OXYGEN FROM COLOURED BACTERIA. 153 
operate, for in the darkness the Bacteria at first come nearer to 
the surface, as if owing to the cessation of the evolution of 
assimilatory oxygen, but after being in darkness for a longer 
time they tend to distribute themselves more evenly throughout 
the upper and lower layers, and after being in darkness for a 
day or two commence to sink to the bottom. 
Various facts seem to indicate that the purple Bacteria are 
really possessed of an actual, though extremely weak, power of 
assimilation. Thus, like the green Bacteria, they can grow and 
multiply in fluids in which but a trace of organic material is 
present. For the life of the red Bacteria the presence of SH, 
is essential, whereas the green Bacteria, though they can with- 
stand the presence of a trace of SH,, thrive best in its absence. 
Winogradsky has shown that the sulphur Bacteria obtain their 
main supply of energy by the oxidation of the sulphur, derived 
from SH,, which they contain. Hence, perhaps, the weaker 
power of assimilation of the red Bacteria as compared with the 
green ones, which are apparently dependent almost entirely, 
under normal conditions, upon the oxidation of the carbon 
obtained by assimilation for their supply of energy. In saturated 
SH, water, or in sealed tubes with pure SH,, even if exposed to 
light, the red Bacteria die in a few days—i. e., they are not really 
anærobes, If, however, kept exposed to light in sealed tubes 
containing a little H-and a trace of SH,, or in capillary tubes, or 
in closed cells, the Bacteria, especially Bacterium photometricum, 
may remain living and motile for weeks, and also divide and 
multiply » the oxygen necessary for both the sulphur and carbon 
respiration being produced by the Bacteria themselves. In dark- 
ness they are robes, in light apparently are anzrobes. 
That the red Bacteria do actually evolve CO, was proved by 
eae almost perfectly pure water cultures free from all SH, 
e gas-chamber by the methods already described as well as 
rie analysis. In light the amounts of CO, evolved and O 
any a atual « slightly less than in darkness, but in no case can 
ordinar SE tion of oxygen be detected analytically. With 
contain, . ; containing cultures, the gas finally analysed always 
of these Ba ne SH, derived from the enclosed fluid. In none 
occludin " eria does the pigment appear to have any power of 
bardly 5 1 een. An evolution of oxygen so weak that it can 
ar as " ected by Spirillum caunot be detected by analysis. 
Liny, wou, evidence goes, a qualified acceptance of the presence 
RN.— BOTANY, VOL. XXXIII. M 
