86 Table 40 (continued) 



PROBABLE VALUES OF THE GENERAL PHYSICAL CONSTANTS 



sured jx, but for a used data, mainly by Chappuis. Henning and Jaeger J note this 

 and adopt merely the Henning and Heuse value 273.20 (which as previously 

 noted should be 273.19°). Roebuck obtained three results, 273. 18 , 273. 16 

 and 273. 12 , average, 273. 15 . He lists all previous determinations, and 

 chooses 273.17°, lying midway between his own result and that of Henning and 

 Heuse. He gives ±0.02° as the probable error. Doctor Birge feels that these 

 two results (273.15° and 273.19°) are entitled to far more weight than any of 

 the older work, but that the second result is probably the most accurate, being 

 based on new determinations of a. Hence he adopts — with the probable error 

 given by Henning and Jaeger — 



r = 273.18 ±0.03°K. 



(Roebuck's ±0.02° may well be more reasonable). 



The mechanical equivalent of heat (J) and the electrical equivalent of 

 heat (J'). — A description of the methods for the evaluation of J, and a dis- 

 cussion of the results, is given by Jaeger in the H.P. 2 The value adopted by 

 Henning and Jaeger in the H.P. 3 is one cal. 15 = 4.i84 2 int. joule = 4.1863 abs. 

 joule. The I.C.T. value is one cal.i 5 = 4.185 abs. joule. The cal.i 5 is defined as 

 the amount of thermal energy required to heat one gram of pure water from 

 i 4 .5°toi5.5 C. 



Joule turned mechanical energy directly into thermal energy, and / was 

 evaluated in abs. joules. In most modern work electrical energy is turned 

 directly into thermal, thus evaluating the electrical equivalent of heat (/', mea- 

 sured in int. joules). Since the relation between the int. joule and the abs. 

 joule (io 7 ergs) is known with considerable precision, the mechanical equiva- 

 lent may be obtained from the electrical equivalent. 



The value of / adopted by the H.P. results from the work of Jaeger and 

 Steinwehr. 4 They determined /', for many different mean temperatures lying 

 between 4.75 °C and 49.60° C. This is undoubtedly the most accurate work 

 now available. They list 67 results. These results are represented as a para- 

 bolic function of t. 



On examining their data, Doctor Birge finds that a parabola is not a suffi- 

 ciently complex function. Their residuals show pronounced trends ; unfortu- 

 nately the largest trend is near I5°C. He accordingly made a separate investi- 

 gation of the best curve for their data. 

 /' = 4.21040 - 2.78958 x io~ z t + 7.73723 X IO~ 5 t 2 



-8.52567 x io -7 * 3 + 3.7540 x io- 9 ^ 4 ( I ) 



This gives /' 15 = 4.i8327 int. joules, and is the most probable value resulting 

 from the work of Jaeger and Steinwehr. Jaeger gives two parts in 10000 

 (i.e., 8xio -4 joules) as the probable error. Doctor Birge therefore writes 

 /'i 5 = 4. 1833.^0.0008 int. joules. We have one int. joule = /><? 2 abs. joule, 

 where pq 2 = 1.0004 T ±0.00010. Hence there results 



7 15 = (4.1833 ±0.0008) (1,00041 ±0.00010) =4. 1 850 ±0.0009 a bs. joules. 



'H.P., 2, 496. 2 H.P., 9, 476. 3 H.P., 2, 497- "Ann. Phys., 64, 305, 1921. 

 Smithsonian Tables 





