Connexion between the Atomic Weights. Ill 



Judged from the percentage standpoint, the elements 

 deviating most from the formula are : — 



W - W 100. 

 to 



Calcium — 5*8 



Cobalt -4*7 



Carbon -4'0 



Sulphur —3*8 



Fluorine -j-3'7 



Vanadium +3*6 



Potassium ,. -f3'3 



Boron- -f 3'3 



Iodine 4-3*1 



Of these, calcium, carbon, fluorine, sulphur, boron, and 

 iodine are exceptions to Dulong and Petit* s law. 



Argon is an exception to the periodic law, which would 

 lead one to expect its atomic weight to be between those of 

 chlorine and potassium. If, however, its atomic weight 

 should be proved to be greater than that of calcium, the atomic 

 weight of the latter would be given correctly by the rule. 

 The value to be then expected for argon would be 42*1. 



Extension of the Formula to all the Elements. 



The evidence which has been adduced in favour of the 

 formula 



W=(n + 2)j. 21 



warrants an attempt being made to apply it to the remaining 

 elements. 



The plan adopted has been to replace the (n-f 2) of the 

 above formula by N, where N is the order in what will be 

 called for short the augmented list. This augmented list is 

 one in which, though the real order of arrangement is never 

 reversed, a value for N is taken so that the particular element 

 shall have its atomic weight given as nearly as possible by 

 the formula. Thus the augmented list is a list of the elements 

 in ascending order of magnitude, with gaps left in it so as 

 to make the experimental results fit the formula. 



The justice of this process may be tested by seeing whether 

 the necessary gaps are too numerous, or are in any other way 

 improbable ; or whether the gaps can be reconciled with 

 acknowledged chemical principles. 



Over the range from lithium to samarium the change is 

 .-imply to replace n-\-'2 by N. That is, instead of regarding 



