398 
Lincoln—Electrical Conductivity. 
therefore, to determine to what extent Arrhenius’ theory, which 
is based upon the behavior of electrolytes in aqueous solu¬ 
tions, can be applied to the solutions that are non-aqueous. In 
presenting the results of this investigation, an inquiry will also 
be made as to whether there is any relation existing between 
the dissociative power of solvents and their other general prop¬ 
erties. 
EXPERIMENTAL PART. 
Qualitative Determinations. 
In an investigation of the electrical conductivity of non- 
aqueous solutions, the chief requirement is that both the solv¬ 
ents and the dissolved substances be absolutely free from water. 
To accomplish this is not an easy task. Then, too, the ques¬ 
tion arises as to what anhydrous salts are soluble in the vari¬ 
ous solvents. In many cases this could only be answered by 
experiment. The work of other investigators on the conduc¬ 
tivity of non-aqueous solutions has been confined almost exclus¬ 
ively to the salts^of the alkalies, ammonium and the substituted 
ammonias. It was soon learned by direct experiment that the 
chlorides of many of the heavy metals are soluble in a number 
of organic solvents and that the solutions formed conduct elec¬ 
tricity. Therefore the investigation was confined chiefly to the 
conductivity of the solutions of the salts (mostly chlorides) of 
the heavy metals. Even when the salts were readily soluble, 
conducting solutions were not always obtained. In order to de¬ 
termine what solutions conduct and to estimate roughly, at least, 
the relative magnitude of the conductivity, it was necessary 
to perform an elaborate series of qualitative experiments to as¬ 
certain what solvents yield solutions having a conductivity suf¬ 
ficient to justify quantitative measurements. 
The method employed in making the qualitative determina¬ 
tions was as follows: Into a resistance cell of the Arrhenius 
pattern, the electrodes of which were about three millimeters 
apart, was introduced the solution to be tested. This cell, a 
rheostat, and a delicate galvanometer were placed in the circuit 
of a Leclanche element. The circuit w T as then completed and 
the deflection of the galvanometer needle noted. It was very 
