﻿250 



Mr. J. H. T. Roberts on a Method of 



We may perhaps notice that the screening effect of the 

 slab (b, fij is independent of its position between P and the 



l'lg. 



o. 



p 



$ / b N \ yW J^ 



w 



h ^ 



active slab, as far as the primary rays are concerned. 

 Formula (14) is of considerable generality, and includes all 

 the results already given. 



XXI. On a Method of Investigating the Transpiration of 

 Gases through Tubes. By Joseph H. T. Roberts, M.Sc, 

 University of Liverpool* . 

 TJ^HE study of the transpiration of a gas through tubes 

 JL affords a means of determining the viscosity, and thus 

 of investigating certain of the molecular properties of the 

 gas. The earliest systematic observations on gaseous tran- 

 spiration were those of Graham f; but since the discovery in 

 1860 J of Maxwell's law of the independence of viscosity 

 and pressure, in virtue of which the viscosity of a gas is 

 deducible from its transpiration coefficient, numerous experi- 

 menters have adopted this method for the determination of 

 gaseous viscosities. The most expeditious way to find the 

 viscosities of a number of gases is to determine the coefficient 

 absolutely for one of them, and to deduce the coefficients for 

 the others from a comparison of their transpiration times under 

 the same circumstances. 



With a view to the fulfilment of the latter condition some 

 observers have used a maintained vacuum for the removal of 

 the transpired gases ; this, however, is not only unnecessary, 

 but undesirable, as the transpiratory behaviour then becomes 

 complicated with the effusive, and only with very long tubes 

 can the one be considered to be separate from the other. 

 Another objection to the use of the vacuum is that, if the 

 higher pressure be comparable with that of the atmosphere, the 

 transpiration tube requires to be very long to ensure normal 

 flow. Indeed, Graham in some cases used capillaries 22 feet in 

 length (made by fusing together shorter pieces), his results only 

 becoming concordant for tubes above a certain relative length. 



* Communicated bv Prof. L. 



t Phil. Trans. 1846, 1849. 



R. 



Wilberforce. 

 \ Phil. Mag. 



1860. 



