Chase.] •^t)U [April 21, 



this an isolated fact, little importance could be attached to it, but when 

 we bring it to the test of mathematical probability, it becomes suggestive 

 of relations which may, perhaps, lead to important discoveries. In the 

 following comparison D indicates the estimate of atomic heat which 

 deviates least from 6 {Meyer, Modernen Theor. d. Chem. Ed. of 1880, 90, 

 106) ; ^1, the ratio of deviation from exact correspondence with the theo- 

 retical value ; C, the observed multiples and values of the perissad and 

 artiad divisors (Notes 201 — 3) ; d^, their deviation from the theoretical 

 values : 



