€4 



SCIENCE 



[N. S. Vol. LIV. No. 1386 



Atoins. — In these compounds the electrova- 

 lence of every atom must be negative, for if 

 the eleetrovalenee of any element is zero (inert 

 gases) it can form no compounds. If we let 

 ■y„ represent the numerical value of the nega- 

 tive valence we obtain from Equation 4 



St)c:=St;„. 



(6) 



Since Vn=^s — e, the value of v,i is fixed for 

 any particular atom. For any given group of 

 atoms, we can find %Vc from (6) but we can 

 not find the values of v^ for the individual 

 atoms, in this way. 



If, however, we place v^ = v„ for each atom 

 it is evident that Equation 6 will be satisfied. 

 The residual charge on every atom (being 

 — •Wre + '"c) is then zero. Thus in any group 

 of atoms Postulates 1 and 3 are hath com- 

 ■pletely satisfied if the covalence of each atom 

 is equal to the negative valence of that atom. 

 The negative valence of carbon, nitrogen, oxy- 

 gen and sulfur are 4, 3, 2 and 2 respectively, 

 while that of hydrogen and the halogens is 

 one. If therefore we follow the custom of the 

 organic chemist and write structural formulas 

 using these valences we obtain results in com- 

 plete accord with Postulates 1, 2 and 3. 



Thus these 3 postulates lead us to a rational 

 derivation of the empirical valence rules which 

 constitute the foundation of the science of 

 organic chemistry. Moreover we are brought 

 to see clearly the limitations of this empirical 

 theory. "We now realize that it is only nega- 

 tive valences that should be used in structural 

 formulas (i.e., as covalences) and that even 

 these can only legitimately be used in com- 

 pounds in which electropositive atoms are en- 

 tirely absent, for if some of the atoms have 

 a positive residual charge {v == Ve) then from 

 Eq. 5 it is evident that other atoms must have 

 a negative charge, and for these as well as the 

 electropositive atoms the covalence is not equal 

 to the negative valence. 



From this viewpoint it is incorrect to write 

 structural formulas such as 'Na — CI, 



H— O O 



V. 



,-^\ 



etc., in which the covalence of one atom is 

 taken as equal to the positive valence of that 

 atom. 



It should be kept in mind that Postulate 3 

 does not require that V/. should be equal to v„. 

 There is merely a tendency for these valences 

 to be equal. Among compounds of electro- 

 positive elements we saw that there was a con- 

 flict between the tendencies of Postulates 1 

 and 3 so that v was always different from zero. 

 With compounds formed exclusively of electro- 

 negative atoms, however, there is not neces- 

 sarily a conflict and it is for this reason that 

 we have such a large class of compounds in 

 which V is zero (i.e., Vc = Vn). There may be 

 various causes that make it difficult for v 

 to be zero even for some compounds of electro- 

 negative elements, so that in individual cases 

 V may differ from zero by one or two units. 

 It must be remembered that we deduced the 

 relation v = minimum from Postulate 3 only 

 by assuming the two atoms which share a 

 duplet are of substantially the same size, etc. 

 From Coulomb's law we should expect that 

 either a large charge on the kernel of an atom 

 or a small radius for the kernel should cause 

 electrons in the sheath to be held more firmly 

 and should make it easier for the atom to 

 acquire a negative residual charge. As an ex- 

 ample let us consider the electronegative ele- 

 ments of the first two periods. 



As we pass from carbon, through nitrogen 

 and oxygen, to fluorine, the kernel charge in- 

 creases and the size of the kernel presumably 

 decreases. The residual atomic charge should 

 thus tend to become more negative as we pass 

 towards fluorine and more positive in com- 

 parison as we pass towards carbon. In other 

 words, in compounds of these elements, we 

 should expect a tendency for fluorine to have 

 a covalence a little less than its negative va- 

 lence while for nitrogen the covalence should 

 tend to be greater than the negative valence. 

 Since there are only eight electrons in the 

 sheath of these atoms, the covalence of the 

 carbon atom can never exceed four. All these 

 conclusions are in perfect accord with experi- 

 ence. Thus we find the following covalences:* 



