The Limitations of the Method of Pure Culture. 123 
study, and they were first isolated by Winogradsky upon media 
composed of a silica-jelly impregnated with certain inorganic salts. 
The detailed further study of the pure cultures proved them to be 
highly susceptible even to traces of organic matter which might 
be in their neighbourhood, and this fact, surprising enough in the 
case of soil organisms, has been found also true of the organisms 
concerned with nitrification in sewage filters, where the amount of 
organic matter accumulated is comparatively large. An explanation 
for such an apparent anomaly may doubtless again be found in the 
artificial conditions obtaining when these organisms are separated 
from the others accompanying them in nature, and this view was 
in one case experimentally confirmed as follows. The organism in 
question, isolated from a sewage filter, was concerned with the 
oxidation of nitrite to nitrate, and when in pure culture was quite 
unable to grow in bouillon ; inoculated culture tubes remained 
indefinitely without sign of growth, and on re-inoculation into a 
nitrite-containing medium no oxidation occurred, shewing that the 
organism had perished. This Nitrobactcr was accompanied in nature 
by, and separated with difficulty from, an extremely small bacillus, 
which was not a nitrate-producer, and grew abundantly in bouillon 
and other nutrient media. The mixed culture of these two organisms 
proved very instructive. When inoculated into bouillon a copious 
growth occurred, and after six days the growth was sub-cultured into 
an inorganic nitrite-containing medium and also into fresh bouillon. 
After another six days the second bouillon culture was similarly 
treated and the operation repeated four times. It was found that 
even after four successive generations in bouillon, the inoculated 
nitrite medium showed production of nitrate, and one is forced to 
conclude that th e Nitrobacter was able, under these conditions, not 
only to survive inoculation into bouillon, but even to multiply there. 
This behaviour of the Nitrobacter must clearly be attributed to 
the presence of the accompanying organism, which in some way 
protects it from substances present by which its growth would 
otherwise be inhibited. How this protection is compassed is at 
present quite unknown, but we cannot too strongly insist on the 
importance of attempting to study more closely such cases of 
symbiosis, which abound in nature. 
The pure culture then, in such instances, is only a stage— 
though a necessary one—in the evolution of a perfectly scientific 
method of study. It is obvious that the next step should be to 
reproduce nature more exactly, by making and studying mixtures of 
