a a7 
containing urine and ordinary atmospheric air! liber-’ 
ated the liquor potassz after the flask and its con- 
146 THE EVOLUTION OF LIFE 

PiG..9. 
Retort with platinum electrodes for the liberation of oxygen. 
tents had been boiled, with the effect of finding fer- 
mentation take place several hours earlier than it 
did in other companion retorts whose contents were 
similar, except for the fact that they contained no 
atmospheric air. The liberation of the liquor 
potassze from the tube with a capillary neck can only 
be brought about with great difficulty in a retort 
containing air; and this was the reason of my giving 
up the conditions more favourable to the fermenta- 
tion of the boiled urine, in order to avail myself of 
the facile and automatic emptying of the liquor- 
potassee tube that takes place in the retort from 
which air has been expelled by boiling.? 
' The flasks having been hermetically sealed while the urine was 
cold. 
2 The first account of these experiments is to be found in Zhe Pro- 
ceedings of the Royal Society, No. 172 (June 1876) ; and there is also 
