o8 Table 40 (continued) 



PROBABLE VALUES OF THE GENERAL PHYSICAL CONSTANTS 



the last being Coblentz' own value. Laclenburg 1 gives 1.432 ±0.006. The 

 separate results which he used are 1.425 to 1.441, 1.4295 ±0.007, I -435» I -43 I 8, 

 1.430. The chief error arises from the various corrections applied to the 

 observed values. Coblentz' original value 2 of 1.4456 has become 1.43 18 in his 

 latest article, 3 Doctor Birge believes 0.003 i s a much more reasonable estimate 

 of error. Both Coblentz and Ladenburg agree on the absolute value. Hence 

 he adopts 



c 2 — 1 .432 ± 0.003 cm ' deg. 



h= (6.548 ±0.01 5) x icr 27 erg • sec. 



The radiation constant c 2 occurs in the Boltzmann factor e~ e/kT , (c = energy, 

 T= absolute temperature) in the form e' C2V/T =e~ C2/XT , where v in cm -1 , or A in 

 cm, is the quantum equivalent of e ergs. 



(f) The Stefan-Boltzmann lazv and the Planck equation. — The second 

 method for determining h by the radiation constants is through the Stefan- 

 Boltzmann law, E = aT i = acT i /4. h is connected with a, using Planck's law, 

 by the relation 



h={2^k i /lSC 2 a)K 



As in the case of c 2 , there is a difference of opinion concerning the accuracy 

 with which a may be measured. The best value, in 19 19, was that obtained by 

 Coblentz, 3 namely (5.722 ±0.012) X I0~ 5 erg • cm -2 • deg. -4 • sec. -1 . In his more 

 recent discussion, Coblentz 4 gives all available data, and concludes that the 

 most probable value lies between 5.72 and 5.73. 



Since this 1922 article by Coblentz, there have been two new determinations 

 of a, one by Hoffman (method of Westphal), giving {7=5.764 ±0.052, and 

 the other by Kussman, 6 using the modified Angstrom pyrheliometer. This latter 

 method was used also by Coblentz 3 giving 5.722 as stated, by Gerlach 7 giving 

 5.80, and by Kahanowicz 8 giving 5.69 to 5.73 as corrected by Coblentz. 4 Kuss- 

 man obtained (7 = 5.795 ±one per cent. Ladenburg 1 quotes the four results by 

 Gerlach, Hoffman, Coblentz, and Kussman. He adopts the unweighted mean. 

 He agrees with Gerlach that Coblentz' true error is more nearly 0.06 than 0.012. 

 The experimental results of Kussman 6 and Coblentz 3 are in almost perfect 

 agreement. The discrepancy in their results is due to the correction for the 

 lack of complete absorption of the receiver. Michel and Kussman 9 claim to 

 prove that the correction Coblentz applied is too small. The values of a by 

 Kussman and by Hoffman, as well as Gerlach's earlier value of 5.80, corre- 

 spond to impossibly low values of A. Coblentz' result gives an h in good agree- 

 ment with that obtained by more accurate methods. This tends to indicate the 

 correctness of Coblentz' correction for incomplete absorption, as opposed to 

 Kussman's. 



1 H.P., 23, 303. 2 Bur. Standards Bull., 10, 1, 1914. 3 Proc. Nat. Acad. Sci., 3, 504, 

 1917. 4 Bur. Standards Bull., 17, 7, 1912. 5 Zeit. Phys., 14, 301, 1923. 6 Ibid., 25, 58, 

 1924. 7 Ann. Phys., 50, 259, 1916; Zeit. Phys., 2, 76, 1920. 8 Nuovo Cimento, 13, 142, 1917. 

 'Zeit. Phys., 18, 263, 1923. 

 Smithsonian Tables 



