It appears as if Griffith may have been exploring
the qualities of various immune environments, including
concurrent vaccination, when he discovered the cell --x
cell effects.   If he had tried mixed culture in vitro
(in line with paragglutination precendents) he does not
say so. And without a special selective environment it
would of course not have worked.

THE PNEUMOCOCCLS

213

diffused uniformly throughout the serum. R colony cultures obtained at
any stage of the passage through serum are incapable of multiplying in
the mouse, and do not produce a fatal septicaemia (unless reversion to the
S form takes place), though large doses intraperitoneally may cause death
from toxic action.  There is some evidence, as shown by differences in
capacity to revert to the virulent form, that strains from individual R
colonies are not equally attenuated.

2. Growth in bile causes attenuation and changes in the character of
the colonies similar to those obtained in immune serum.
3. Other methods of producing attenuation are by growth in optochin,
in meat infusion, in acid broth and at temperatures over 39" C.
4. Attenuated R strains have also been obtained from the blood of
highly immunized horses inoculated with living virulent pneumococci.
Pneumococcus cultures which have become partially attenuated in
the course of cultivation on artificial media and have dissociated into a
mixture of R and S forms regain their virulence after passage through
animals.  The effect is in part the result of the elimination of the
attenuated forms, and the survival of those best suited for multiplication
in the animal body.

Selection of particular cells, however, is not the only mechanism
involved in increase of virulence.  Avirulent strains derived from single

pneumococci, and who& composed of R forms, may revert to the
virulent S form after inoculation into animals (see p. 209).  It must
also be mentioned that Dawson and Avery (1927) and Dawson (1928)
found that virulence and type-specificity could be restored to R pneumo-
cocci bv growth in an antisem prepared by immunizing rabbits with
avirulent R pneumococci.  Moreover, Felton and Dougherty (1924) have
reported that virulence may be increased & vitro by frequent transfers in
milk at 4-hourly intervals ; they also found that peptone in 2 per cent.
solution maintained and even increased the virulence of a strain of
pneumococcus.  The possibility of determining virulence by the supply
to the `growing organism of definite chemical substances is a subject
worthy of further study.

Pneumococci obtained from the lesions of pneumonia patients are
generally of high virulence for mice, causing fatal septicazmia within
48 hours in a dose of 10e7 c.cm. of broth culture.  On the other hand,
these same sources have yielded cultures which failed to kill-in much larger
doses, and Neufeld and Handel found strains of pneumococci which killed
mice when injected in very small doses, but only after a period of 5 to
7 days. As a rule pneumococci freshly isolated from human beings only
attain their maximum virulence for a particular species after several
passages through animals of that species.  Moreover, high virulence for
one species does not necessarily signify equally high virulence for another
species.  Tillet (1927) examined 11 strains of Type III, obtained from
cases of pneumonia, 10 of which were of low virulence and 1 of high
virulence for rabbits, while all killed mice in a dose of lo-' c.cm.


THE PNEUXOCOCCUS            209

artificial cultivation under suitable conditions. After many years sub-
cultivation the chief type strains sfm~ no divergence from their original
characters, provided that their virulence is maintained by occasional
passage through animals.  \Vhether pneumococci may change from one
type to another under natural con&ions, e.g. in human infections, has
not been shown, though it is difficult to conceive that the chief types, as
well as the innumerable serological t-arieties of Group IV, are absolutely
fixed and unalterable.  It is well known that after recovery from
pneumonia the infecting t&ype, I or II, tends to disappear from the
respiratory tract and is often replaced by a Group IV strain. Frequently
several different varieties of Group IV may be found in the sputum of a
convalescent pneumonia patient-  According to the more generally
. acckpted view, the chief types die out and the Group IV strains, origin&ly
present in the nasopharynx before the development of pneumonia, can
alone be demonstrated.  As an alternative to the above hypothesis one
may consider the possibility, as the author has done (Griffith, 192S),
that under the influence of the immune substances developed during
recovery the pneumococcus responsibll- for the pneumonia may assume
different serological characters, that i: to say, the Group IV strains may
be derived from one or other of the chief types.

M~difcatiaP3 Experiments.

Morgenroth, Schnitzer and Berger (1925) have reported that they were
able by special methods to transform pneumococci into streptococci ;
Reimann (192T), following the methods devised by Xorgenroth and his
collaborators, repeated their e.xperiments and also obtained variants.
These he identified v,irh the R form of pneumococci obtained by various
other methods. The R pneumococcus produced or. blood-agar colonies
resembling those of S. z~irida~ts, but, unlike the latter, the R colonies were
invariably bile-soluble though slightly more resistant to the action of bile
than the type-specific pneumococcus.

  The writer (Griffith, 1928) has described a method by which one type of
pneumococcus can apparently be converted into another. The principle
of the method is to supply to fhe liting R form of one type of pneumococcus
a pabulum consisting of a dense suspension, killed by heat, of a virulent
pneumococcus of another type fmm which to build up its type-specific
antigen. The change of type has been obtained only when the mixture
of living and heated cultures are injected into mice, and not in the test-
tube. A preliminary observation which suggested the procedure was the
discov&y that the attenuated R form of the pneumococcus when
inoculated subcutaneously into a mouse together with a mass of killed
virulent culture of the same type readily reverted to the original type-
specific form.  The R form inoculated alone rarely reverted unless very
large doses were given, e.g. the deposit of 50 to 100 c.cm. of broth culture.
The temperature to which the uirulent culture was heated esercised an
important influence on the results and conversion from one type into