40 Dr. H. J. S. Sand on the Measurement of the 



by clashes. The values for k derived from this equation are 

 no doubt often vitiated by the fact that hydrogen may come 

 off before complete reduction of the nitrobenzene has taken 

 place, so that the values found for Jc are too small, and those 



n n 



for — ^ — - too large. This does not matter seriously, however, 



as we only desire to find an upper limit for the latter quantity. 

 Eliminating F and partly k from the equations (2), (3), (4), 

 (5), we find the following relations : — 



and 



C^ _ 96540 kQ w 



As an example, by which to test whether concentration 

 changes played an important part in Haber's results, we take 

 numbers from the paper recently published by him in con- 

 junction with Russ*. According to Table II., p. 279, when 



, £ i ., 4*5 milamps. , ^ , 



a current ot densitv - - — — - J_ was passed through 



' 1-5 X 10 cm. 2 * & 



a vigorously stirred alkaline solution of nitrobenzene of the 



concentration 0*8 qillV ' , an absolute potential of 0*0737 



c.c. 



volts was observed ; whereas when the current had a density 



£ 36 milamps. n . .. . _ _,.,_., .. .. 



ot T~k — T7v $ — < the potential was 0*1171 volts, the 



1*5 x 10 cm. 2 ' L 



difference being 0*0434 volt. Now, according to the Table 



Vila, on p. 285, when a current of the density ^Tk — tt\ 

 mi limns 



— s— * was employed in a solution of the concentration 



cm/ r J 



0*8 - g '" eqmV -, the potential 0*0993 volt was recorded. In 

 c.c. 



order to obtain a rise of potential of 0*0434 volt by reduction 

 of the concentration of nitrobenzene, the current strength 

 being left unchanged, we can infer by interpolation from 

 numbers given in the same table that the concentration must 



be reduced to 0*115 — — — ', i. e, to 14 per cent, of its 



c.c. x 



* Loc. cit. 



