244 K. IKEDA 



for the recalculation are very imperfect ; and as hunting for these 

 in the various periodicals consumes more time than can well be 

 spared, and moreover as the determination of the atomic capillary 

 constant of an element requires the value of K to be known for 

 many compounds of the same type containing the element under 

 consideration, only about one hundred determinations have been 

 submitted to recalculation. Still it is hoped that these will suffice for 

 establishing the approximate validity of the new formulae. 



These one hundred and odd substances consist of compounds of 

 carbon, hydrogen, oxygen, nitrogen, chlorine, bromine, iodine and 

 sulphur. The atomic capillary constants have been calculated from the 

 data, and from these KV* & /T^ï 7118 , as well as A', have been recalcu- 

 lated, with what agreement the following; tables will show. 



The atomic constants for the various elements (a, b, c, &c. in 

 the formula) used in the calculation of the first table are as follows : 



for carbon* 600 in all cases 



„ hydrogen 75 „ „ „ 



,, oxygen 170 in alcohols. 



,, ,, 450 when it is entirely combined with one or more 



carbon atoms. 



,, chlorine 1100 when only one chlorine atom is united with a 

 carbon atom. 



,, ,, 925 when more than two chlorine atoms are united 



with a carbon atom. 



,, bromine 1500 



„ iodine 2150 



„ nitrogen 650 in amines, isocyanates, and nitro bodie s. 



,, ,, 450 in nitriles and cyanates, 



„ sulphur 1500 



* The value fur carbon is most probably higher than this in unsaturated compounds. 



