188 CHEMICAL DISCOVERY AND INVENTION 



one atom of chlorine to another, but in its turn also replaces an 

 atom of copper, forming chloride of copper and sulphuric acid," 

 and so forth. It is not necessary to quote further, but this idea, 

 which we owe to Williamson, of atomic motion and exchange of 

 partners in a liquid or gas, is now acknowledged to lie at the 

 root of all modern ideas of chemical action. 



The same idea was made use of in 1857 by the German physicist 

 Clausius, in order to explain electrolysis, but he could not answer 

 the question as to the proportion of the dissolved substance 

 which must be supposed to be in the act of exchanging con- 

 stituents, and therefore dissociated into ions. To Arrhenius we 

 owe the knowledge that from the conductivity and the osmotic 

 pressure of the solution of an electrolyte the relative proportions 

 of the undissociated and dissociated molecules can be calculated. 

 It is well known that water is an extraordinarily bad electrolyte. 

 Experiments originally made by Davy a hundred years ago, and 

 repeated by numerous experimenters down to our own day, have 

 shown that in proportion as the dissolved impurities in ordinary 

 water are removed or excluded the conducting power, small at 

 any time, is steadily reduced, and at one time it was supposed 

 that pure water was a non- electrolyte and non-conductor. What 

 is called " conductivity " water, that is water specially distilled 

 under special precautions, does possess a very small conductivity 

 which is attributed to the presence of minute quantities of 

 hydrogen ions and hydroxyl, HO, ions. It will therefore behave 

 as an extremely feeble acid or base. 



Hydrogen chloride or hydrochloric acid, in the liquid state 

 but absolutely free from water, is a colourless liquid, but it- 

 exhibits none of the characteristic properties of water as a 

 solvent. It is a non-conductor. It has none of the properties of 

 an acid, for it does not alter the colour of litmus, nor does it act 

 on alkaline oxides or carbonates. If, however, this liquid is 

 mixed with water the solution is one of the best electrolytes 

 known, and with the development of conductivity the chemical 

 activity commonly attributed to hydrochloric acid at once 

 becomes manifest. It is obvious, therefore, that some change 

 has taken place in the constitution of both water and hydrogen 

 chloride when in presence of each other. 



It appears probable that the ions available in the liquid in a 

 moderately strong solution are chiefly those of hydrochloric acid, 

 inasmuch as the products actually evolved are hydrogen and 



