1911.] = On Variation and Adaptation im Bacteria. 558 
‘through the test media, and showed other changes of reaction as exhibited in 
‘Table VII above. 
The general bearing of the results obtained in the experiments recorded 
above on the problem of the identity or diversity of human streptococci, and 
on the value of test media in the diagnosis and differentiation of these 
organisms may now briefly be indicated. 
The streptococci shown in Tables I, II, and III undergo such agen changes 
under cultivation in stab agar as to exhibit collectively 28 alterations of 
individual reactions out of a possible 70, namely, nine gains out of a 
‘possible 28, and 19 losses out of a possible total of 42. 
Among the seven streptococci in Table III which each gave a different series 
“of reactions in 1908, and could therefore at the date have been classified as 
-different varieties, L, P, and V give identical reactions in 1910, and 
S, M, and H are also identical, but give a slightly different series of 
reactions from the former, while G still stands alone, although itself 
somewhat changed. Examined for the first time at this date these seven 
‘streptococci might, therefore, have been regarded as representing only three 
varieties. 
At the end of the mannite experiment (Table VI) M is no longer 
identical with 8, but has assumed the reactions of the L, P, V group, while 
V has left this group and now clots milk. 
Streptococcus G, after three months in agar following eight weeks in 
inulin (Table V), gives the same reactions as M (vi, 1910) in Table III. 
And L, 8, and G,in Table V, give identical reactions at the third stage— 
all positive, and only differ from M in that the latter still lacks the power of 
‘clotting milk, which it possessed when first observed in 1908. 
It is unnecessary to adduce further instances, though a number occur in 
the tables, that streptococci, which are at one time different, may at 
another give identical reactions, while those which are apparently identical 
at a given date may later on exhibit totally different series of reactions in 
the test media. But it is clear that no trace of specificity has been found in 
the tests in question in the present observations. 
Looking through the tables one finds that Streptococcus L exhibits no less 
than seven different series of reactions in different circumstances, P exhibits 
six different series, S exhibits nine, G exhibits seven, and M exhibits four. 
These five streptococci have, in fact, given rise to 32 different varieties or 
variants in the course of this investigation. It follows that the method of 
identifying varieties of streptococci by means of the series of test media 
which have been employed in these experiments rests on no fixed or specific 
differences in the organisms themselves. The differences observed cannot in 
