480 Mr. W. Sutherland on Molecular Force 



Table III. 



]S T aCl. KC1. NaN0 3 . KN0 3 . MgCl,. 0a01 2 . Sr01 2 . BaCl 2 . 



b 6-47 3-8 4 6 313 65 ~ 6-67 4-6 12 



c 36L 38 2-7 333 29 271 1*7 18 



Na 2 C0 3 . K 2 C0 3 . Na.,S0 4 . K 2 S0 4 . MgS0 4 . ZnS0 4 . CuS0 4 . 



b 91 6-68 5"-5 18-2 7*9 5-2 73 



c -8 1-80 1-8 16 -8 -2 2 



Now a 2 /p2 2 i s proportional to 2 A 2 (m 2 //) 2 )*> so that m> being 

 actual mass and M the mass when the atom of H is 1 

 (a 2 V/°2) M 2 (M 2 //o 2 ) — i is proportional to (M 2 2 2 A 2 )*, and therefore 

 to (M 2 2 / 2 )* ? the form under which the parameters of molecular 

 attraction have been studied in my previous papers ; therefore 

 if b is really « 2 /p 2 2 , then PM 2 (M 2 //p 2 ) — I is proportional to 

 (M 2 2 Z 2 )*. In " Further Studies on Molecular Force " (Phil. 

 Mag. Jan. 1895) the values of (M 2 2 Z 2 )* for most of the above 

 substances are given, either as calculated from their surface- 

 tensions at the melting-point, or according to the Kinetic 

 Theory of Solids. In the next table are given the values by 

 the latter method from tables xiii., xiv., xix., and xxi. 

 of " Further Studies," as well as values of M 2 /p 2 and the 

 ratio ^M 2 /(M 2 / i o 2 )-i-r-(M 2 2 / 2 )* in the last row. 



Table IV. 



NaCl. KC1. NaN0 3 . KN0 3 . MgCl 2 . CaCl 2 . 



M.Jp 2 27'3 38-3 39-0 48'8 43-6 50-0 



(M 2 %)5 ... 5-7 6-8- 7-1 8-3 73 79 



151 11-6 13-9 11-3 177 189 



SrCl 2 . BaCl 2 . Na.,COg- K 2 C0 3 . Na 2 S0 4 . K 2 S0 4 . 



M 2 /p 2 51-9 541 43" BOO 54-0 660 



(M 2 %)* ... 8-7 9-5 6-2 7-6 7"8 89 



202 12-3 19-5 16-8 15-5 294 



The mean value of the ratio for the compounds of the 

 monad metals with the monobasic acid radicals is 13, and for 

 the others involving either dyad metal or dibasic acid radical 

 the mean value is 19, which is half as large again as 13; but 

 the departures from the mean such as 12*3 for BaCl 2 , and 29*4 

 for K 2 S0 4 are too large te allow us to determine whether 

 MM 2 (M 2 /p 2 ) — J as determined from solutions is proportional to 

 (M 2 2 Z 2 )* as determined from the solid, but they do not 

 determine that the proportionality does not exist, because b 

 being the coefficient of the smallest term in (8) is very 

 seriously affected by experimental error or slight defect in the 



