the Elementary Law of Hydrod [fusion. 491 



which differ from the observed values by only 

 -0-7, —5-2, + 2*0. -4-5. 



The difference between observation and calculation therefore 

 does not in any case reach 1 per cent, of the quantity measured. 



Two adjacent solutions of sulphate of zinc, with the concen- 

 trations : 2 and z\ generate therefore, between two zinc elec- 

 trodes immersed in them, an electromotive force the quantity 

 E of which can be represented bv the expression 



E = A(.- 1 ,-.- 1 ):i + B(.- 2 + c 1 >] . . . . (1) 

 as soon as z 2 and z^ come within the interval from c = 0'15 to 

 z=0"35. 



By aid of the relation (1) it becomes possible in certain 

 cases, from the measured electromotive force which makes its 

 appearance between two zinc electrodes immersed in solutions 

 of zinc sulphate of different degrees of concentration, to make 

 a sure induction respecting the amount of difference of con- 

 centration of the solutions. 



On this is founded the principle of the method of measure- 

 ment which I have applied to the investigation of the elemen- 

 tary law of hydrodiffusion. 



By the reduction of the measurement of the concentration 

 to the measurement of an electromotive force advantages are 

 gained which are offered by none of the methods hitherto used 

 for the investigation of diffusion: the accuracy of the mea- 

 surement can be carried to almost any degree desired: the 

 concentrations measured are not. as in previous methods, mean 

 concentrations of thin layers of liquid, but are those which 

 occur in certain surfaces, namely the boundary surfaces in 

 contact with the electrodes ; lastly, the method employed per- 

 mits (and this might be an important advantage) the diffusion- 

 vessel to be carried to any distance we please from the mea- 

 suring-place — for instance a place free from agitation, and 

 which offers the most constant temperature possible. 



Two essentially different forms of investigation were made 

 use of: one of them requires, like all former methods, a pro- 

 portionally long time of observation, a space of several days, 

 and, in consequence of this, prepares for the exact experi- 

 mental investigation of the process of diffusion sundry diffi- 

 culties and casualties : the other permits the investigation to 

 be brought to a conclusion in a few hours, and thereby affords 

 the advantage that one is able to decide in a simple way, and 

 in a very brief time, some important questions, such as, for 

 example, the influence of the concentration upon the a con- 

 stant of diffusion." or the influence of temperature upon the 

 progress of diffusion. 



2L2 



