74 BELL SYSTEAf TECHNICAL JOURNAL 



The first five are the stellar examples of concordance, therefore support for 

 the two conjoined assumptions. In the last five the discrepancy is deemed 

 to exceed the possibility of doubt, and some explanation ought to be found. 

 How is one to explain a "calorimetric" value less than the "spectroscopic" 

 — as occurs in all five of these cases? The comparison, I recall, is essentially 

 between (I— So) and /. In all of these five cases, I — So turns out to be less 

 than 7; therefore we are tempted to assume a "disorder" in the solid, sub- 

 sisting even unto the absolute zero. The case of hydrogen is the clearest, 

 the disorder being traceable in part to the intermingling of two types of 

 hydrogen molecule known as the "ortho" and the "para" type. In the 

 other cases, the explanations have a vagueness which suggests that they 

 are speculative. This is not a very satisfactory statement to end with. 

 Yet even these discrepancies are so small, that even if there were no closer 

 agreements one would still feel assured that the truth is to be found in 

 some minor alteration of the theory rather in rejecting the theory altogether; 

 and the five excellent agreements, combined with the excellent agreements 

 for so many monatomic gases, convert this feeling of assurance into 

 certitude. 



