I9I4.] ELECTRON CONCEPTION OF VALENCE. 27 



ing" upon the loss or gain of corpuscles, the knowledge of the elec- 

 trical state of an atom in a compound is of importance. Ionization 

 in solution is an invaluable aid in determining the distribution of 

 these charges. Dissolving a substance does not produce electric 

 charges on atoms but only makes these charges manifest to certain 

 experimental methods. These are, in fact, the only direct experi- 

 mental methods for determining valence. For substances which 

 ordinarily do not ionize, a knowledge of the distribution of the 

 charges is also important. This question was taken up for organic 

 compounds and reactions by Professor J. M. Nelson, of Columbia 

 University and the speaker in a number of papers.^ It was shown 

 that satisfactory classifications could be developed with the electron 

 conception of valence alone, but that the use of both polar and non- 

 polar valences leads to contradictions in reactions which are fun- 

 damentally similar. 



With compounds which do not ionize, the Periodic System serves 

 in a general w^ay as a guide for developing valence structures in 

 which relative positive and negative properties of the different atoms 

 are involved. When two atoms are united by a single bond (one 

 unit of valence), where one corpuscle is transferred in the produc- 

 tion of the bond, there is ordinarily no question of the direction of 

 transference of this corpuscle, that is to say, which element is posi- 

 tive and which is negative. If isomers exist, the difiference in the 

 relative affinity of the atoms for the negative corpuscle may be small 

 (as in iodine monochloride) and the less stable modification may 

 possess the structure in which the corpuscle is transferred in the 

 opposite direction from that of the stable modification. 



Some interesting questions are raised when the double bond is 

 considered from the electronic point of view. As used in the past, the 

 justification for the double bond lies in the desire to maintain con- 

 sistently, constant values for the valence of certain atoms. Practical 

 work during the past fifty years has borne out within certain limits 

 the usefulness of this conception. Before discussing the significance 

 of the double bond with the newer ideas of valence, some general 



^ Jour. Amer. Client. Soc, 32, 1167 (1910) ; 33, 440 (1911) ; 55, 1810 

 (1913) ; 36, 2og (1914). 



