THE RANGE OF NATURE's OPERATIONS. 209 



ward in an almost^ straight line till it gets close to another molecule. 

 Then an encounter takes place; the molecules struggle together for an 

 excessively brief period, after which they fling asunder in two new 

 directions. The average velocity with which the molecules dart about 

 had been known before Maxwell's investigation. It is about 500 meters 

 per second in the air which we breathe. It was also known that, except 

 In very high vacua, the molecules are so crowded that their journeys 

 between their encounters can be but short, but the length of these jour- 

 nej^s was not known. What Professor Maxwell efl'ected was an actual 

 determination in certain gases of the average length of these "free 

 paths.-' He did this by showing that upon this average depends what 

 is called viscosity in a gas — that property which gradually brings a gas 

 to rest after it has been disturbed and currents established in it. He 

 further showed that the average length of the free paths is what deter- 

 mines the rate at which gases diffuse into one another. Accordingly, 

 from experiments on viscosity made b}^ Sir George Stokes, and from 

 Graham's experiment on diffusion, he was able to ascertain what the 

 average length of the free paths must be to produce the observed 

 amount of effect. He thus found it to be about six eighthets ^ of a 

 meter — that which would be represented arithmetically by 0.00000006 

 of a meter — in atmospheric air at the temperature and pressure of 

 the experiments, which we may take to have been a barometric pres- 

 sure of 7B0 millimeters of mercury and a temperature of about 17"^ 

 Centigrade. This length is smaller than any interval which the micro- 

 scope can show, and 3^et it is a length which must be regarded as very 

 large among molecular magnitudes. 



nature's w^ork at closer quarters. 



We can, however, extract from Maxwell's determination informa- 

 tion about still smaller quantities. In fact, Clausius had previousl}' 

 been able to show ^ that in the more perfect gases, at ordinary temper- 

 atures and pressures, the mean lerjgth of the free path is about sixt}" 

 times what the average spacing of the molecules is at any one instant 



^ The gravitation of the molecules toward the earth must bend the free paths, but 

 the curvature is insensible until, near the boundary of the atmosphere, the attenua- 

 tion of the air far exceeds any that can be reached in artificial vacua. This bending 

 of the free paths keeps the atuKJsphere that accompanies the earth from extending 

 outward beyond a short distance. It moreover makes the denser constituents of an 

 atmosphere come to an end sooner than the lighter constituents, so that in the upper 

 regions of an atmosphere the law of the ecpial diffusion of gases no longer holds. See 

 "On the physical constitution of the sun and stars," Royal Society's Proceedings, 

 No. 105, 1868, p{). 13 and 14; or "Of atmospheres upon planets and satellites," Royal 

 Dublin Society's Scientific Transactions, Vol. VI, 1897, p. 305, or Astrophysical Jour- 

 nal, Vol. VIII, 1898, p. 25. 



^Subsequent experiments by ^laxwell himself on the viscosity of air (Phil. Trans. 

 1866, p. 258) assign a length of 10.6 eighth-metrets to the average free path. The 

 mean of all the deterininations is 7.6 eighth-metrets. 



^Pogg. Ann. 1858, Vol. Ill, p. 251; or Phil. Mag. 1859, Vol. XVII, p. 89. 

 SM 99 U 



