288 MORRIS LOEB 



m = 0.0008, ^ of the molecules of argentic nitrate are disso- 

 ciated, we can fairly assume that the limiting values of X for 

 the other silver salts may be obtained by raising by 0.75 per 

 cent the values found for m= 0.0008. 



Table IV summarizes the measurements made to deter- 

 mine the effect of temperature upon conducting power; in 

 these the platinum electrodes usually employed were replaced 

 by silver ones, which gave very sharp readings. The mea- 

 sured conductivities at 0, 18, and 28 are denoted by XQ, X 18 

 and X 28 respectively; to obtain the coefficient of temperature 



Y^ must be diminished by 1 and divided by 10; ^- was 



A 18 A 25 



calculated from such an interpolated value of X 2 5- 



TABLE IV 



AgNO s ....... 0-1 0-638 1-213 0-555 



0-02 0-638 1-217 0'554 



0-005 0-632 1-222 0'548 



AgClOg ....... 0-005 0-626 1-222 0-542 



'AgClO 4 ....... 0-005 0-632 1*224 0-547 



AgC IO H 7 SO 3 ...... 0-005 0-615 1'242 0-526 



AgC,H 5 SO 3 ...... 0-005 0-607 1-241 0'519 



AgC 2 H 3 2 ...... 0-005 0-611 1-237 0-524 



AgC 9 H u SO 3 ...... 0-025 0-609 1-242 0'521 



0-006 0-606 1-246 0'517 



Ag 2 S 2 O e ....... 0-025 0-631 1-222 0'547 



0-006 0-631 1-226 0'544 



[119] 7. TEST OF THE LAW OF KOHLRAUSCH FOR 

 EXTREME DILUTIONS 



For this purpose we need, besides the limiting values for 

 conduction, which we have just shown to be attainable, the 

 limiting values for the rate of transference, likewise for ex- 

 treme dilution. 



At the close of section 5 we noted that Hittorf found the 

 value of n to be independent of the state of dilution where 



