Messrs Laby and Carse, On a relation, etc. 291 



Ionic Velocities. 



The velocities used in tbe table have been obtained from 

 Bredig's memoir*. To convert the velocities as given by Bredig, 

 into centimetres per second per volt per centimetre, the formula 



k dV 



u + v = -5— 



nq ax 



must be used, where u and v are the velocities of the cathion and 



anion respectively in cms. per sec, k the specific conductivity in 



reciprocal ohms, n the concentration in gram-equivalents per cub. 



cm., q the electrical charge in coulombs carried by a gram-equiva- 



dV 

 lent of ions, and -7— the potential gradient in volts per cm. 



Bredig's units are the same as these except that the conduc- 

 tivities are in Siemens' mercury unit, and the potential gradient 

 is one volt per cm.; to be applicable to his data the expression 

 becomes 



1-063 A? ln in _ 5 & __ in _ 



« + *>= nn * ac = 11 X 10 -5 -= l'l X 10 -1i /A . 



n x 96440 n ^°°. 



Thus the values of the velocities given by Bredig were 

 multiplied by 11 x 10~ 5 . 



Densities. 



These were got from several authorities, chiefly from Clarke's 

 "Table of Specific Gravities." In some cases the original authority 

 does not state the temperature of the water used in the determin- 

 ation of the specific gravity. When it is given, the density in 

 grams per cub. cm. has been calculated. 



Explanation of Table. 



In the table there will be found the ionic velocities and the 

 data used for deducing the cube roots of the ionic volumes. The 

 densities could not all be obtained at the same temperature, but 

 the temperatures selected were, in as many cases as possible, the 

 same for each series. The variation with temperature of the ionic 

 volume in the range of temperature used is small, e.g. the ionic 

 volume of the propionic acid anion is 67 "7 at 0° and 71 "7 at 50°, 

 a change of "1 p. c. per degree. An error in a density affects the 

 product in the last column to about one-third of the extent that 

 the same error in ionic velocity does. 



* Zeit. f. Plujs. Chem. xin. 11)1 (1894). 



20—2 



