ption spectrum almost identical with that of
triphenyl-rosaniline. In the case of the scarlet and crimson red pigments
of _B. prodigiosus_, _B. ruber_, &c., the violet of _B. violacens_, _B
janthinus_, &c., the red-purple of the sulphur bacteria, and indeed most
bacterial pigments, solution in water does not occur, though alcohol
extracts the colour readily. Finally, there are a few forms which yield
their colour to neither alcohol nor water, _e.g._ the yellow _Micrococcus
cereus flavus_ and the _B. berolinensis_. Much work is still necessary
before we can estimate the importance of these pigments. Their spectra are
only imperfectly known in a few cases, and the bearing of the absorption on
the life-history is still a mystery. In many cases the colour-production is
dependent on certain definite conditions--temperature, presence of oxygen,
nature of the food-medium, &c. Ewart's important discovery that some of
these lipochrome pigments occlude oxygen, while others do not, may have
bearings on the facultative anaerobism of these organisms.

[Sidenote: Dairy bacteria.]

A branch of bacteriology which offers numerous problems of importance is
that which deals with the organisms so common in milk, butter and cheese.
Milk is a medium not only admirably suited to the growth of bacteria, but,
as a matter of fact, always contaminated with these organisms in the
ordinary course of supply. F. Lafar has stated that 20% of the cows in
Germany suffer from tuberculosis, which also affected 17.7% of the cattle
slaughtered in Copenhagen between 1891 and 1893, and that one in every
thirteen samples of milk examined in Paris, and one in every nineteen in
Washington, contained tubercle bacilli. Hence the desirability of
sterilizing milk used for domestic purposes becomes imperative.

[Illustration: FIG. 18.--A similar preparation to fig. 17, except that two
slit-like openings of equal length allowed the light to pass, and that the
light was that of the electric arc passed through a quartz prism and
casting a powerful spectrum on the plate. The upper slit was covered with
glass, the lower with quartz. The bacteria were killed over the clear areas
shown. The left-hand boundary of the clear area corresponds to the line F
(green end of the blue), and the beginning of the ultra-violet was at the
extreme right of the upper (short) area. The lower area of bactericidal
action extends much farther to the right, because the quartz allows more
ultra