496 



Notices respecting New Books. 



course, comparatively small, so that it was necessary to use 

 smaller resistances for R than were used in the previous 

 cases. R was therefore made equal to 3000, 2000, and 1000 

 ohms respectively. 



R. 



r (direct 

 experi- 

 ment). 



r (reverse 

 experi- 

 ment). 



(observed). 



(calculated). 



D 2 



(observed). 



D 2 



(calculated). 



3000 ... 



ohms. 

 92-9 



ohms. 



98-8 



volt. 

 1-389 



volt. 

 1-3875 



volt. 

 1-4515 



volt. 

 1-4496 



2000 ... 



88-3 



957 



1-370 



1-3667 



1-4580 



1-4585 



1000 ... 





951 







1-4825 



1-4837 



For this cell D = 1-4315 at 16° C, B = 2*0434. 

 Mean value for r= 94*16. 



With R = 1000 ohms, it was found that the current was 

 much too large, as the E.M.F. of the cell decreased very 

 rapidly when running the current. But the reverse current 

 gave a fairly good result, as might be expected. For the 

 direct current would cause the solution of Hg 2 S0 4 to become 

 weaker, and the solid present would not immediately saturate 

 the solution. But the reverse current would cause Hg 2 S0 4 

 to be formed in the solution, which would therefore deposit 

 some solid, but would at the same time remain saturated. 

 Thus, all the results from the reverse experiments are better 

 than the corresponding direct results ; but considering the 

 nature of the cell, these latter are fairly good. 



Thus we may say that the Clark cell, besides the others 

 treated here and those similar to them, is reversible. 



These experiments were carried out at the suggestion and 

 under the direction of Dr. Lehfeldt, of the East London 

 Technical College, to whom I gladly acknowledge my debt. 

 Physical Laboratory, 

 East London Technical College. 

 October 1899. 



XL VII. Notices respecting New Books. 



Groupements Cristallins. Par Fred. Waxleha^t. Paris : Georges 

 Carre & C. Naud, 1899. Pp. 1-81. 



WHEN crystallization is allowed to take place in a solution which 

 is not subject to any disturbance, we have the formation of 

 well-defined and complete crystals. If, on the other hand, the liquid 

 is violently stirred, we get, instead of large crystals, masses of small 



