bages, cucumbers, potatoes, &c. In the carnation disease and in
certain diseases of tobacco and other plants the seat of bacterial action
appears to be the parenchyma, and it may be that Aphides or other piercing
insects infect the plants, much as insects convey pollen from plant to
plant, or (though in a different way) as mosquitoes infect man with
malaria. If the recent work on the cabbage disease may be accepted, the
bacteria make their entry at the water pores at the margins of the leaf,
and thence via the glandular cells to the tracheids. Little is known of the
mode of action of bacteria on these plants, but it may be assumed with
great confidence that they excrete enzymes and poisons (toxins), which
diffuse into the cells and kill them, and that the effects are in principle
the same as those of parasitic fungi. Support is found for this opinion in
Beyerinck's discovery that the juices of tobacco plants affected with the
disease known as "leaf mosaic," will induce this disease after filtration
through porcelain.

[Sidenote: Symbiosis.]

In addition to such cases as the kephir and ginger-beer plants (figs. 19,
20), where anaerobic bacteria are associated with yeasts, several
interesting examples of symbiosis among bacteria are now known. _Bacillus
chauvaei_ ferments cane-sugar solutions in such a way that normal butyric
arid, inactive lactic acid, carbon dioxide, and hydrogen result;
_Micrococcus acidi-paralactici_, on the other hand, ferments such solutions
to optically active paralactic acid. Nencki showed, however, that if both
these organisms occur together, the resulting products contain large
quantities of normal butyl alcohol, a substance neither bacterium can
produce alone. Other observers have brought forward other cases. Thus
neither _B. coli_ nor the _B. denitrificans_ of Burri and Stutzer can
reduce nitrates, but if acting together they so completely undo the
structure of sodium nitrate that the nitrogen passes off in the free state.
Van Senus showed that the concurrence of two bacteria is necessary before
his _B. amylobacter_ can ferment cellulose, and the case of mud bacteria
which evolve sulphuretted hydrogen below which is utilized by sulphur
bacteria above has already been quoted, as also that of Winogradsky's
_Clostridium [v.03 p.0170] pasteurianum_, which is anaerobic, and can fix
nitrogen only if protected from oxygen by aerobic species. It is very
probable that numerous symbiotic fermentations i