ELECTROLYSIS 191 



cules, H 2 , 2 , S 2 , S 4 , S 6 , P 4 , N 2 , etc. This appears again to 

 indicate that the atoms of these elements must be capable of 

 presenting opposite charges to each other, in virtue of which 



+ - + - 



they combine together to form molecules, thus HH or C1C1. So 

 that their positive and negative constituents are both capable 

 of changing their attitude according to circumstances. There 

 is, however, no condition known in which hydrogen ions go to 

 the anode or chlorine atoms to the cathode in the process of 

 electrolysis. 



Here we are on difficult ground. Chemists are far from 

 general agreement as to the valency of the elements and the 

 cause of the differences which are observed in so many cases 

 between the principal or ordinary valency, which seems to rule 

 the composition of the most stable compounds, and the extra 

 valency which is frequently developed. By the latter they 

 endeavour to account for the composition of what are still 

 frequently called molecular compounds. 



The measure of the principal valency is the behaviour of the 

 compound in electrolysis. Thus there is no difference of opinion 

 as to the constitution of common salt, which whether in the 



i 



fused state or dissolved in water, yields the ions Na and Cl. But 

 when a saturated solution of common salt in water is cooled 

 prismatic crystals are formed which contain NaC12H 2 0. An 

 immense number of compounds of a similar character are known 

 containing this " water of crystallisation " ; for example : 



Washing soda . . Na 2 C0 3 .10H 2 

 Glauber's salt . . Na 2 S0 4 .10H 2 

 Epsom salt . . MgS0 4 .7H 2 0, etc. 



What is the nature of the bond which holds together the mole- 

 cule of the salt and the molecules of the water superadded no 

 theory yet explains in a manner generally accepted by chemists. 

 If we look through the periodic scheme of the elements many of 

 them are found to exhibit two degrees of valency ; thus carbon 

 appears to be bivalent in carbonic oxide, CO, while it is quadri- 

 valent in marsh gas, CH 4 , and in nearly all other compounds. 

 Nitrogen appears to be bivalent in nitric oxide, NO, while it is 

 trivalent in ammonia, NH 3 , and probably quinquevalent in 

 sal-ammoniac, NH 4 C1, and in nitric acid, HO.N.0 2 , etc. Oxygen 

 is usually bivalent, but it is now agreed that in certains caes. 



