[king] DETERMINATION OF AVOGADRO'S NUMBER 63 



The above determination of n compares favourably with Ruther- 

 ford's 1 2-78, Planck's 2 2-77, and Millikan's 3 (2 -70S -00 ±5), while 

 the value recently obtained by Fowle 4 from a somewhat different 

 treatment of the Mount Wilson data gave 2-56. 



Taking as the volume of a molecule-gramme the value 22412 

 at 0°C. and 760 mm., Avogadro's Number is given by the relation 5 

 N = 22412 X n , from which we obtain N = (6-23 3 ± -02 8 ) XlO 23 . 

 Taking the atomic weight of silver to be 107 • 88 and its electrochemical 

 equivalent to be 1-11800 mgm. per coulomb, we have Ne = (107-88) 

 -f- (1-11800) X 10 2 , where the electronic charge is measured in e.m. 

 units; we thus easily obtain the final result e = (4-64 4 ± -020 X 

 10" 10 e.s. units. 



Although the above reductions of a series of self-contained ob- 

 servations on atmospheric extinction yield a determination of n Q 

 to an order of accuracy not very much less than that of the best ex- 

 isting determinations, their chief interest lies in the fact that they 

 constitute as rigorous an experimental test of Rayleigh's Law as may 

 be expected in view of the practical impossibility of securing absolutely 

 perfect atmospheric conditions. From the value of Y may be cal- 

 culated the fraction of radiant energy converted per centimetre 

 of path into thermal molecular agitation ; taking a value 7 = -032 

 for air under standard conditions it is easily shown that in a stream 

 of radiation corresponding to the solar constant the rate of increase 

 of temperature amounts to -015°C. per hour. 6 As the above value 

 of 7 , even for the comparatively dust-free air above Mount Wilson, 

 includes to a certain extent the effect of volcanic haze, it follows that 

 in a pure gas partition of energy cannot take place at a rate greater 

 than is represented by the above-mentioned rate of increase of tempera- 

 ture. We have in this case an excellent illustration of two interpene- 

 trating dynamical systems (the aetherial system of electromagnetic 

 waves and the molecular gaseous system) allowing of partition of energy, 

 if at all, at an excessively slow rate compared to the rate of equilization 

 of energy distributions which is capable of being realized in each system 

 considered separately. It is interesting to notice also that this rate 

 is enormously increased by the presence of constrained molecular 



1 Rutherford, E., and Geiger, H., Roy. Soc. Proa, A, Vol. 81, 1908, p. 171. 



2 Planck, loc. cit., p. 172. 



3 Millikan, Phys. Rev. 2, Ser. 2, pp. 109-143, Aug. 1913. Phys. Zeitschr. 14, 

 pp. 796-812, Sept. 1, 1913. 



4 Fowle, Astrophysical Journal, 38, No. 4, p. 398, Nov. 1913. 



6 The values of these fundamental constants are those recently given by Kolow- 

 rat, "Le Radium," II, i., p. 1, Jan. 1914. 

 6 King, loc. cit. p. 394 



