886 



SCIENCE. 



LN. S. Vol. XV. No. 388. 



supposed that the water was primarily de- 

 composed and that the metallic copper was 

 a secondary product reduced by the nascent 

 hydrogen. He says:* "I have experi- 

 mented on many bodies, with a view to de- 

 termine whether the results were primary 

 or secondaxy. I have been surprised to 

 find how many of them, in ordinary eases, 

 are of the latter class and how frequently 

 water is the only body electrolyzed in in- 

 stances where other substances have been 

 supposed to give way. ' ' Prom our present 

 point of view many of us would rather say 

 that the direct electrolysis of water very 

 rarely occurs, except to a very small extent. 

 In 1839 Daniell began a series of ingen- 

 iously devised and skillfully executed ex- 

 periments with the view of determining, in 

 the case of salt solutions, whether it is the 

 salt or the water which is primarily electro- 

 lyzed. The results appeared in two letters 

 from Daniell to Faraday in 1839 f and 

 1840, J and in a paper by Daniell and W. 

 A. Miller in 1844,§ all published in the 

 Transactions of the Royal Society. The 

 purpose of these investigations was at- 

 tained, and it was completely proved that 

 in reference to their behavior as electrolytes 

 there was no difference between say potas- 

 sium chloride and potassium nitrate, except 

 that in the latter some ammonia was 

 formed at the cathode by the reducing ac- 

 tion of the nascent hydrogen, and it was 

 clearly shown that from an electrolytic 

 point of view all the oxygen acids and their 

 salts fell into line with hydrochloric acid 

 and the chlorides, and the NH^ was electro- 

 lytieally perfectly analogous to K. There 

 is, however, an interest in these papers be- 

 yond this important result. In the earlier 

 part of the woi-k the authors measured the 



* Faraday, ' Experimental Researches in Elec- 

 tricity,' par. 751 (1834). 



t Daniell, Phil. Trans., 1839, p. 97. 



t Op. cit., 1840, p. 209. 



§ Daniell, Phil. Trans., 1839, p. 97. 



amount of electrolysis not only by 'the 

 amount of ions disengaged at eithei" or both 

 electrodes by the primary action of the cur- 

 rent or the secondary action of the elements,' 

 but also tried to obtain a cheek to this way 

 of measuring, by using a diaphragm in the 

 electrolytic cell, and analyzing the contents 

 of the two parts of the cell, the one on the 

 anode side and the other on the cathode side 

 of the diaphragm. Thiseheck was' founded 

 on the hypothesisthatthevoltaiedecomposi- 

 tion of aji electrolyte is not only effected 

 by the disengagement of its anion and ca- 

 tion at their respective electrodes, but by 

 the equivalent transfer of each to the elec- 

 trodes, so that the measure of the quantity 

 of matter translated to either side of the 

 diaphragm might be taken as the measure 

 of the electrolysis.' They soon found 

 that this hypothesis was unfit to give 

 any such measurement, and in the paper of 

 1844 state that their results show that the 

 hypothesis of equivalent transfer of the 

 ions, 'although generally received, is itself 

 destitute of foundation. ' 



The non-equivalent transfer of the ions, 

 incidentally observed by Daniell and Miller, 

 and imperfectly measured by them in a 

 few cases, was made the subject of a long 

 and elaborate series of experiments by 

 Hittorf . The work extended over six years 

 from 1853 to 1859* and is a monument of 

 patient labor and of happy adaptation of 

 means to a clearly jDerceived end. The im- 

 portance of the work was not at first recog- 

 nized by either physicists or chemists; in- 

 deed its meaning was scarcely understood. 

 I shall try to put before you as shortly as 

 I can an outline of the ideas involved in the 

 work, and of the most important conclu- 

 sions arrived at by Hittorf. As the anions 

 and the cations are separated at their re- 



* Hittorf, Pogg., LXXXIX., p. 177 (1853); 

 XCVIII., p. 1. (1856); cm., p. 1 (1858); CVL, 

 pp. 337 and 513 (1859). Arch. N^erland. (II.), 

 VI., p. 671 (1901). 



