1902. J on the Ions of Electrolysis. 37 



the electrodes, which were preliminarily inserted into the upper parts 

 of the two limbs of the tube, with a battery with high difference of 

 potential a current will pass, and a transference of ions will take place, 

 cations (K) towards the cathode and anions (NO3 and MnO^) towards 

 the anode, and the column of pink colour will rise in the limb contain- 

 ing the anode and fall by an equal amount in the other. By this 

 means an approximation can be made to the rate of travel of the 

 ions. 



We now come to a new chapter beginning with 1887 ; but before 

 entering on it we must turn aside for a little to a subject which does 

 not at first sight seem to have a very close relation with the matter 

 we have in hand. The subject is that of what may be called the 

 osmotic phenomena. These are all connected with the concentration 

 or with the dilution of solutions. They all involve the idea of the 

 work done in concentrating a solution. We need not discuss the 

 theory of these phenomena, we are interested in them now only as 

 they give us methods of ascertaining the molecular concentration of 

 a solution. In 1883 * Eaoult showed that in the case of a great 

 many substances, equimolecular solutions (with the same solvent) 

 have the same freezing-point. In 1886 f he showed that equi- 

 molecular solutions with the same volatile solvent have the 

 same boiling-point. Molecular is here used in its ordinary chemical 

 sense. These discoveries were eagerly taken up by chemists as pro- 

 mising an important addition to the means at their disposal for deter- 

 mining the molecular weights of substances. Convenient arrange- 

 ments for applying the methods were devised by Beckmann, { and 

 soon came into use in nearly every laboratory. They were almost 

 exclusively used for the determination of the molecular weight of 

 organic substances, and have been found trustworthy in such cases. 

 When, however, van't Hoff § in his study of the theory of solutions 

 concluded from theoretical considerations that the depression of the 

 freezing-point and the rise of the , boiling-point are proportional to 

 osmotic pressure in the case of dilute solutions, the observations made 

 by Baoult and others furnished a number of facts ready for testing the 

 theory. He found that, while in many cases the osmotic pressure cal- 

 culated from his formula PV = RT agreed, within the limits of experi- 

 mental error, with the value calculated from the observation, there were 

 a very considerable number where the observed value differed from that 

 given by the formula. He accordingly modified the formula by the 

 introduction of a factor i, so as to make it PV = iRT. This factor i 



* Raoult, Compt. rend. xciv. p. 1517; xcv. pp. 187 and 1030 (1882); xcvi. 

 p. 1653 ; xcvli. p. 941 (1883). 



+ Raoult, Compt. rend. ciii. p. 1125 (1886); civ. pp. 976 and 1430; cv.p.857 

 (1887). Zeitschrlft f. physik. Ohemie, ii. p. 353 (1888). 



X Beckmann, Zeitschrift f. physik. (3hemie, ii.'pp. 638 and 715 (1888) ; iv. p. 532 

 (1889) ; viii. p. 223 (1891). 



§ Van't Hoff, Zeitschrift f. physik. Chemie, i. pp. 500-508 (1887). 



