Patten—Solutions of Hydrochloric Acid in Solvents. 323 
sulphuric acid, calcium chloride and phosphorous pemtoxide as 
rigorously as the hydrocloric acid gas just described was run 
into' this bottle with the test-tube containing the solvent still 
unbroken. This hydrogen was left running all night and the 
experiment finished in the morning, when the bottle was closed 
by screwing down the pinchcocks, connected to the hydrochloric 
acid generator and the pinchcocks next the generator opened 
first and allowed to stand for fifteen minutes in order that any 
moisture from the air in the tube by which connection had been 
made might be absorbed. 
Before taking any hydrochloric acid from the generator the 
gas was allowed to run through all the train, with the bottle 
containing the solvent and metals cut out, until there could be 
no question that all air had been displaced. Then the gas was 
run into the bottle under considerable pressure in order to pre¬ 
vent the solvent rushing back into the drying towers—hydro- 
chloic acid gas is much more soluble in some of the solvents 
used than might be supposed. 
2. The second method of preparing the solvent is that given 
in detail under the experiment in which benzene is used as solv¬ 
ent. Owing to the controversy concerning the action of a solu¬ 
tion of hydrochloric acid gas in this solvent, I have given every 
point. To save repetition, the reader is referred to' this experi¬ 
ment. In other cases, to' relieve all doubt as to the conditions 
of experiment I have been very explicit even at the risk of repe¬ 
tition. 
Method 1 is more satisfactory for solvents which are self- 
dehydrating; that is, which decompose water with formation 
of oxychlorides and hydrogen chloride gas. While it is true 
that the oxychlorides might promote chemical action, the con¬ 
ductivity measurements show that at least no electrolytic dis¬ 
sociation exists in these solutions. 1 
1 I mean simply that it would be carrying the theory of electrolytic dis¬ 
sociation to an extremity to claim ionization in a solution or solvent which 
conducts electricity less well than a dry air gap. That we do get a cur¬ 
rent even across the air-gap is unquestionable, since the voltmeter used 
as ammeter is deflected. But that this current is carried electrolytically 
is a point which requires experimental proof more rigid than any yet at 
hand. 
