Atomic Laws of Thermochemistry. 23 



that in the halogens as elements, atom is combined with atom 

 to form the diatomic molecule, in contrast to the state of 

 affairs known for at least the five monatomic metals Na, K, 

 Zn, Cd, and Hg. This difference between the atoms of 

 halogen and of metals suggests that the relation 



(I) = (Br) = (Ol)=0 



does not by symmetry necessitate that for the metals (R) =0. 

 But for hydrogen, in which the molecule is diatomic, we ought 

 by symmetry to expect (H) =0 : hence, as H(HI) is small, say 

 zero, we must have either yfr(I) = or iJr(H)=0, or both are 

 zero. Of these alternatives the preferable one is that which 

 makes ^(1) =0, because of the comparative chemical inertness 

 of iodine; and, accordingly, the heats of formation of the 

 iodides given in Table V. may be taken as approximate values 

 of (R) . For ^Srl 2 and -JBaI 2 the data are lacking, but may 

 be approximately derived from those for the chlorides by 

 subtracting 2*6 x 9 x 1*06 and 2'6 x 8 x 1*06, when we get 

 71*9 and 77*5; but it should be noted that Thomsen states 

 that the absolute values for the heats of formation of all the 

 Ba compounds are uncertain, because he was unable to get 

 pure enough barium to give a reliable initial datum, and 

 preferred to assign a value for the formation of the hydrate 

 by analogy with the data for Mg, Ca, and Sr : on this 

 account the value for -JBaI 2 is uncertain and will be marked 

 with a ?. The next table contains the values of the heats of 

 formation of the iodides arranged in order of magnitude, and 

 in the second row these have been divided by 3*8. 



Table VIII. 

 Approximate Values of (R) for the Metals. 



|Ba. K. ±Sr. Cu. |Ca. Na. Li. 



(R) 77-5? 74-2 71-9 66'5 638 636 60-9 



(R)/3-8 ... 204? 19-4 189 17'4 16'8 167 16-0 



Ag. pig. *A1. *Cd. §Zn. ^Pb. $Hg. £As. 

 (R) 54-8 527 38-2 33-4 33'1 267 15-4 15-4 



(R)/3-8 ... 14-4 13-9 10-0 8-8 87 7'0 4-0 4-0 



Several of the numbers in the second row will be seen to be 

 close to the values 4, 9, and 16, that is 2 2 , 3 2 , and 4 2 , a fact 

 which suggests that for the metals (R) is a quadratic function 

 of the numbers 1, 2, 3, 4, and so on. In the magnesium 

 family we find on passing from JMg to ^Ca an increase of 

 2-9, and from JCa to |Sr 2*6, which is nearly the same ; on 



