Structure and Affinities of Leuconostoc mesenteroides. 83 
isomorphous with Streptococcus. There is thus no fast morpho- 
logical distinction between the two genera, but they can be 
easily distinguished by the fact that the dominant phase in the 
life-history is different in the two cases. 
The obvious affinity of Leuconostoc with Streptococcus does 
not enable one to connect the genus with autotrophic forms. 
In fact many of the species of Streptococcus appear to be more 
advanced both in their morphology and biochemistry. The 
parasitic habit of Stveptococcus has already been referred to. 
The genus Diplococcus, which resembles Leuconostoc in its paired 
or sometimes seriate cells and in its frequent formation of cap- 
sules, is even more specialised chemically, ‘‘ growing poorly or 
not at all on artificial media” (11). Morphological complexity is 
seen in the development of long chains of cells in some species 
of Streptococcus whilst the special larger cells, already referred 
to in Leuconostoc, are even more conspicuous here, although their 
nature does not appear to have been studied. 
The envelope of Leuconostoc has already been noted as 
staining with the stains used for demonstrating those of bacteria. 
It thus reacts, on the methods of Friedlander and of Welch, in 
the same way as the capsule of Dzflococcus lanceolatus. We 
have little information concerning the capsules of bacteria in 
general. The observations of Brautigam show the gelatinose of 
Micrococcus gelatinosus to be similar to, if not identical with, the 
chief constituent of the envelope of Leuconostoc. On the other 
hand the presence of nitrogen compounds was established by 
Beijerinck in Streptococcus hollandicus, and Hamm found bac- 
terial capsules consisting of protein(s). It would thus appear 
that, although we can trace no near affinity of Leuconostoc 
mesenteroides with the Cyanophyceae on the grounds of its 
membrane the latter differs still more remarkably from that of 
certain other bacteria and in its carbohydrate nature would 
seem to be relatively primitive. 
The genus Leuconostoc is thus allied to certain bacteria and 
must be provisionally placed in the Coccaceae. A more detailed 
analysis of the systematic characters of the various other repre- 
sentatives of this family will, of course, be necessary before 
their phylogenetic relationships can be fully understood. That 
Leuconostoc may be a fairly primitive member of this plexus is 
suggested by the fact that its mode of life is not so highly 
specialised as that of the majority of the Streptococceae and 
by its palmelloid character which, by analogy with the algae, 
must be regarded as a more primitive stage of development 
than the filamentous. The capsulated forms in the young 
colonies of the higher Streptococceae thus appear as phases 
recapitulating the ancestral type. The resemblance of Leuco- 
6—2 
