OXYGEN FROM COLOURED BACTERIA. 137 
collected after 15 minutes, contained 0:2 per cent. CO, and 
("0 per cent. oxygen; a third sample, collected after 25 minutes, 
contained no appreciable trace of either O or CO, The tube 
was then at once sealed and, by inverting, the contents were 
mixed with HgCl, Examined after 8 hours, the enclosed H 
contained 0:5 per cent. CO, and 2'8 per cent. oxygen (per gramme 
of baeteria—0'2 c.c. oxygen). A prolonged current of H there- 
fore apparently gradually drives out the oxygen held in a state 
of loose combination by the bacterial pigment. The loss of oxygen 
is still more rapid if HgCl, has been previously added. Thus 
a tube of Bacillus brunneus to which HgCl, had been added 
before H was passed through evolved per gramme of bacteria 
(bouillon-culture) 0-057 c.c. oxygen; whereas if H was passed 
through a similar tube of normal material, and then, after sealing, 
this at once heated to 90? C., per gramme 0°11 c.c. of oxygen 
was evolved. 
This method cannot be claimed to give really accurate quanti- 
tative results, though from a qualitative point of view it is 
perfectly satisfactory. One serious error is due to the fact that 
any liquid enclosed has the power of holding in solution a certain 
amount of the oxygen present. This error is obviated by using 
relatively large amounts of hydrogen and reducing the amount 
of fluid enclosed to a minimum. Also, before sealing, the flame 
is allowed to play upon the upper part of the tube before the 
"amowing, and then the tube at once drawn out. On cooling a 
Th Vacuum is produced in the sealed portion of the tube. 
Basten once placed in water at 90° C. to 100° C., when the 
of a na " killed and at once cease to respire, and the presence 
"d h l vacuum prevents the tube from bursting. Minute 
bacterial ; gas can generally be seen to be evolved from the 
xo the taba when thus heated. After the lapse of an hour or 
off under TH. 18 taken out, allowed to cool, and the tip is broken 
show that S when the Hg as a general rule runs in so far as to 
atmosphere “stinet negative pressure varying from j to 4 of en 
lected and in still present in the tube, The gas is then col- 
t e measured , ysed. From the percentage of oxygen present in 
evolved can bent of H enclosed, the actual amount of oxygen 
€ calculated. In this way, using a moist weight of 
acter; . i 
in. 18 varying from 0-2 erm. to 1 grm., the following calcula- 
Ons Were made " 
