OXYGEN FROM COLOURED BACTERIA. 127 
zinc dust is added, so that any blue tinge acquired whilst the air 
is being driven out can be removed by shaking. After a certain 
time the tubes between the flasks are clamped, and after a few 
hours the current is renewed, when the gas in the 2nd flask, which 
contains a bouillon culture of Bacillus brunneus, is slowly driven 
through the 8rd flask. The reduced indigo-carmine in the latter 
acquires in the upper layers a distinct blue tinge, showing that 
oxygen has been evolved from the culture of B. brunneus. The 
time necessary to drive out all oxygen from the flasks is deter- 
mined by previous experiments made with sterile bouillon. On 
reclamping and leaving for a few hours, no further evolution of 
oxygen can be detected from the B. brunneus culture. 
That the oxygen is evolved in the normal form as O,, and not 
as ozone or O, can be shown by testing with a hanging-drop of 
KT and starch in an atmosphere of hydrogen. 
The following experiment seems to show that the evolution of 
oxygen is a vital process continuing only as long as the Bacteria 
remain living. The usual chamber arrangement having been set 
up and a quantity of B. brunneus enclosed, a drop of concen- 
trated HgCl,, previously contained in a bulb in the tube leading 
to the chamber, is driven by the current of hydrogen, on 
inclining the tube, into the chamber and comes into contact with 
the B. brunneus lying on the floor of the latter, and separated 
from the outer half by a streak of vaseline. The evolution of 
oxygen from the bacterial mass is at first quickened, but then 
tapidly weakens and ceases entirely in about an hour or so. 
The process therefore appears to be a vital one, ceasing as 
Soon as the HgCl, reaches the centre of the bacterial mass and 
the Bacteria are killed. As will be shown later, however, this 
phenomenon has really a quite different bearing on the problem 
at issue. 
inne sight it appears as if we had here to do with an 
assimilati process, the oxygen evolved being derived from the 
latory pigo, O, and the bacterial pigment being an assiml- 
the sment. Were this so, it would be natural to expect 
evolution of oxygen to take place more actively in an atmo- 
sphere of H + a little } n" f H alone 
That oxygen ittle CO,, than in an atmosphere o : 
that it is not n be evolved in an atmosphere of H is no proo 
alga exposed evel from a process of assimilation, for green 
Oxygen to k o light m an atmosphere of H evolve sufficient 
eep Bacterium Termo in a hanging drop in active 
