ce have arisen from the foregoing.
About 1880 Pasteur first showed that _Bacillus anthracis_ cultivated in
chicken broth, with plenty of oxygen and at a temperature of 42-43deg C.,
lost its virulence after a few "generations," and ceased to kill even the
mouse; Toussaint and Chauveau confirmed, and others have extended the
observations. More remarkable still, animals inoculated with such
"attenuated" bacilli proved to be curiously resistant to the deadly effects
of subsequent inoculations of the non-attenuated form. In other words,
animals vaccinated with the cultivated bacillus showed immunity from
disease when reinoculated with the deadly wild form. The questions as to
the causes and nature of the changes in the bacillus and in the host, as to
the extent of immunity enjoyed by the latter, &c., are of the greatest
interest and importance. These matters, however, and others such as
phagocytosis (first described by Metchnikoff in 1884), and the epoch-making
discovery of the opsonins of the blood by Wright, do not here concern us
(see II. below).

[Sidenote: Form and Structure.]

MORPHOLOGY.--_Sizes, Forms, Structure, &c._--The Schizomycetes consist of
single cells, or of filamentous or other groups of cells, according as the
divisions are completed at once or not. While some unicellular forms are
less than 1 [micron] (.001 mm.) in diameter, others have cells measuring 4
[micron] or 5 [micron] or even 7 [micron] or 8 [micron], in thickness,
while the length may vary from that of the diameter to many times that
measurement. In the filamentous forms the individual cells are often
difficult to observe until reagents are applied (_e.g._ fig. 14), and the
length of the rows of cylindrical cells may be many hundred times greater
than the breadth. Similarly, the diameters of flat or spheroidal colonies
may vary from a few times to many hundred [Sidenote: Cell-wall.] times that
of the individual cells, the divisions of which have produced the colony.
The shape of the individual cell (fig. 1) varies from that of a minute
sphere to that of a straight, curved, or twisted filament or cylinder,
which is not necessarily of the same diameter throughout, and may have
flattened, rounded, or even pointed ends. The rule is that the cells divide
in one direction only--_i.e._ transverse to the long axis--and therefore
produce aggregates of long cylindrical shape; but in rarer cases
iso-diametric cells divide in two or three directions, produc