488 Mr. W. Sutherland on Molecular Force 



the type MgCl 2 in place of our former 7*25/19 we should 

 take 8*5/19 or -45, and for the type K 2 S0 4 4*4/19 or *23 ; 

 thus for the type EaCl we have the equation 



(M// 2 )* = « 2 *(M 2 /p 2 )7l3=-08V(M 2 / ft )?, . (16) 



and for the types MgCl 2 and K 2 S0 4 the equation 



(M/y»=<(M,/p 8 )Vl9=-05^(M 2 / ft )^. . (17) 



Now for the three types we ought to have the same 

 equation, for the fundamental equation for giving a value of 

 (M 2 2 / 2 )^ from a value of a 2 measured at § of the critical tem- 

 perature is ( u Laws of Molecular Force/'' Phil. Mag. 5 ser. 

 xxxv. p. 258), in terms of 10 12 dynes as unit of force which 

 has been used throughout my values of M 2 /, 



(M// 3 ) } =-109«/(M„/ ft )§;. . . . (17) 



or if a 2 is measured at the melting-point a rougher approxi- 

 mation to the truth is (** Further Studies/'' Phil. Mag. Jan. 

 1895) 



(M 2 2 y*=*097a 2 *(M 2 /^)-i (18) 



Thus the difference in the equations (15) and (16) above is 

 probably due to some fact not taken account of in interpreting 

 S for all the types as being « 2 //o 2 2 ; the difference would be 

 explained if If for the type NaOl were too small in the ratio 

 of 13 to 19, or in the types MgCl 2 , K 2 S0 4 too large in the 

 ratio of 19 to 13; that is, if b were too small in the ratio of 1 

 to 2*1 in the one case, or too large in the ratio 2*1 to 1 in the 

 other : in round numbers we have to do with an unexplained 

 factor 2 or ^ in one case or the other. 



Let us now compare the equations (15) and (16) with (17) 

 and (18) ; the coefficients "08 and *05 are smaller than the 

 •109 and *097, because a 3 , being measured at the much lower 

 temperature of 15° C, is larger, for the lower the tempera- 

 ture the smaller must the coefficient be. We can determine 

 approximately whether the difference in the coefficients really 

 corresponds to the difference in temperature, for according to 

 the discovery of Eotvos (Wied. Ann. xxvii.), confirmed by 

 llamsay and Shields, the change in a 2 (M 2 /p 2 ) 3 for 1 degree C. 

 change of temperature is nearly *227 for all normal liquids; 

 and it" we extend this principle into the solid state for, say, 

 NaCl, whose absolute melting-point is 1045, then at 15° C. 

 or 288° absolute a 2 (M 2 //9 2 ) f increases by *227 x 757 above 

 what it is at the melting-point; but at the melting-point a. 2 is 

 11*6 according to Quincke and Traube, and at the melting- 



