

 cule ; u the "velocity of mean square " of the molecules ; and s thedis- 



* Pliil. Mag. xix., xx.; 1860. t Phil. Mag. xix., 434. 



