( 248 ) 



He also remarks that this formula agrees in character with my 

 theoretical one, viz. 



' A =c g I 



M ' J 1 + c' ' T 



because 1 -(- c' slowly increases with decreasing T. 



This, indeed is the case. For instance when we assume q = 3.70X10 18 



or = 10, and consider the spherical function of order 2 we find 



after Schuster (Phil. Tr. Vol. 180 p. 496): 



1 + c' 

 rf 1 + c' 



The exponent accordingly changes slowly, and reaches as a maxi- 

 mum 0.126, a value, which approaches the value 0.2 required by 

 the empirical formula. 



1 think this points to a possibility to bring agreement between the 

 theoretical assumptions and the observed tacts for the daily variation 

 of earth-current and magnetic component. However this is not the 

 ease for the short oscillations. 



I regret that mj near return to Europe prevents me from entering 

 now on those questions, or making new experiments. 



Batavia, Aug. 1908. 



Chemistry. - - "The nitration of toluene", by Prof. A. F. Holleman. 



(Communicated in the meeting, of September 26 1908). 



On account of its great technical importance the nitration of toluene 

 has heen studied repeatedly; the determination of the quantity of 

 0- and p-nitrotoluene contained in the product of the reaction has 

 also been carried out a few times. Raoul Pictet, (C.r. 116, 815) 

 states that when toluene is nitrated at - 55° 5.5 times as much 

 p-nitrotoluene is formed as when the nitration is carried out at 0°. 

 HoldbrmatsN, (1>. 39,1250) tried to modify the proportions in which 



