138 Mr. G. J. Stoney on the Internal Motions of Gases 



with air. Again, Professor Maxwell has deduced from observa- 

 tions on the viscosity of gases, and also from the rate of diffu- 

 sion of olefiant gas into air, the mean length of the path which 

 a molecule of air describes between successive collisions. It is, 

 at the temperature of 15° C. and the pressure of an atmosphere, 

 about 



7 eighth-metres*, 



using the term eighth-metre as a convenient abbreviation of the 



-TTgth part of a metre f. This is about the ninth part of the 



* The length of the mean excursions of the molecules deduced by Max- 

 well from a determination of the coefficient of viscosity of air which Stokes 

 obtained from Baily's pendulum-experiments is 5 7 Vlllth-metres (Phil. 

 Mag. 1860, vol. xix. p. 32). 



From an experiment of Graham's on the rate of the diffusion of olefiant 

 gas (a gas of nearly the same specific gravity as air) into air, he deduces 

 the value 6"5 Vlllth-metres (Phil. Mag. 1860, vol. xx. p. 31). 



Recent experiments by Maxwell himself on the viscosity of air (Phil. 

 Trans. 1866, p. 258) assign the value 10-6 Vlllth-metres. 



These determinations are very different ; but even greater discrepancies 

 would be of little moment where our object is merely to gain some insight 

 into the order of magnitudes with which we are dealing. 

 The mean of the values obtained by the two methods is 7'3 Vlllth-metres. 

 t I here venture to employ a nomenclature for large multiples and sub- 

 multiples which I have found so convenient that I think I may safely re- 

 commend it. 



In this nomenclature the successive decimal subdivisions of the metre 

 are : — 



the decimetre, 



the centimetre, 



the millimetre, 



the fourth -metre, IVth-metre, or IVth m. 



the fifth-metre, Vth-metre, or Vth m. 



and so on ; 



a fourth-metre meaning a metre divided by 10 l , and so of the others. Thus 

 microscopical quantities can be conveniently measured in fifth-metres, the 

 Vth-metre or metre divided by 10 1 being rather more than the diameter of 

 a corpuscle of human blood ; the wave-lengths of light in Vlllth-metres, 

 as they all lie between 39 and 77 Vlllth-metres; and so on. A fourth- 

 metre is about the thickness of a sheet of medium writing-paper. 



Again, a sixteenth-second of time is to be understood to mean a second 

 divided by 10 10 . This is the period in which the visible vibrations of light 

 are to be measured — the shortest violet double vibration occupying rather 

 more than 13 XVIth- seconds, and the longest red about 25*6 XVIth- 

 seconds. 



So, again, decimals may be conveniently spoken of as fifth-units, ninth- 

 units, &c, or, as I would suggest, for euphony, and since there is so fre- 

 quently occasion to speak of decimals, as fifth-eins, ninth-eins, &c, using 

 the affix -ein as equivalent to unit. Thus the coefficient of the dilatation of 

 platinum is 8'56 Vlth-eins, that of zinc 29'4 Vlth-eins, and those of most 

 other metals between these limits. These measures are equivalent to 



