IsMied March 12, 1907.

U. S. DEPARTMENT OF AGRICULTURE,

BUREAU OF ANIMAL INDUSTRY. BULLETIN No. 95.
A. D. MELVIN, CHIEF OF BUKBAU. .

THE BACTERIOLYTIC POWER OF THE
BLOOD SERUM OF HOGS.

B. M. BOLTON, M. D.,

Bacteriologist, Biochemic Division, Bureau of Animal Industry.

WASHINGTON:

GOVERNMENT PRINTING OFFICE.
1907.

Issued March 12, 1907.

U. S. DEPARTMENT OF AGRICULTURE,

BUREAU OF ANIMAL INDUSTRY. BULLETIN No. 95.
A. D. MELVIN, CHIEF OF BUREAU.

THE BACTERIOLYTIC POWER OF THE
BLOOD SERUM OF HOGS.

B. M. BOLTON. M. D.,
Bacteriologist, Biochemic Division, Bureau of Animal Industry.

WASHINGTON:

GOVERNMENT PRINTING OFFICE.

1907.

BUREAU OF ANIMAL INDUSTRY.

\ I' MIIVIV

A**i*lnnt 1'liiff: \. M. I- \aiti s. TON.
Chief Cl.rL I- 1 .. B. Joi

M. KOK.XKT. chief; .1 VMK* A. KMKHV, a istant chief.
/Ann/ I hn.t i,,n: Ki>. 11. WKK-TKI:, chief ; (' I'.. I. A SB, assistant chief.
Iit*lcti'i Itirisinn: KICK I'. Si KIM* >M. chief . I". .. HOICK, as-ociati- .-Im-i; M..i!i!i-

rnthul,H/icnl I)iri*',n,t: .!<iu\ K. Moin.KK. chief; HENRY. I. \V \MIIH i;\. ai-tant <-hi*-f.

tjiiiinililini Itiiition: K|CIIMM>\V. II !< K MAN, chief.

i <>f '/.>K>IIJ>I: I>. II. ll \s-4iM. chief.

HiiiitSliitiiin: I-!. ('. S< IIHOKDKH. Hi]M-rmtcn<lciit ; \V . I:. < orTux,
JIiitlHiii'hiKin: (tKtiKiiK M. KOMMKI..
JAMK- M I'i KKN-.

KHH-IIKMIC IMVIMoN.

'7,/: M. IMK.SKT.

n,t 1'lii>f. .1 \MK> A. K.MKHY. .

Mint-in."/-, liinn Isihnrnlnr'ii.*: T. M. I'HK K. chcini>t in chariM-of central laUmitor\ :

A. K. (iltMIAM. K MTII Ho \i.l.A\Ii. ( '. II. SWAM.KH. WlI.I.IA.M I'.. SMITH. 1 \

IH^KK. CI.AKENJ-K T. N. MAIOII. I'IIILII- C\>IIKMAN. I'.. II. IM.KK.-OI.I , I'KTKK

VAI.AKK, itssitiUiiit clu-mi-ts in Iminch laboratories.
IliMj-rlinli rn Iit>f#ti,/,iti<>u.<<: Cliief of Division in charge: 1'.. M. I'>OI.TO\ ami ('. N.

Mi HRYDE, Iwcterioloiri.^ts; \V. 15. NII.K-. in>|H-ctor in charp- of t'n-l<l c\|K-riii)cntM.
I'millri/ l> I,, rimmt.*: K. \V. HIMWX, a^intant rhiMiiir-t in charge.

' l>i/,* <unl lHxinfi-1-tmit*: JAM !> A. MMKHY. chiMni>t; ( '. N M I'.i

r<--lirnln.n of Tiih,r<;ilin >n,>l Mntl,-in: Chief of IMvision in charge: A. M. Wf>T ami
II. .1. SIIOKK, ansistant ItacteriologiHts : \V. S. STAMTKH anl II. S. M- AIIKS.
arista nt.-.

LETTER OF TRANSMITTAL.

U. S. DEPARTMENT OF AGRICULTURE,

BUREAU OF ANIMAL INDUSTRY,

Washington, D. C., December 27, 1906.

SIR: I have the honor to transmit for publication the accompanj'ing
paper entitled "The Bacteriolytic Power of the Blood Serum of
Hogs," by B. M. Bolton, M. D., of the Biochemic Division of this
Bureau.

The experiments herein described were carried out with certain
strains of the hog-cholera bacillus, and are a part of the general
investigations concerning hog cholera which have been conducted for
some years by the Biochemic Division. The subject of bacteriolysis
has claimed much attention from bacteriologists in recent years, and
Doctor Bolton's work is believed to be of interest on account of the
light it throws upon the defensive mechanism of the animal bod} 7 in
its fight against infectious diseases.

I recommend that the paper be published in the bulletin series of
this Bureau.

Respectfully, A. D. MELVIN.

Chief of Bureau.
Hon. JAMES WILSON,

Secretary of Agriculture.

CONTENTS.

Introduction 7

Definition of sera used in experiments 8

Nature of the substance causing destruction of bacteria 9

Methods used in drawing blood, and description of cultures employed 14

The bactericidal potency of serum from the same hog at different times . . 15

Modifications which occur in the bactericidal power of serum upon standing . 19
Comparison of the bactericidal potency of the same serum for different strains

of bacteria 21

The potency of serum from arterial blood compared with that from venous

blood 22

Effects of heat upon the bactericidal power of hog serum 24

Effects of injecting cultures of B. cholerx suis upon the bacteriolytic power of

hog serum for this organism

Effects of dilution upon the bacteriolytic power of normal and immune serum.

Dilution of normal serum

Dilution of immune serum

Summary of the dilution experiments

General discussion

Summary of the conditions affecting bacteriolysis

The theory of complement diversion

Conclusions

Bibliography

ILLUSTRATIONS.

Ki<;. 1. Illustration of tin- incrliunism uf hai-ti-rinly-i- a<v..rdinir to the Khrlii'h

hypothesis 11

1'. I>iversin of complement in niiililiit-.l immune serum 13

.".. Partial bacteriolysis arnl partial diversion of i-iiiii|>li-iiicnt in ililuifd

iiiiiniiiic H-runi 13

4. DivenkMl Of Complement in ininnun' serum, not licatcl. ami without

tin- ail- lit ion of foreign -<un|.lcni-nt 51

THE BACTERIOLYTIC POWER OF THE BLOOD SERUM

OF HOGS.

INTRODUCTION.

Aside from a few preliminary experiments with cultures of the
typhoid and colon bacilli, the present investigation was confined to
the reactions taking place between three strains of B. cholwae ,y///.v,
and normal and immune serum' of hogs..

The bactericidal properties of the blood serum and other fluids of
the body have of recent } T ears attracted a great deal of attention from
bacteriologists, and the investigations of the subject have led to very
far-reaching results of both theoretical and practical importance.
Without attempting to enumerate in full the different properties
which blood serum has been found to possess in its various reactions
with bacteria and their products, those properties which bear upon
the present investigation may be briefh' stated as follows:

The blood serum of manj^ health} 7 animals when drawn into test-
tubes has the power of destroying a larger or smaller number of
bacteria when these are introduced into it. Thus large numbers of
anthrax bacilli are destroyed by rabbit's blood serum in test-tube
experiments. Typhoid, colon, and Asiatic-cholera bacteria are silso
destroyed by the blood serum of rabbits.

The same serum is not equally potent for different bacteria, and
the serum from different animals of the same or of different species
varies in bactericidal potency for the same organism. The chemical
reaction of the serum seems to exert some effect upon this quality of
the serum. The more alkaline the serum the more potent is its action
apparently, and consequently the venous blood has been found some-
times to furnish more potent serum than the arterial blood from the
same animal. 9 "

The injection of animals with the products of growth of bacteria
produces different effects upon the properties of the blood serum, de-
pending upon the kind of bacteria employed for injection. The injec-
tion of the bacteria of the group to which B. cholerx .*///*, B. typhosus,
and B. col! communis belong appears sometimes to have no effect upon

"The figure references refer to bibliography at end of bulletin.

8 HACTKKIOL1 IK 1 l'"\\ 1 K K KI.....M M HUM OF HOO8.

the bactericidal |>otclicy of tile scrum. At other times it sccin- t'

increase the potent \. but frequently it has :i very peculiar .lint, for
in some cases tin- scrum after the injection (if tin- J'liilual will no lei
kill any of tin- bacteria of tin- kind used to inject tin- :iniiii:il. l)iit on
the dilutioiiof the serum \\ ith an inditierent fluid, -m-h a- -alt Dilution.
it becomes strongly bactericidal for tin- bacteria of this kind. I'.ut
this serum beha\cs like the serum from an uninociilated animal with
other Imctrria than tho-r with which the animal is inoculated. Thus
the M-riini from the blood of a rallit injected with typhoid l>acilli will
often fail to kill the typhoid bacilli unless the serum i- givath diluted.
but it will kill ju.-t a- many Asiatic-cholera spirilla tin it did before the
animal was injected with the typhoid bacilli. 4 The peculiar behavior
of the serum from an animal injected with a culture of bacteria, in it-
failure to kill the species of bacteria with which the animal is inject, d
unless it is diluted, is called the Ncisser- WechslxTg '* phenomenon,
after the two investigators who first observed the reaction.

DKKIXITION OF SERA ISKD IN KXI'KKI M KM -.

The serum obtained from an animal which has been injected with
bacteria is spoken of as immune serum in contradistinction to the serum
from an uninociilated animal, which is called normal or fresh serum.
Thus si-rum obtained from a hog injected with //. <-Iioh r;e, suls would
be known as //. ,-//,,/, /, A-///.S- immune ho^ serum. So wherever immune
serum is mentioned in the present paper it is to be understood that it
refers to the serum obtained from an animal which has received at
last one injection of bacteria, but not necessarily that the serum has
either immunizing or germicidal pro|>ert ies. On the other hand, where
normal or fresh serum is mentioned, it is to be understood to refer to
serum obtained from an uninociilated animal.

^-ruiii loses in bactericidal potency on standing after being drawn
from the animal: and the higher the temperature to which the serum
i- e\j)osed the more rapid the loss of potency. It may remain jx>tent
for several days or for even a week or more in the refrigerator. Imt if
kept at body temperature it generally lo.-es all bactericidal properties
in three or four hours. Heating at .".: or f)'> ('. for ten or tifteen
minutes uUo robs the so nun of its bactericidal power, or rather this
treatment of immune scrum suspends its bactericidal power, which is
restored by the addition of a -mall amount of fresh scrum. This sus-
pension of Iwctericidal power by heating at 55 or 56 C. u called
inactivating the serum, and wherever inactivated serum is mentioned
it is to IM- understood to refer to serum which has Ix-en treated in
this \\ay. Inactivated serum to which fresh serum is added, and which
has hail its bacteriolytic properties restored in this way, \a termed
\ated scrum.

HOW BACTERIA ARE DESTROYED.
NATURE OF THE SUBSTANCE CAUSING DESTRUCTION OF BACTERIA.

In regard to the nature of the substance or substances in the serum
which cause the destruction of the bacteria which are introduced, there
is some difference of opinion. Some authorities maintain with Buch-
ner that there is only one substance, called by him " alexin," which
causes bacteriolysis, while others and these appear to be in the major-
ity maintain that there are two substances concerned. Those who
hold to the former view regard the effect produced by standing or by
heating as due to the weakening of the alexin, while those holding
the latter view explain this effect by a modification of one of the two
substances which they regard as necessary for bacteriolysis. All
are not agreed, however, as to the nature of these two substances.
Bordet 1>2>3 and the French school generally look upon the substance
which is not modified by heat as a sensitizing agent merely, which acts
upon the bacteria in such a manner as to make them susceptible to
the destructive action of the other body to which the bacteriotytic prop-
erties of the serum are directly due. Bordet borrows Buchner's term
"alexin " for this active agent, and he gives to the other body the name
"substance sensibilisatrice." According to the French school, then,
bactericidal serum owes its power to two substances, called, respec-
tively, substance sensibilisatrice and alexin. The former resists heating
up to 75 C. or even somewhat higher temperatures for an hour or
more, while the latter is destroyed by heating at 55 C. or higher in
ten or fifteen minutes. Moreover, the substance sensibilisatrice accord-
ing to this view is a specific substance in each case. The substance
sensibilisatrice for one kind of bacterium sensitizes this one kind only,
and while it can not of itself cause bacteriolysis, it nevertheless pro-
duces certain changes in the bacteria. On the other hand, according
to the view of Ehrlich and his school, this body, the thermostabile
body, causes no change in the bacteria themselves, but serves merely
as an intermediary, a binding link, serving to connect the bacteria to
the active body the alexin of Bordet which according to this view
is the active agent in bacteriolysis. To the two hypothetical bodies
concerned in the process Ehrlich has given, at different times, differ-
ent names. At first he called the thermostabile body "intermediary
body," afterwards "immune bod\ r ," and finally "amboceptor." The
last two designations are those now generally employed by the Ehrlich
school for the body which corresponds, as is evident, to the substance
sensibilisatrice of Bordet. To the other body, the alexin of the French
school, Ehrlich has given variously the names "addiment" and "com-
plement," the last being that now exclusivel}' employed. The torm
intermediary body has been employed in a recent publication for the
arnboceptors present in the normal serum in order to distinguish these
19500 No. 9507 2

10 HA. I 1KI..I.VTI. iMiWKK <>| i:|. ...... sKKl'M "I l!<"

from iiiiiiiunc amboceptor-. or ambo. -, pi,.! > produced l\ injecting
animal- with bacteria or their pro.lu.i-.

But among those who have accept.-. I mirror less completely the
Khrlieh conception of the nature of the Inxlir-. . <>IK . i n.-d in )>: -i, ! -i
olysts there is difference of opinion a- t tin- mode of a<ti<.n of
the complement. Khrlich himself says in regard to tin- matter
'that one will not go amiss if lie a-stime^ \viih IM'eiller that the pro.-
ess of bacteriolysis is analogous to digestion, ami attrilmte- i<> the
addiment (complement) the character of ad i^e-iive ferment." (Jruhcr/
on the contrary, contends that the complement do.-> not art lik< :(
ferment, and that it is erroneous to draw any analogy l-t \\een t Ill-
complement and an enzym, since the complement is entirely u-ed
up in bacteriolysis, whereas in the process of fermentation, as is well
known, the ferment is not used up, hut may In- recovered after tin-
action is ended, and used for the fermentation of other material.

But whatever the exact mode of action may IM>, it is evident from
what has just been said that both the Khrlieh and the IJordet -ehooU
attribute Imcteriolytic action proper in normal serum to a >ultam-e
easily changed by comparatively low temperatures, and called, respec-
tively, complement and alexin by the two schools. To the other
body concerned in bacteriolysis the amboceptor of Ehrlich and the
substance sensibilisatrice of Bordet is assigned by the former the
role of a binding link between the complement and the bacterium.
while by the latter is assigned to it the property of a sensitizer, or of
a mordant as in dyeing. In the one case the bacterial cell i> regarded
as not at all injured or otherwise changed by the union with the arnbo-
ceptor; in the other case it is the opinion of those holding thi> view
that the cell is acted upon and changed by the sensihili.sttrice in such
a way that the alexin can penetrate it.

Bordet* summarizes the difference l>etween his theory and that of
Ehrlich as follows:

According to Khrlieh and Morgenroth the sjHvilic antibody (aensihilisatric.- > ]>la\>
tin- !,!. of an actual intermediary (7.wiachenkor|H>r, amlxiceptor), a link of union
attaching itaelf on the one haud to the cell, on the other to the alexin. In other
words, the aheorption which tin- alexin undergoe** in the presence of the *-n-iti/i-<|
cell i.- not due to an affinity nianifcfted hy the cell it-.-li to thix tuilistance. The
absorption of the alexin is only indirect; the cell jin- itself to the intermediary
gulwtance, which i- itself united chemically by its other |tole to the alexin.

( >ur idea of the phenomenon, which we feel we are justified in holding, i.- alto-
p-ther different from thin. To us it seems that the Hengibilteatrice which uniu-.-
with the cell modifies this in a way which permits it to abxorb the alexin <lir.-.-tl\ .
The action of the sensibilisatrice upon the cells is comparable to that of certain fixa-
tivi- agents or mordants which confer upon certain substances, (or to the cell- i
these substances, as is the case in histological technic) the fiower of absorbing colons
which they refuse to absorb before treatment * * * It is to be clearly under-
stood, however, that when we speak of mordants in this connection we do not int-nd
t<> apply in all details the phenomena of dyeinx to the matter at present under rn-
; \\.- nu-rely mean to draw a comparison which will serve to make our

HOW BACTERIA ARE DESTROYED.

idea clearer. The hypothesis which we wish to bring out in relief is that in the
presence of hemolytic serum, the cell becomes capable of absorbing directly the
alexin by means of its own proper elective affinity, and that this power is due to a
change caused by the sensibilisatrice. In other words, we do not believe that one is
forced to admit, with Ehrlich and Morgenroth, that the sensibilisatrice itself com-
bines with the alexin, and that this union is indispensable for the latter substance to
attack the cell.

Bordet furthermore states in the same connection that he agrees
with Buchner in regarding the alexin for blood cells and for bacteria
as identical that one and the same alexin may attack the most diverse
cells; whereas Neisser and others of the Ehrlich school believe that
alexins or complements are different in one and the same serum.

While it is evident from the above that the terms amboceptor and
sensibilisatrice are used to designate the same substance, it is scarcely
correct to use them interchangeably, since they connote somewhat dif-
ferent attributes in the body to which they refer. The same is true
of the terms complement and alexin, though to a less degree.

The following diagrams, obtained from various sources and modified
to suit the purpose, will serve to illustrate the process of bacteriolysis
according to the views of the Ehrlich
school.

Figure 1 represents in its simplest
form the mechanism of bacteriolysis
according to the Ehrlich hypothesis,
and serves to illustrate the process suffi-
ciently for the purposes of the present
paper.

In the diagram the bacteria are rep-
resented b}- the parts marked 5, the am-
boceptors by those marked , and the
complements by those marked k. In
No. 1 the bacterium, amboceptor, and
complement are represented as just on
the point of uniting. No. 2 represents
the bacterium and the amboceptor united and the complement on the
point of uniting with the unoccupied end of the amboceptor. No.
3 represents the process of uniting of bacterium, amboceptor, and
complement completed; the bacterium in this case would undergo
bacteriolysis.

It should be borne in mind that according to this theory bacterioly-
sis can take place only where the bacterium becomes united to an
amboceptor which is itself united with a complement. A bacterium
may become united with a free amboceptor i. e., an amboceptor
which is not united with a complement but the bacterium in such a
case does not undergo bacteriolysis unless a complement subsequently
becomes attached to the amboceptor. The complement is incapable of
uniting directly with a bacterium; it can do this only through the

FIG. l. Illustration of the mechanism of
bacteriolysis according to the Ehrlich
hypothesis.

12 BACTKIUnl.VIlC I'nWKK <>K MU't-H -Mil M "I BOfl

intervention >f tin- amboceptor. Hut when the complement

linked t> tin- bacterium by mean- of tin- amhoceptor tin- l>a> i,-rjum

become- broken ti|) into minute granule- and nit imately di-appear-.

The twiiid- liy which t lu> ambocept< >r at tache- it-elf to the bacterium
on ili'- one hand, and I the complement on the other, an- called hapto
phor groups or haptophors (/<). and similarly thi- name i- given to the
bond* of union of the bacteria and of the complement. The ambo
eeptor thus has two haptophor-. one by mean-, of which it attache-
it-elf to the bacterium, the cytophylie ha|)tophor, an<l one by mean-
of which it attache-, itself to the complement, the complenientophylic
haptophor. The liacteria probably po e each many haptophor- all
of the same kind i. e.. hapto|>hors capable of uniting with anil .<! ptm--
of the same kind but for the sakeof simplicity the bacterium is repre-
sented in the diagram as having only one haptophor. The complement
ha- one hnptophor group and one so-called toxophor group (/). and it
is by means of the latter jrroup that the coinplenient a-ts upon the
bacterium. The complement may be deprived of this toxophor group,
and although it is still capable of uniting with the amlx>ceptor in such
a case, it can no longer cause bacteriolysis. This loss of the toxophor
group is caused by heating, and it also occurs spontaneously in the
-erum on standing. Bacteria subjected to the action of heated serum
do not undergo bacteriolysis, but become fixed to the aml>oceptor8,
and the amboceptors become united to the haptophor group of the
complement which are left unaffected by the heating. It will thus be
readily understood why bacteria treated with heated immune serum are
subsequently protected from bacterioly-is even when indicated immune
-erum or when indicated complement is added to them. The comple-
mentophylic haptophor of the amboceptor is in such a ca-e already
occupied by the haptophor- of the heated complement, which has
in thi- way become deprived of its toxophor group.

The amboceptors found in ordinary normal -erum are either all
alike and in this case they u\\\-\ po---- atlinity for a great many dif-
ferent kinds of bacteria or they must differ from one another; and
in this case there inu.-t evidently be a great many specific aotbooepton,
-ome titled for the hade-Holy .-is of one -pecie- of bacterium, -ome for
otheis. This matter -eeni- not yet to have been settled. Hut it is
certain that the injection of an animal with certain bacteria or their
products causes the formation of a large number of specific amlio-
ceptor-: that i- to -ay. of ainlxx'cptor- having affinity only for tin-
kind of bacteria with which the animal is injected. Such injection-
-eem not to increase the amount of complement.

Complement is found normally in the serum, that of some animals
|x>--e ing more than that of others. The horse- appear- to have a
Ia:-L r e amount of complement in the serum. It is not yd -dtled
whether the complement i- -peci tic that is, whether the complement
for one kind of immune serum can unite with the aml>oceptor- of thi-

HOW BACTERIA ARE DESTROYED.

FK;. 2. Diversion of complement in undiluted immune
serum.

serum only and not with the immune serum of a different sort or
whether complements are general; though they seem for the most
part not to be specific. The complement in the serum of horse's blood
seems capable of reactivat-
ing heated immune serum
of various kinds. Still in
some cases it would appear
as if they were specific.

With the explanation
given above of the nature
of amboceptors and com-
plements, the phenomena
which take place in im-
mune serum become more
or less satisfactorily ex-
plicable. By means of the characteristics ascribed to these bodies it
is possible to account for the peculiar behavior of immune serum
stated above, consisting in the fact that such serum is frequently
more potent when diluted than when it is undiluted.

Neisser and Wechsberg were the first to observe" this phenomenon,
and the theory which they advance to explain it the}* very appropri-
ately call the theory of the diversion of complement. As the name
implies, they attribute the lack of bacteriotysis in the undiluted
immune serum to the turning aside of the complement from the
bacteria, or rather from the amboceptors which are attached to the
bacteria. They hold that this diversion is brought about by the free
amboceptors themselves. In other words, where there are more
amboceptors than there are complements present in a serum, a part of
these attach themselves to the bacteria and a part to the complements.

The accompanying diagram (fig. 2),
taken from Neisser and Wechsberg,
and modified to suit the present descrip-
tion, shows two amboceptors a, at-
tached to bacteria >, and four ambo-
ceptors attached to complements k.
Bacteriolysis is not possible in such a
condition, because the complements
have been diverted from the ambo-
ceptors which are attached to the bac-
teria. Bacteriolysis can take place
only when the complement becomes
attached to the bacterium through the
medium of the amboceptor.
Figure 3 is meant to show the same serum diluted with an equal
amount of salt solution. In this case, with the same number of
bacteria added, it is evident that one-half of them would be killed, as

FIG. 3. Partial bacteriolysis and partial
diversion of complement in diluted im-
mune serum

14 HAITI Kh'I.VIH ToUlK <>K HI IMM >F U<(18.

i- indicated ly tin- combination lietwecii Iwteria. amboceptdr. :m.l
^implement in No. I. Tin- other half ol tin- bacteria \\<.nl.| e\ i
dently escape. the complement being dixcrted by tin 1 free ambo,-epi.r.
us shoxxn I iy Nds. _' and H.

Although these statement- iii regard id Uactei iolx -i- and I In- niechan-
i-m probably involved are by no mean- e\haii-iix e. they \\ ill. perhap-.
serve tin- purpd-e- df tin- present invcstigatidii and td explain the
results hero obtained. It will !> noticed that r\pn -iiiirnt- lia\- i< .-n
made with a view t> throuin^ additional li^ht npdii \\\n*.t> pha-< - .f
Imctrriolysis already nicnlidiifd. a- they weiv di^rrved in hoo-\ l>lod
-ennn, and it Wdidd appear that the re-id t". although diJferin^ in

-dine in-tance- frolll those of dtlier dh>er\ er-. are neverthele -U--

i-eptiMe of interpretation in hannony with more or less linnl\ e-tal-
lished hypothec -.

METHODS USED IN DRAWING BLOOD, AND DESCRIPTION OF
CULTURES EMPLOYED.

The blood wa> drawn from healthy hojjs, or from liojrs injected with
cultures of li. cholerw Unix, siibeiitaneously or intravenously. a> stated
in each case, at various lengths of time before the drawing. It wa-
obtained (1) by cutting off a piece of the tail and allowing the blood to
tlow into sterili/ed tubes or tla>k>. (L'| by blei-ding from an artery in
the ear, or (tt) by in-erting a canula into the carotid ai'terx <>] jugular
vein. In some cases the blood wa- drawn from the carotid and from
the jugular at the same operation. After drawing, the blood un-
usually placed in the refrigerator to allow the serum to -eparate. but
in a few ca-es it wa- u.-ed immediately after drawing.

All the hogs used in the experiment- were in a perfectly healthy
condition to start with. and. as far a- could le a-certained. had ne\ i
been -ick. They weighed from S<> to 4<> pounds each.

The cultures employed were obtained during the course of former
exjM'riments. One of them, G. 1'. 4' ( ( ,^. had been repeatedly pa ed
through guinea pigs; another, Crawford, had been carried along fora
number of years on artificial media without passage through animals;
the third, F. '20, had also been carried along for -everal years on arti-
ficial media without animal passage. The-e culture- piv.-ented minor
point- of difference from one another, but they were all quite typical
for II. linl,r tint*. All of them give characteristic growths upon
artiticial culture media with the usual fermentation reaction- <>f the
u-. and the other features of H. cliolt-rtf #//*, both macroscopic
and microscopic, though it is true that the Crawford strain grew more
vigorously and gave larger and denser colonies than the other two,
ind that the (. 1*. }>'.fJ .-train gave the least vigorous growth on arti-
ticial media. Al-o. in the test* whi< h were made to determine the
point, the (i. 1*. Jt.:-j -train was much more stronjrly pathogenic foi

BACTERICIDAL POTENCY OF SERUM.

guinea pigs, rabbits, and hogs than the other two strains. The Craw-
ford strain was the least virulent of the three. So the virulence of
the three organisms was in inverse ratio to the vigor of growth upon
artificial culture media. The culture of B. coli oommwtu was obtained
in the course of former experiments from the feces of a normal hog.
It showed no pathogenic properties.

THE BACTERICIDAL POTENCY OF SEBUM FROM THE SAME HOG
AT DIFFERENT TIMES.

The bactericidal potency of the serum from the same hog on differ-
ent days was tested by bleeding a hog at intervals of a few days,
drawing off the serum after the blood had stood in the refrigerator
for twenty-four hours or less, and distributing it into test tubes 1 c. c.
into each tube then adding a definite, measured amount of a suspen-
sion in salt solution of the bacteria to be tested. At the same time a
tube of 1 c. c. of salt solution was inoculated with the same measured
amount of bacterial suspension and plates made immediately to deter-
mine the number of bacteria added. Plates were made from the serum
tubes on the following day. The serum from 4 hogs was used in this
experiment, the blood in all four experiments being drawn from the
tail. The results will be found in the following tables:

TABLE I. Bacteriolytic action of normal hog serum from the same animal at different
drawings. Blood drawn from the tail. Hog No. 1783.

Date.

Time after drawing.

No. of
drawing
of blood."

Culture used to test
bacteriolysis.

Number of
bacteria
introduced
per 1 c. c.

Number of
bacteria
per 1 c. c.
after 1 day.

1905.
April 7

3 hours

G. P. 4692

4,620

79,800

B coli

5 880

30 800

G. P. 4692

30,800

48,300

.do

B. coli

7,000

1 day

III

G P 4692

4 160

9 800

III

B. coli

225

. . .do

G. P. 4692

9,800

4,690

B. coli

9,520

9,800

G. P. 4692

8,655

142,900

do ... .

B. coli

2,600

2 days

G. P. 4692.

4,968

43,200

B. coli

1,940

240

1 day

VII

G. P. 4692

1,800

1,220

VII

B. coli

2,800

16, 870

May 3

3 hours . .

VIII

G P. 4692

22,628

17, 510

VIII

B. coli

23,380

G. P. 4692

1,004

B. coli

869

2,594

u Roman numerals denote the serial numbers of the drawings of blood.

As will be seen from Table I, the blood was drawn from hog 1733
on nine different days and tested simultaneously upon cultures of G. P.
4692 and on the culture of B. coli communis. In four of the draw-
ings the serum apparently had no bacteriolytic power for the strain of
B. choleras suis employed, while in four of them, although there was
also no marked bacteriolysis, there was nevertheless no evidence of

BACTKKIOLYTK' I'oWKK "I I|.M,| S1-KTM

EIOO8.

in\ irivat iiHTi-a-e <>f tin- bacteria introduced. In tin- serum from one
of the drawings tin- bacteria were reduced to one-half of the number
introduced; this was the greate>t amount of potencx >hox\n iii :HI\ of
the drawings.

With the colon bacillus there \\a^ marked liacteriolysis in one draw-
in jf, loss marked in four other drawing. und neither i IK rea.se nor
decrease in one of the drawing, while in the other three draxx in;:-
there was an increase of the Itacteria introduced.

It is evident that the serum from this hog Allowed different degrees
of potency or alienee of potency upon the different cca-ion- when
the bl<x>d was drawn.

T \III.K II. />'</(/ r/o/i/fiV Hi-linn of normal Innj nfrtiin fnun tin' tainr animal ut iliffrrmt
dnnnnyx. ///<*.// ilru >n, from tin- tail. liny .V". 1740.

Date.

Tlnif after drawing.

No. of

IrauiiiK

..n.i.Nni.

Culture lined to tot
Uu-tcrlolyria.

Number of

ImrU-ria
Introdurcd
per 1 <. c.

Number of
tmcteri*
per 1
after 1 day.

190T>.

H ln>iir> . . .

O. P. 4692. . .

1.750

do ...

H. n>li

975

...do..

'. (' 4G9J.

l .' .

1 276

...do...

//. ntti

.. 771

...fin...

, IV .<

102

1,296

...do...

<i. I'. -KW2

1.010

1,562

.. do...

II ciAi . .

S.110

do ...

Q. P. 4692

848

...do...

O. P. 492

B. coH

1.700

1.840

The blotxl of hog 1740, as is shown in Table II, was drawn on six
different days and tested on It. >li and on (i. P. 4>>t'J. It showed
marked Imctericidal power for the B. cot! culture in three of the
four drawings tested with this organism, but in one there was no
decrease in the number of bacteria introduced. With the B. <-/,,>/ f -/';r
Kin* culture it will be noticed from the table that the effect varied
greatly with the different drawings. While there was no very marked
decrease in any case, it is evident in general that the bacteria did not
multiply abundantly in the serum in any case. It would seem as if
growth had l>een inhibited by the serum, although active lacterioly-i>
was lacking except in the serum from one drawing.

TABLE III. Ikttirrioliitir miion of nttrinnl hoy emm from the mme animal <il Mffertiit
Wo<i ilrnirnfrom the tail. Hog \o. 1741-

I'..t.

Time after drawing.

No. of

.'.: ,'.v ;: i:

of blood.

Culture tued to tect
bacterioly*!*.

Number of
barteria
Introduced
pt-r 1 r. r.

Number of
bacteria
per 1 c. c.
nfd-r Iday.

1906.
May B

;; ln>i;r

;

4 926

H. mii .

1.778

Q i 1 4692

102

166

.!,,

II |

R nil ....

1 970

390

III

,, |- |, ,

1 010

1 9S2

III

H rnli

3 110

.. do

. 1 ;

848

B. rnli

1,700

BACTERICIDAL POTENCY OF SERUM.

As Table III shows, the blood was drawn from hog 1741 on five
different occasions. The serum from four of the drawings was tested
upon the culture of B. coli, and it showed marked bacteriolytic prop-
erties for this organism in all the tests. The serum from four of the
drawings wasjtested upon the G. P. 4692 strain of B. cholerse, suix, but
only* two of the drawings showed bacteriolytic properties for this
organism.

TABLE IV. Bacteriolytic action of normal hog serum from the mme animal at different
drawings. Blood drawn from the tail. Hog No. 17 4%

Date.

Time after drawing.

No. of
drawing
of blood.

Culture used to test
bacteriolysis.

Number of
bacteria in-
troduced
per 1 c.c.

Number of
bacteria
per Ice.
after 1 day.

1905.
May 8

3 hours . . .

G. P. 4692

4 926

do ...

Ji coli

3 788

686

...do...

G. P. 4692.

102

B coli

1 970

III

G. P. 4692

1 010

798

.do

III

B coli

3 110

.... do

G P 4692

'848

B coli.

1 700

...do...

G. P. 4692

10,880

3,376

Crawford

40 500

2 640

F 26

19 040

18 280

...do...

VII

G. P. 4692

1,682

VII

Crawford

1 808

180

VII

F 26

12 040

1 640

VIII

320

. .. do. .

VIII

F 26

1 960

F 26

1 740

4 480

...do...

Crawford

2,320

do ....

G P 4692

1 920

960

June 5

s hour

F 26

2 840

26 640

do...

Crawford

3,380

180

G P 4692

1 380

4 420

In the experiments with the blood of hog 1742, as shown in Table
IV, ten different drawings were made on separate days, and four of
them tested on the culture of the colon bacillus, eight on culture G. P.
4602, five on culture Crawford, and five on culture F. 26.

The colon bacillus was greatly diminished in all cases. In one case
where 1,700 per 1 c. c. had been introduced into the serum all of them
were destroyed, and in another test 3,000 out of 4,000, in round num-
bers, were killed.

The tests with the G. P. 4692 culture showed that the serum was
bactericidal in seven tests out of eight. In one of these about 1,700
bacilli were introduced and all of them were killed. On the other
hand, in the serum from one of the drawings there was an increase of
the bacilli introduced.

The tests with the Crawford culture showed that the serum from all
of the drawings tested was actively bactericidal, but in some more
than in others. In one of the drawings about 38,000 bacilli per 1 c. c.
of serum were destroyed out of the 40,000 introduced. But in another
case all were not destroyed although only 1,808 were introduced.
19500 .No. 95-07 3

18 BACTKKIoI.YTK POWE1 0* HI. ..... > -IKl\l 01 I ..... -.

with culture F. _'' showed bacteriolysis in om* drawing; in
another drawing there was no increase nor decrease of tin- bacteria
introduced. In three drawings there was marked increase of the
organisms introduced.

That the bactericidal potency of the scnini of hogs -should vary :it
different times. a- these refill- seem to indicate. should |>crliaps'not
be a matter of surprise. Then- arc probably nmny circumstances
which influence this property of the serum. A- i- stated eUe \\hrre in
this paper, ihe chemical reaction of the serum has Ixvn found ly
others to influence the lactericidal power, and doubtle^ there are
other as yet oli-cure circuin.stances u hich raise or lower this p,\\er of
the sei-uin. The nature of t>aeteriol\ -in- will he found discussed at
some length helow in a different conne<-tion. and it would not seem
at all improhaMe that there may he more of the-e nt one time than
at another present in the blood serum. Indeed, the production of
hacteriolysis in serum of the living animal seems to lx> easily Influenced
one way or another, and it would not be unreasonable to regard them
us \aryinjj from time to time under even slightly changing' conditions
of the body.

Trommsdorf " noticed that human sera derived from normal indi-
viduals as well as from those suffering from various diseases \ a ry
greatly in bacteriolytic j>ower. Petterson 15 found the same thing
with chickens.

Morgenroth and Sachs 1 " found great variation incytolytic power in
serum of various sorts. Thu.s the serum from a horse at one drawing
\\ as hemolytic for rabbits' corpuscles but not for those of guinea pigs;
three days later the serum from the same horse was strongly hemo-
lytic for guinea pigs' corpuscles, but only very slightly for rabbits'
corpuscles; twenty-three days later the serum from this hor-e \\ as not
hemolytic for guinea pigs' corpuscles, but strongly hemolytic for
rabbits" corpuscles.

It is therefore evident that the cytolytic power of serum is very
variable. Not only does the blood from different individuals of the
same species differ in this respect, but the serum from the same indi-
vidual differs from time to time. This is prolmbly the case with all
animals, and. as is apparent in the experiments with hogs, these are
no exception to the rule.

In regard to the above exjM'riiuents. while there is more or less
variation between single tests, if one strain only of II. !,,, I, ,
considered, there seems t<> be a general tendency toward either active
l>a. teriolysis or else inhibition of growth.

MODIFICATIONS IN BACTERICIDAL POWER OF SERUM.

MODIFICATIONS WHICH OCCUR IN THE BACTERICIDAL POWER
OF SERUM UPON STANDING.

As has been previously stated, the blood was placed on ice in all
cases where it was not at once used for making tests. In some cases
the bactericidal properties of the serum were tested after the serum
had stood for various lengths of time. The results of these tests, as
will be seen from the accompanying tables, show in general that the
bactericidal potency is retained in some cases for as long as nine days.
Occasionally, however, the bactericidal power of the serum is dimin-
ished by even two days' standing.

TABLE V. Bacteriotytic action of normal hog serum after standing for various lengths of
time in the refrigerator. Blood drawn from the tail, eighth drawing. Hog No. 174%-

Date.

Time after drawing.

Culture used to test
bacteriolysis.

Number of
bacteria
introduced
per 1 c. c.

Number of
bacteria
per 1 c. c.
after 1 day.

1905.
May 20
June 2
May 26
June 2'

3 hours

Crawford

320
1,960
80
3,280

380
1,960
2,980

7 davs .

do .

3 hours

F. 26 "...

7 davs

...do...

The serum from the blood of hog 1742 at the eighth drawing, as is
shown in Table V, was tested upon two of the organisms only,
namely, Crawford and F. 26. For the Crawford culture the bacteri-
cidal power of the serum was retained for seven days, whereas for the
F. 26 culture the serum was bactericidal at the start and only inhibi-
tory after seven days.

TABLE Vf. Bacteriolytic action of normal hog serum after standing for various lengtlis of
time in the refrigerator. Blood drawn from the tail, ninth drawing. Hog No. 1742.

Date.

Time after drawing.

Culture used to test
bacteriolysis.

Number of
bacteria
introduced
per 1 c. c.

Number of
bacteria
per 1 c. c.
after 1 day.

1905.
May 29
June 3
May 29
June 3
May 29
June 3

3 hours

F. 26...

1,740
1,220
2,320
1,060
1,920
700

4,480
1,180

960
1,100

5 days

3 hours ..

Crawford

5 days

3 hours

G. P 4692

5 davs ;

At the ninth drawing from hog 1742, as Table VI shows, the serum
retained its bactericidal potency for the Crawford culture apparently
unabated for five days, whereas for the other two organisms there was
apparently more or less inhibition of growth, perhaps, both at the start
and after five days, but no very marked bacteriolysis for either of
them.

I'll

i BIOL1 I ic

"F lfl.>"|. -i 1:1 M

I'MII K VII. /.'./.7, -riitli/tn- 'i.Vi/.M ,,f i, HI-UK il liny *'nnn nfl, r 'l'ii,,l,i,. : //,*

tun, in tin r<_lriii>rnti,r. ///<*.</ ,,i tfi, tail, hull, >lnt 1749.

lt.

Jillir

7
N

9
19
1

7
1

19
J>
1

9
19

1 mi.' nft.-r .Iriiwin.-

cultiitv IIM-.I i.. I.-M
ttacterloljrvla.

Nun.
: n*
Inir..
,-,-r 1 . (

Nilin'
Iwi t.
iM-r 1

.lil.T

I hour...
1 -U\ ....

! I

/ .
M.760

M
H

700
Ml
' 1

4.430

2,000

- - . .

...do...

; H

11 .ln>-

: M

! Imii'r

Crawford

1 ,lnv

N 200

...do ..

7 440

:'. -lax -

...do...

lltln\

1 tin

i hour

u. I 1

2 davn

'. "l:l\

1 'l;tV.

.1..

' -V.

11 ilitVK

oTh^ i-liiinn-t.-r -itrnitii- tlmt tlu-r<- w-rt' too many roloniiii (.. iinint.

>iTiun from tho l)lood at the truth drawing from hor ITli'. :i-
Table VII sliows. prcsci vrd it> hactcricidiil powt-r to a marked degree
for tlu Crawford culture uj) to and including the >rcond day of stand-
ing. It appeal^ to liave IMTM >oiuc\vhat more potmt on tho fourth
than on the third day of ^tandiiiir. Imt tho dinVrenee in potency of the
^eruin on the-e t\\<> days i> hardly >uliicient. perhaps, to IK- of any
great signiricance. For the other two strains there was at most inhi-
bition shown in some of the tots. lut no decided baeterioly-U either
before or after standing, except perhaps in the case of the F. L''> -train
after t\.<> day-. But in the case of this organism the serum appeared
to have been more |M>tent after -landing for three da\ - than it \\a-
previou- to thi> time.

TMII.K VIII. -- Itiiilrrinhftir ,i,-lii, i, <,j normal IK><J trrutn aflrr ulnitilhuj t'r utriinin length*
"f tini' in tin- rifri'firntor. liliMxl ilrnnn fruin tin tail. I'-n nlli ilrnn imf. /// \.

I'M

Tlm nflrr drawing.

('ilium- ii^.-<l in t.-t
bactcrlolyd*.

Sui In-rol

Hilr."ltn . -I
pt-r 1 r <

Niini>N-rof
bacteria

p<-r I
af l.-r 1 day.

1905.

Jinn- U

| hour . ...

1 |

2.SMO

. .

1,000

1 INI

4. MO

in .1.1 \ -

., ..

i hour

('rnwlord. ..

400

I Ml

2*0

1.440

32C

.V> 400

' I' Ifl92 ..

800

do.

1 71"

, ..

iiarmctcr iii(nlRe> that then- were too many >l<>m<- lo count

COMPARISON OF BACTERICIDAL POTENCY OF SERUM.

The serum from the blood of hog 1742 at the eleventh drawing, as
shown in Table VIII, had marked bactericidal power for all three
strains at the start, and this was retained for as long as seven days for
the Crawford culture, whereas there was inhibition only for F. 26
after four days, and after this there was not even inhibition for this
organism. The effect upon the G. P. 4692 culture was similar to that
upon the F. 26 strain.

TABLE IX. Bacteriolytic action of normal hog serum after standing for various lengths
of time in the refrigerator. Blood drawn from the jugular rein, twelfth drawing. Hog
No. 1742.

Date.

Time after drawing.

Culture used to test
bacteriolysis.

Number of
bacteria
introduced
per 1 c.c.

Number of
bacteria
per 1 c.c.
after 1 day.

1905.
June 22

i hour

Crawford

4 320

1 (lav

. do

4,060

i hour

F. 26

3,380

1 dav

640

i hour . .

G. P. 4692.

1,240

1 day

340

a This character signifies that there were too many colonies to count.

The serum from the blood of hog 1742 at the twelfth drawing (Table
IX) was taken from the jugular vein, and it will be observed from
the table that the serum showed no bacteriolytic power either at the
start or on standing. In order to test whether the lack of bactericidal
power was due in this case to the fact that the blood was drawn from
the vein, experiments, which are hereinafter given (page 22), were
made to compare the bactericidal power of arterial blood serum with
that of venous blood serum.

COMPARISON OF THE BACTERICIDAL POTENCY OF THE
SERUM FOR DIFFERENT STRAINS OF BACTERIA.

SAME

If in the foregoing tables comparison is made between the results
obtained with the same drawing of blood upon the various organisms,
it will be seen that quite decided differences appear.

Selecting the most striking contrasts, in one case there was a decided
increase of the B. choleras, suis, culture G. P. 4692, while in the same
blood 7,000 B. coli per 1 c. c. were all destroyed. It is true that in
this case a great many more of B. choleras suis were introduced at the
start than of B. coli, and this may have influenced the result some-
what. But in another case where approximately the same number of
organisms of the two kinds were introduced at the start there were
veiy few of the B. choleras, suis, if any, destroyed, whereas nearly all
of the B. coli were destroyed. In still another case there was a very
great increase of B. choleras suis, while all of the B. coli were de-
stroyed. In one case there was no notable increase of the B. choleras
suis. while there was a decrease of the B. coli.

HACTKKH'I N I h POWKB OF HI..H,|. s|KI\| Ql BOOS.

It is e\ ident, therefore. that tin- bactericidal pouer of hog >erum is
different for tins strain of />'. ,!,,,/, n .-//. and for the II. ,,/;, \\\ go
far as this rim In- determined l>\ tin- method- :it pre-.-nt in use.

THE POTENCY OF SERUM FROM ARTERIAL BLOOD COMPARED
WITH THAT FROM VENOUS BLOOD.

As hereinlM'fon- staled. thr following experiment! were undertaken

because of iln- iv-ull- .ilit.-iin.-d \\ith tin- -.-rum fnun tin- \en,.ii- M ..... 1
in tin experiment in which it appeared that there was neither inhibi-
tory nor bactericidal power in the serum. In the experiments pre-
vious to this one the blood was drawn from the tail, and was
con-et|uently mainly arterial. It spurted from a severe. 1 artery when
the tail was cut. In the experiments at present under consideration
the blood was drawn from the carotid artery or from the jugular vein
either at the same operation or on different occasion-.

TABLK X. Itacteriolytic action of arterial nrrmn i-<n}M\rl irlth tlml of i-faou* *mtm,
Ixilli x,-nt I'roin tit,' K'lint- inirninl luxj (\<>. 17.1.1) nt the Mine operation. T'nll, .//-./
nf liltMnl. Titt* inmle at intt'ri'al a

Time- after drawing.

Culture u..l in ti-t
tiactcriulysls.

NtlllllMTI'f

ItwctcriH
intnxtiioMl
perlc.t-.

NlllIllNTl

|HT 1

1 .1,,J.

AM. -rial
M-niin.

>( bacteria
f. c. after

\'.

Hma

1 ilav

F. 26...

\,m

m
IM

1,480
420
. II-
9. IWO
900

60
1A.320
2,200
40
200
140

lino

100

,,.

80
120
40
6.820

, ,

140

J <\H\*

<\<>.

do. . .

1 (lav

Crawfonl . . .

J iln\

...do...

1 .lav
J duv

0. F.46W
...do...

9 djn

The serum from hog 17H3 at the tenth drawing, as shown in Table
X, was taken at the same operation from the carotid artery and from
the jugular vein. In the t<--N with the Crawford culture the serum
from the arterial, as well as that from the venous blood, seem both to
have |>o-^r>M-d considerable jx>tenry, and to have ] I it in about

equal degree. They furthermore appear to have retained their
potency, much or all of it. for nine days. For (i. P. 4'.'.'i > culture
neither serum appears to have had strong bactericidal pouer in any
They both -eem to have had some inhibitory jxjwer for thi^
organism both at the start and after nine days. For culture F. 26
there appear- to have leen no difference between the two kind-
serum after having stood for one day, but the serum from the venou-
bloo.l would seem to have retained it- potency upon -landing more
tenacioii-lv than that from the arterial blood.

ARTERIAL AND VENOUS BLOOD SERUM COMPARED.

TABLE XI. Baeteriolytic action of arterial strum compared with that of renous serum,
both Kern from the same normal hog (No. 1733) at the game operation. Eleventh draw-
ing of blood. Texts made at intervals after drawing.

Time after drawing.

Culture used to test
bacteriolysis.

Number of
bacteria
introduced
per 1 c. K.

Number of bacteria
per 1 c. c. after
1 day.

Arterial
serum.

Venous
serum.

1 dav

F. 26...

5,420
6,520

10,480
20,160
280
10,960
160
400
20,640
7,280
2, 520
2,960
21,240

220
27,680
8,960
2,600
10,860
360
580
60
3,300
3,340
1,360
2,320

2 days

r> davs

...do...

8,580
240
8,160
8,660
26,240
1,920
2,160
2,720
a oo
2,240

8 davs

1 day

Crawford

2 days

...do...

5 davs

8 days

. ..do

Idav

G. P. 4692

2 days

5 days

8 davs

"This character signifies that there were too many colonies to count.

The blood obtained at the eleventh drawing from hog 1733, as shown
in Table XI, was drawn from the artery and from the vein at the
same operation. There seems to have been a difference in the potency
between the two kinds of serum for all three organisms. For the
Crawford culture the serum from the arterial blood seems to have
been quite strongly bactericidal at the start, but the serum from the
venous blood was apparently not so. The tests made after the serum
had stood for two days would indicate that both kinds of serum were
potent. The serum from the venous blood in this case seems to have
gained in potency upon standing, a result previously noted in the
case of culture F. 26 with a different serum, but it seems necessary to
conclude that an error of some kind must have crept in where such
results are obtained, since they are at variance not only with the
a priori probabilities of the case but also with the usual experience.
On the whole, the results in this experiment seem to indicate for the
Crawford culture a greater potency, or at least a retention of a greater
amount of potency, on the part of the venous than of the arterial blood.

For culture F. 26 in this experiment the arterial serum appeared less
potent at the start and after eight days than the venous serum, but
after standing five days it was more potent than the venous serum.
Here again the serum seems to have gained potency on standing.

For G. P. 4692 the stronger retention of potency of the venous
serum as compared with the arterial is apparent in the tests made five
days and eight days after drawing. No difference between the two
sorts of serum is noticeable in the tests made with this organism one
day and two days after drawing.

The fact that blood through which a stream of CO 2 was conducted
becomes more alkaline led Hamburger 9 to make tests of the compara-
tive bactericidal potency of serum from ordinary blood and that from

i RIOL) 'I. r>\\ IK OP MI. -I 1:1 M "I

Mood llin.nyli \\ In. h ( ( li:i.l In-.-n pa-xi-d. M- \\.-ll a- !' the difference

in tlii- rexp-ci Itetueen arterial and venou- Uood. Mr found. in fa.-i.
i h;ii serum from \enou> |I|IMH| and from l>l....:l through which CO, had
(Missed is more potent than UterbU-blood serum.

EFFECTS OF HEAT UPON THE BACTERICIDAL POWER OF HOO

SERUM.

A large number of (.>!- \\.-re made parallel with those above recorded
to determine the effect of heat upon the gerwicidal power of hog
serum, mid also to determine whether there i- anv difference lct\\ < -n
the potency of the heated -crum for the different strains of It. cholerx
SHIM upon which it was tested.

The method employed in making the te-K consisted in immer-inir
the tubes containing the serum up to the plugs in a water Imth which
\\a^ kept by means of a thermoregulator exactly at the temperature to
be tested. The results are shown in the following tables, from whieh
it will be seen that the serum was subjected to temperatures of 51,
52, 53, and 54 C. for ten minutes and for thirty minutes at different
times. Tests were also made at temperatures al>ove 54 C.. but these
are not tabulated for the reason that they gave results similar to those
at 54.

TABI.R XII. Effect* <>f hi-nt n/xm the barlirinhilli- jtowtr of normal hog terum. Blood
drawn from ll toil mul /.</./ in tin refrigerator for varion* length* of time. Temper-
ature ftnjtloiffil, :,i C.

Date.

Titm- nftT

ilruwiiii:.

N". i.f hiv
and drawiiiK
of bl.xxl.

Culture iix-<l

til !

tiactfriolysi*.

Number of

tiarlcrin
intr.Mlurtil
IKT 1 . .

Arterial blood.

l^'iiirth of
expxiirc to
ht-nt.

Numberr

[*Tl <..-. n

N..t
h.-at.-il.

fbnct-ri
ftiT lilny.

M.-nl.-.l.

1*K
April 7

June 12

:t li-uir*...
.1..
i liniir
|.lnv ....
T.IMV- ....
I."l\- ..

J li<>iir
I >\n\~ ....

10 <1 >...

; liniir

T .lm

17881...

<!<>
ITU XI
.lo
do
... .do
do
do
do

.. I- MB...

F. a;
do

d.. :..

( 'raw ford

4.GX

I'. H-'M)

l.UUU
4.040

1,860

t,m

1 IKI

.. i..,

16. 1'JO
1.740
8. WO

TC.flOO
30. KX>

I.IK)

M,MO

400

:NI

800
1.2HO

,, nog
,, no

1.100

<l x

ao
440
1. Ill)
4,200

11.200
I.K40
2K.MO

10 minute*.

1)0.

90 minutes.
Do.

i...
Da
Do.
Do.
Do.

Da
Da

-. i- u;s
.1..
.... .1..

Thi huni. i. r -ii:niii.- that there were U many <-il..ni-> io . ..nut

It would appear from the results recorded in Table XII that heat-
ing for ten minutes at 51 C. produced no etfert in the two tests which
were made in this manner, one on the G. P. 4692 strain of B. chol- >:
.>"/>. and the other on the colon culture. The same temperature for
thirty minutes robbed the serum of much or all of its power in 8 of
the 11 tests.

EFFECTS OF HEAT UPON BACTERICIDAL POWER. 25

TABLE XIII. Effects of heat upon the bacteriolytic power of normal hog serum. Blood
drawn from the tail. Temperature employed, 52 C.

Date.

Time
after
draw-
ing.

-9

No. of
hog and
drawing
of blood.

Culture
used to test
bacterio-
lysis.

Num-
ber of
bac-
teria
intro-
duced
per
1 c. c.

Number of bacteria per 1 c. c.
after 1 day.

Length of
exposure
to heat.

Arterial blood.

Venous blood.

Not
heated.

Heated.

Not
heated.

Heated.

1905.
June 28

July 12

1 day . . .
2 days . .

1733 X..
do ....

F. 26
...do...

3,880
4,920
380
940
1,480
420
2,440
9,880
300
5, 420
6,520
2, 160
2,720

60
16, 320
2,200
40
200
140
9,760
8,160
100
10, 480
20, 160
2,620
2,960

1,540
15, 200
1,720
3,960
460
300
1,920
9,840
100

320
600
400
80
120
40
5,920
6,560
140
220
27, C80
3,300
3,340

10,720
3,360

a 30
25, 080

3,080
10,000

30 minutes.
Do.
Do.
Do.
Do.
Do.
Do.
Do.

Do.
Do.
Do.

9 days

1 day

do .

Crawford

2 days . .

do....

...do...

9 days . .

...do...

1 day

.do

G. P. 4692...

2 davs

do....

...do...

. do

1 day . . .
2 days . .
1 day

1733 XI ...
do....
do....

F. 26
do
G. P. 4692...

5,680
41,720
2, 240

(?)

14,880
2.440
2,180

2 days '

...do

"This character signifies that there were too many colonies to count.

It will be seen from the above table that the effect of an exposure
of thirty minutes showed a marked influence in some cases while in
others it was apparently without effect. In these experiments the
blood was drawn at the same operation from the carotid artery and
from the jugular vein, and it would appear that the serum from the
venous blood was somewhat more sensitive to the action of the heat
than the serum from the arterial blood, if any conclusions are per-
missible from the somewhat limited number of observations. In the
arterial there was in only one case a marked difference between the
potency of the heated and of the unheated serum out of the 12 cases
in which it was tried. In the serum from the venous blood, on the
other hand, there was a marked difference between the heated and the
unheated serum in the 11 te^ts which were made. On the whole the
effect of heating at 52 C. would appear to be uncertain; sometimes
heating seems to have a marked effect upon the bactericidal power of
the serum; at others it seems to lack this effect.

19500 No. 9507 4

BA< rKKlOLYTIC I'oXVKK <>K HI. -MMM "! HOGS.

T \IIIK XIV. Kipi-ct* a f hull III>H tin Inn-It riot (///< fxmrr f iu>rnuil )nj *nnn.
ilrnirii I'niiit lltr tail. 1'i-iii/n-riitnri- nni>l'ir<l, .',.1 ('.

/;/,./

Time ftT
drawing.

NO. Of In < " lid

drn liik- (

blood!

Cotton

llwll
burl.

U-i-

Nunilx-r
of l>ai-(i-Ma

Intro

Sum
hartrrU

nfli-r

Sol

\' "'

XT 1 C. r.

1 da>

l.i-IIL-t '

n- to

190ft.

\|T I 1 .'

i : . . v

U. P. 4092...

- , . ,

i , ...

; . .-

10 nn i

J davs

ll.roli

', r IMW

i ' -

; .

..: -i n
1 ...

-"'

<|,i

H riili

I 940

1.800

Mnv

;i hittip* .

1 1

::. i. -

17 M

H.eoli
<i T. 4fi92

i m

I 004

9.920

IS 720

17401

H. ro/i
1-.92

975

.'. 04
1,750

,.,.

Do.

1740 II

'1746111!!!

171" IV

<;. i'. !

H.oJi

i, I'. !

i MB

1.010
3.110

70
1 ] |
i M
994

1,900

. '
1 1 .

Do.

Do.
Do.

|l<>

11
12
12

do
do

1740 VI

1711 1

<;. r ii

^. P. 4692

1,700

450

1,840

.' -l

15,900
4 1 II

Do
Do
Do.

...do..

. 1T ",',!::::::

17M III

I'.'.'J . .
li. -1,
G. 4692

3,778
102
1,970
1,010

390
|,M

1 ! I
910
1.2X0

i :>

Do.

. . .do . . .

3,110

11
12

...do...

1741 IV
1711 V

;.. 4692...
Hi i

Ml
1,700

3.286
1.1M1

Do
Do,

17 1'. 1 I . .

; i (092

4 925

Do.

I r

1712 II

. 1'.

j rrrf.

1 970

7:t4

U
U

do
do

rum.

do
171.' IV

i 1'. ?J2

i. r i

1,010

:l. no

798
30
50

Mil

1.758

Do.

Julv 12

i VlHV ""III!

1742V
173:1X1 ' .. .

It. f,tli

r H

1,700

5 420

10.480

1,181

Do.

1 " minute*.

12
13

-'day-
1 day
J diiv
1 dy

do
do
do
do
do

do
Crawford

.... do

(J. p. 1-

2,100

20 160

400
2,960

H -"

2! 440
2.180

Do,
Do
Do.

1 dav ...

17: XI'-..

I M

5. 120

13
12
13

'.'lav*
1 da\

Ida? .'.'.

2 ilny-

do
do
do
...do
do

r raw lord
.... do
(!. P. 469 1 .?

6.520

10. MO

.. . M

41.720
2R.160

(T)

Do
Do
Do

"This i-luirartcr *i>riiit:rs (lint tln-n- \\-rr hm iimny t-nlonio in rouiil.
* Arteiial I>I<>1 M-riiin.

' Venous M.H| MTtlll).

A- -lin\vii Ity the results recorded in Tahlr XH". lie:itin<: at
had tlie ftlect in most casos of jrreutly diminishing the liitctericidal
power of the -enmi. though this is not to lie noted in all <-:i-e-. e\-.-n
\\here the xeriim \\:i- exposed for thirty minute^.

EFFECTS OF HEAT UPON BACTERICIDAL POWER.

TABLE XV. Effects of heat upon the bac'eriolytic power of normal liog serum,
drawn from the tail. Temperature employed, 54 C.

Blood

Date.

Time after
drawing.

No. of hog and
drawing of
blood.

Culture used
to test
bacteriolysis.

Number of
bacteria in-
troduced
per 1 c. c.

Number of bacteria
perlc. c. after 1 day.

Length of
exposure to
heat.

Not
heated.

Heated.

May 26
June 2
May 26
June 2
May 29
June 8
May 2'J
June 3
May 29
June 3
5
6
7
8
9
19
5
6
7
8
9
19
5
6
7
8
9
19
May 23
23
23
24
24
24

3 hours

1742 VIII..
do
...do...

Crawford . . .
do
F. 26

320
1,960
80
3, 280
1,740
1,220
2, 320
IOM
1,920
700
2,K40
2,240
3,440
2,200
1.360
4,040
8,380
5,200
7,440
3,500
2,640
1,440
1,380
3.C40
180
360
l.SSO
1,740
10, 8SO
40,500
19,040
1,682
1,808
12,040

380
1,960
2,980
4,480
1,180

960
1,100
26, 640
51,760
900
1,340
18, 280
a oo
180
120
760
3,160
1,380
oo
4,420
6, 320
3, 420
2,000

3.376
2,640
18,280

180
1,640

1,160
2,380
5,280
9, 360
13.600
1,180
'., 520
1,880
5,040
1,260
119,700
41,840
6,440
7,280
20,480
oo ,
28,000
8, 320
13,000
4S,000

00
00

4,820
27, 760

17,960

13,180
58,800

2, 720
7,600
26, 320

10 minutes.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
Do.
30 minutes.
Do.
Do.
Do.
Do.
Do.

7 days
3 hours

7 davs

. do

do . ..

3 hours

1742 TX

5 days

do .

3 hours

Crawford
do

5 days

...do

3 hou rs

G. P. 4692.

f> davs

J hour

1742 X .

F. 26

1 dav

2 days

...do...

3 days

do...

4 days
14 davs

do..
...do ...

do
do

i hour

do.

Crawford
do

1 dav do

2 days .do

...do ...

3 davs

.. .do .

.. do

4 days

. do..

14 days do

hour do
1 dav do

G. P. 4692
do

2 days do

3 davs . ' do . . .

.do . .

4 days
14 days

...do ...

do
...do...

. .do 1742 VI

do do

do ' dn

Crawford
F. 26

. ...do.

1742 VII

G. P. 4692

do
....do...

do
do

Crawford
F. 26

a This character signifies that there were too many colonies to count.

As is evident from the preceding table the effect of exposure for ten
minutes at 54 C., as well as exposure for thirty minutes at the same
temperature, is practically in all cases to weaken or suspend the
bacteriolytic power of the serum. The few exceptions to be seen in
the table are without significance in view of the many cases in which
the serum lost in power.

In summing up the results of heat upon the bacteriolytic power of
normal hog serum it would seem evident that temperatures below 54
C. are uncertain in action for the lengths of time employed, but that
54 C. even for ten minutes serves quite uniformly to suspend or at
least to weaken the bacteriolytic power of the blood serum. A num-
ber of tests were made at 55 C. and 56 C., but as these did not show
any results that were not to be expected from those at 54 C. it would
seem superfluous to give them in detail.

EFFECTS OF INJECTING CULTURES OF B. CHOLEILffi SUIS UPON
THE BACTERIOLYTIC POWER OF HOG SERUM FOR THIS OR-
GANISM.

Very conflicting results have been obtained by various investiga-
tors in regard to the effects produced upon the bacteriolytic power of
blood serum by the injection of animals with cultures of bacteria.
Some have failed to note any effect of such injections, either to in-

BAOTJCBIOLTTIC

-M:IM

i ..... -.

civa-e ..r .Iiinini-h till' haeteriolytic potency; other- ha\e found that

lli> potcncN was diminished, others again that it wa- increased. I'n

der the supposition that this discrepancy in iv-ult- was due pn-hap-

to tin- te-ts having been made at different interval- of time after in
jeetion. and that the serum taken from th- -aim- animal at dillen nt
period- might exhibit dihVrence- in potenc\. <.r that it might at
tain times he devoid of power, te-t- \\eiv made of -erum taken at

various intervals after OM or more injection-. The re-uli- < ( f these

experiment- are -nmmari/ed in the following table-.

TMII.K \\'l. H'l'trrioli/tii- ftoinr of xt-runt fruiii tin <iin: // ( .\<>. 17.', 1 1 I,. !,,,-> nn<l
ilt,r iiijti-ti'ni intli II. rlin/,-r,i xnin, ,-nlt'

Ciiltur. H-..-.! in i.-i
IwirliTiiilyM-.

Time lil.MMl \Mi-ilniwn r.-lali\. (.. inj.-i-in.n.

NiiniU-r ..f
lia.-tcriH in

triMlllcell

IHT 1

NniiilH-r <if

iMleterill |T
1 < e. llfl.T

1 d.l>.

Cnif..r.l

jilt.-r lir-t in jit -ti< >M

1".

K. ji;

! iln\ > ai'i.T s.'coti.l injection
loilnv- ai'i.T third injortiim
B.'forc injii-tinii

6.200

IL.

:i '1 n - aft.T lir>t inji-rti.ni

11 800

'. il -i\ - afl.T MTiiinl i ii j 11 -lii Hi

: MM

'2 &W

in i!av- after Hi inl inject inn

fi,200

| j .

h,.

ltd on- inject inn
:! .In \ x after tirvt injection

1,060

tu P..

12 WO

]>,,

'.' i|n\ ~ alter NOOOO inject inn

1 nJO

Di.

10 i\n\- after third injection

4,100

16,700

The results uiven in Table XVI were obtained with the serum from
a hojr which was injected nt various intervals suhcutaneou-lv \\itha
In-ef-broth suspension from an arar culture of the (i. 1*. ''.'_' -tmin.

The lir-t injection was made with 5 c. c. of a su-pen-ion e(pii\alent
in den.-ity to a twenty-four hour typhoid culture diluted about one-
half. The second injection was made with 10 c. c. of a sii-jH-n-ion of
about the same strength. The third injection was made with a -u-
pen-ion con-isting of the entire growth in twenty-four hours of an
a gar culture. Before each injection the blood was drawn from the tail
of the animal and placed on ice for three days. After thi- the -erum
\\a- <li awn otf from the blood and tested on the three -trains in all
cases except the serum. from the blood taken at the tir-t drawing, which
\\a- tested only upon the culture with which the animal was subse-
quently injected culture (J. I*. 4M2.

As there were no test- made with culture Crawford before the
animal was injected, it is impossible to be SUIT whether it \\a- bac-
tericidal for thi- -train at this time or not, but judging from its act ion
on(i. T. H'.!'J before injection it would seem proluihle that it would
have been bactericidal or. at least inhibitory for culture Crawford,
since the latter strain was quite uniformly more sen-it iv- to the
germicidal action of serum in other experiments. The -erum after
the injections possessed quit^ strong l>acteriolytic properties for cul-
ture Crawford.

INJECTIONS OF CULTURES OF B. CHOLERA 9UI8.

It is also possible only to surmise from the results with G. P. 4692
that the serum probably had at least inhibitory power for F. 26 before
the animal was injected. It seems to have retained its inhibitory
power after the first injections for this organism, but to have lost much
in potency after the third injection. With culture G. P. 4692 the
results seem to be more or less in accord with those obtained with F.
26, though it is true that three days after the first injection the serum
exhibited very strong bactericidal power for G. P. 4692 stronger
than for F. 26. Still the results can not be compared too closely in
these two cases, since the number of bacteria introduced differed
widely.

TABLE XVII. Bacteriolytic power of serum from the same hog ( No. 1755} before and after
injection with B. cholerse suis, culture F. 26.

Culture used to test
bacteriolysis.

Time blood was drawn relative to injection.

Number of
bacteria
introduced
per 1 c. c.

Number of
bacteria per
1 c. c. after
1 day.

Crawford

Before injection .".

3 days after first injection

2,180

9 days after second injection. ......

240

10 davs after third injection

5,200

F.26

Before injection
3 days after first injection

800
16,700

47,400
380

9 days after second injection

7,600

11,880

10 days after third injection

8,200

660

G. P. 4692 . . .

Before injection

3 days after first injection

38,400

9,800

9 davs after second injection

6,340

3,800

10 days after third injection

4,100

11,200

The results of the experiments with the serum of hog 1755 are
recorded in Table XVII. In this experiment the injections were made
and the tests applied precisely as with hog 1754, except that culture
F. 26 was used for injecting the animal instead of culture G. P. 4692.
The concentration and amounts of culture were the same in both experi-
ments, and the methods of procedure otherwise were the same. The
serum from the hog before injection was tested upon the F. 26 strain
only; after injection it was tested with all three.

For the Crawford culture, as already stated, there was no test of the
serum made before the injection of the animal. The tests made with
this organism after each of the th'ree injections showed that the serum
possessed considerable potency, and there seems no evidence that the
injections were followed by diminution of potency for this organism.

For F. 26 culture the serum seems to have had but little potency to
start with before injection, but it seems rather to have gained than
lost in potency for this organism after injection.

For G. P. 4692 culture the results of the three tests are somewhat
difficult of comparison, perhaps, but the serum after the third injection
would appear to have been somewhat less potent than that after the
second, and, also, after the first, if the difference in the initial doses
in these two cases may be disregarded as a factor in determining the
results.

HAiTKRInLYTH 1 I'oWKR 'K HI.ooH

<! I!

Tvlll.K XVIII. /.''. ' , r ,,f i-niiii from tin- IMIIIII- lioij ( .\'<t.

utter f/i/Vr/iiiii ii , tit />'. f/u*/rr:r i>ni/>, riill "/.

Ciilmr.- tiM>d U> IWt

baetoriolyrii.

Time blood WM drawn n i.m\.- in mj.-.-n,,ii.

Nutiitx-r <>(
linrtrrin
tinr--'
per 1

\nnii

il |MT

ulii-r

I'niw lunl . .

llcliiro injection

TOO

:illiT tirM llijrrlloli

Mill r -i mini iiiji-< tii.n

I'"
I'l"

III ililN" lllIiT tlllnl lllJiTtlnll

Ih-forv inji-rl inn

after lirl mji-< linn

M "">'

.ifd-r M-.-IHI.I Injection
In iln\ uftcr tliinl Injection

; m
8,200

( , | \(Q> . . .

Kcfnre injection

3 I|HVH niter llrt injection

iij jjj

&.MO

tt <1a>' BfliT M'cnllil injection

. Ml

10 dv ufti-r thinl injection

4,100

- ...

[n the experiment with tin- srnun of the Mood of ho^ 17.'.<;. the
n-.nlt> <>f which :irc ^ivrn in Tahlr XVIII. the method of procedure
u:i- the same ;i- with t he >rruiM of hoyfs 1 7. VI and 1755, just de-i-rilied.
the only difference IMMIIJJ in the strain >t' oiyani>in employed. In the
ca>e of hog 175G the Crawford Culture WM u>ed. and the .-ernin \\a-
tested upon this organism only before injection and on all three
-train> after the injections.

With the Crawford culture the serum seems to have been unaf-
fected ly the injections of the hog. It seems, at any rate, not to Inivc
lt in potency, and from the result after the second injection, if any
conclusion is justifiable, it seems possibly to have gained.

With F. _'; it >eem- to have lo>t in potency, if the result obtained
after the tirst injection is compared with those after the other two.

With (i. 1*. 4t>'.'2 culture the serum seems to have lost in potency
after the third injection, but it is uncertain whether it gained or l"-t
after the lir>t two.

TABLE XIX. llnrlrriiih/tit' f.ovrr of neritm from thf mint hoq ( A". 1798) brfarr <nul
after mju-tinn u-ith H. chdltTH unit, culture F. 6.

Culture nvil
bacterioljmta.

Time blood wn drawn relative to injection.

NiiinlM-r of
bacteria
introduced
per 1

Number of
iMdi-ria |K-r
1 . . c. after
Iday.

f ^

IM reinjeetion

, 1

l.MO

Do ..

in iln\" after tlrM injection

. s

7 -JUO

9 dayx after <-coii'l injection

, ,-,,

. ..

1 1 iliivu after ihlnl inject inn

., ..,

In the experiment with the >erum of hog 1798, Table XIX. the
animal wa^ given three intravenous injections of suspensions in beef
broth from agar cultures of the F. '2K strain. The tirst one was made
with a very dilute -u>pension. weaker than would correspond with a
twenty-four hour In-ef-broth typhoid culture, and the other two injec-
tion-, were made with -imilar. >ii>|>cn>Mons.

It would appear from the results given in the table that the serum
w.i- not actively bactericidal before the injection of the animal, but
it seems to have had some inhibitory power. Ten days after the tirat

INJECTIONS OF CULTURES OF B. CHOLERA SUIS.

injection there seems to have been no striking change. The serum
was still inhibitory, but was neither markedly bactericidal nor can it
be said to have been devoid of potency. But nine days after the
second injection, and fourteen days after the third injection, the serum
seems' to have lost greatly in the inhibitory power which it possessed,
it would appear that in this case the injections had the effect of ulti-
mately diminishing the power of the serum.

TABLE XX. Bacteriolytic power of serum from the same hog (No. 1809) before and after
injection with B. choleric suis, culture G. P. 4692.

Culture used to test
bacteriolysis.

Time blood was drawn relative to injection.

Number of
bacteria
introduced
per 1 c. c.

Number of
bacteria per
1 c. c. after
1 day.

F. 26...

Before injection

5,920

5,900

14 davs after injection ...

7,300

1,800

Crawford .

Before injection . .

5,700

14 davs after injection

13,000

56,600

G. P. 4692

Before injection

1,140

10.9CO

14 days after injection .

7,300

7,900

In the experiment given in Table XX the serum of hog 1809 was
tested on the three different strains of B. choleras suis before the animal
was injected and on the same cultures fourteen days after the hog had
received subcutaneously about 5 c. c. of a dilute suspension from an
agar culture of culture G. P. 4692. In the tests made with the serum
upon the F. 26 strain there seems to have been a slight diminution of
potency after the inoculation, but the difference in the numbers of
bacteria surviving bacteriolysis in the serum from the hog before
injection and those surviving after injection is not large enough to
permit of very definite conclusions. The same may be said in regard
to the organism with which the hog was injected culture G. P. 4692
but the results in this case are perhaps somewhat more significant.
It would seem justifiable, however, to conclude that there was no
marked increase nor decrease of bacteriolytic power for F. 26 culture
or for G. P. 4692 produced by the injection. The result with Craw-
ford seems to show plainly that the serum was less potent for this
organism after injection than it was before.

TABLE XXI. Bacteriolytic power of serum from the same hog (No. 1836) before and
after injection ivith B. cholera suis, culture F. 26.

Culture used to test
bacteriolysis.

Time blood was drawn relative to injection.

Number of
bacteria
introduced
per 1 c. c.

Number of
bacteria per
1 c. c. after
1 day.

Crawford

Before injection

32 200

a 3 840

1 hour after injection

32, 200

a 5, 340

3 davs after injection

1 560

F. 26

Before injection .. . .

30,400

a 247 800

1 hour after injection

30,400

" x

Do ..

3 davs after injection

5 000

G. P. 4692

Before injection

35 000

a 47 4CO

1 hour after injection

35,000

a 36, 600

3 days after injection .

6 800

a Serum diluted in the proportion of 1.4.

6 This character signifies that there were too many colonies to count.

HACTI i:i"i.i ii< i'"\\ i i; 01 BLOOD -i-:i:rjf OF li>

In tln experiment with the >erum from tin- l>l>o<l of hog 1830, given
in Table XXI, the animal \\a- injected int ravenously with a -u-;
-ion from an a gar culture of -train !'. L'I', of about the density of a
:M-hour l>eef-broth culture of the typhoid bacillus. Before injection
the hog was hied to obtain -enim for testing the power of the
normal serum. The amount of blood obtained before injection and
that obtained one hour after injection wa> in-utlicient for the test to
be made with the- quantity of serum usually employed, so the -enim
was diluted in both of tlie-e ea-es with three pan- of physiological
salt solution to one part of serum. There were no tests made with
the undiluted -rum.

With the Crawford culture the serum seems to have been quite
potent before injection, and also one hour after injection, but three
days after injection it appears in have been entirely devoid of bacter-
iolytie power for this organism.

With F. _' -the strain used to inject the animal the serum had
only weak potency before injection, and if there was any effect pro-
duced by the injection it was in the direction of weakening the potency
of the serum for this organism. This apparent diminution of potency
i> to be noted even in one hour after the injection, and is still more
pronounced after three days.

With the (i. P. 45t-J strain the serum was apparently inhibitory at
the start before injection and one hour afterwards, but three days
after injection it seems to have lost even the power to inhibit the
growth of the organism.

The results of this experiment seem to show that in this case the
effect of injecting the hog was to suspend the bactericidal power of
the serum for the homologous" organism a- well a- for the other
-trains upon which it was tested.

TABLE XXII. Baekrtobftfc powtr <>f serum from the Mine hog (,V. 18,59) before and
after injection icith B. cholerx suit, culture (i, P. 4692.

Culture used to test
bacteriolysis.

Time blood was drawn relative to injection.

NiimlM-r "f
bacteria in-
troduced
per 1 c. c.

Number of
bacteria per
1 c. c. after
Iday.

Before HIM rt ion

3 760

Do...

*J hour* after injection ................. . .

.. TI'.I

' i.,.

1 day alter injection ...................... ....

i...

M i"'

F. 26. . .

- niter Injection

600

\ .'..

I-.,.,,
.i

'i hours alter injection .

1 020

DO :.:::...::

1 day alt-r injection
2 da v<< after injection

840

... ...

.' .

O. P. 46M...

2,840

. -v.i

'_' Hour* niter injection

| |.i

9.200

Dor...

1 lav niter inject ion

200

:.j ....

2 day* alter injection . ...

200

1 i"'

Homnln((ou i the term now universally employe*) in denote the organism with which the
animal in any given case Is Injected, aa In the present cmae.

INJECTIONS OF CULTURES OF B. CHOLERA SUIS.

In the experiment given in Table XXII the animal hog 1859 was
injected intravenously with 5 c. c. of a suspension of culture G. P.
4692. The suspension was made in beef broth from a 24-hour agar
culture as usual, and was equivalent in density to a 24-hour beef-broth
culture of the typhoid culture diluted with about equal amounts of
broth. Tests were made of the potency of the serum from the blood
of the animal before injection and after injection with all three strains
of B. cholerse suis. Owing to the fact that the animal was killed in a
moribund condition two days after injection, it was impossible to make
tests with the serum from the blood later than the second day after
injection.

It would seem hardly necessary to discuss the results with the differ-
ent organisms in detail, since they all appear to point to a lessening
of the potency one and two days after injection. Before injection the
serum from the blood seems to have been bactericidal for the Crawford
strain only; still for the other strains it was at least inhibitory, whereas
after injection it had lost much, if not all, even of its inhibitory power
on the first and second day after injection, in the tests made two
hours after injection the results appear somewhat discrepant; the
serum seems to have lost potency for Crawford, and gained for F. 26;
but, after all, the figures are perhaps not very significant in these
cases.

The results in this experiment may have been complicated by the
fact that the animal from which the blood was drawn was suffering
from acute infection with B. choleras suis at the time the blood was
obtained. It may not be correct to regard this serum as immune in
the sense in which this term is usually understood that is to say,
serum from the blood of an animal given injection of bactei 'a and re-
covered. Nevertheless, the results are perhaps not without interest,
since they show certainly a very marked difference between the serum
from the blood of the hog before injection and that from the blood
after injection.

TABLE XXIII. Bocteriolytic power of serum from the same hog (No. 1860) before and
after injection with B. cholerse suis, culture F. 26.

Culture used to test
bacteriolysis.

Time blood was drawn relative to injection.

Number of
bacteria in-
troduced per
1 c. c.

Number of
bacteria per
1 c. c. after 1
day.

Crawford
Do
Do

Before injection
1 day after injection
2 days after injection

1,460
1,240
140

a go

Do
Do

3 days after injection
4 davs after injection

5,400
1,620

F. 26 ...

Before injection

2,100

Do
Do

1 day after injection
2 days after injection

700
1 100

880

3 days after injection

1 500

4 davs after injection

2,260

5,580

Q. P. 4692
Do

Before injection

1 day after injection .

10,200
4 100

17,400
60

Do
Do
Do

2 days after injection
3 days after injection
4 davs after injection . .

1,500
9,900
1 380

1,020

a This character signifies that there were too many colonies to count.

84 HA<TKKI"|.VH I-\\I-.K <>K m.nm -i ui'M m- IKKS.

In the experiment recorded in Table \\lll. (lie animal IIOM- 1
\\lis injected fust \vilh about "J r. r. of a dilute sUs|M'l|sion of MM :ej:il

iiltuiv in beef 'broth, but most of tin- material was injected meie|\
under the >kiii. as tin- attempt to strike a vein \va> not successful.
Two lay> after the first injection a second injection of about I'..". <
of H similar .suspension from the same M r: ij,, ,,f bacterium. in this
all of the suspension, was int roduced into the vein. Two days after
the second injection a third injection was made similar in all r>--\
to the second. The sii-peiisions for all three of the injections \
heated at '>~> < '. for thirty minutes. The blood was drawn from

the animal In-fore injection and on the liiM. second, third, and fourth
daja after the last injection. The results seem to le quite uniform for
the .serum on the second and third days after the injections; on both
days the serum showed apparently entire loss of bactericidal jnwer for
all three organisms. On tl>e fourth day after the last injection the

Beruin had apparently gained in potency, particularly for the Crawford
stniin. The results obtained with the serum one day after the last
injection are not very striking, but they perhaps permit of the conclu-
sion that the serum had not lost in potency at least to any marked
extent.

AS previously stated, the results of the injection of animals with
cultures or with the products of rrowth of bacteria have been most
conflict intr in various hands in >,, far as the effect of >uch injections
upon the bacteriolytic properties of the blood serum from the animal
in test-tube experiments is concerned. 7

A numtier of investigators have found that the injection of an ani-
mal with cultures or with the products of growth of certain bacteria
increases the bactericidal power of the blood serum for the homologous
organism. Others have found that there is no effect upon the bacteri-
cidal potency of the serum produced by such injections, while still
others have found diminution or complete l<s of j>ower for the
homologous organism to follow the injections. Among others Neisser
and Wechsberg 11 have found that the so-called immune >erum i. e.,
the serum from an animal which has received injections of certain
Iwcteria no longer possesses bactericidal power for the homologous
organism when the serum is undiluted. The results with the diluted
serum will be discussed later. Huxton* has found that the serum
from the blood of a normal rabbit is capable of destroying very lar-jv
numbers of the typhoid and paratyphoid bacilli, but that after the
rabbit has been injected with these organism.- it no longer kills any of
the homologous strain.

But while there is undoubtedly conflict in regard to the matter, the
preponderance of evidence certainly seems to show that the >erum i<
frequently, if not always, weakened in bactericidal power or com-
pletely robbed of this power for the homologous organism by th -

EFFECTS OF. DILUTION UPON BAOTERIOLYTIC POWER. 35

injections, and the results recorded above seem in many instances to
bear out this result in the case of hogs injected with B. choleras, *'//'..
But it would also seem apparent that there is a tendency in some ca-f-
to a restoration of the lost potency on further injection of the animal.

These conclusions seem justifiable at least from the experiments \\ ith
the F. 26 and the G. P. 4692 strains. The results with the Crawford
strain differ somewhat in this respect from those with the other two
strains, and this may be due to the fact, already mentioned, that this
organism is of only feeble pathogenic potency. It seemed to be much
more easily killed by the serum generally than the other two strains,
and it probably also has less power than these to cause the reaction
in the serum which is necessary for the production of immune serum
as this is understood. The serum of hogs injected with this strain
would consequently perhaps be expected to behave more like the
serum from an uninoculated animal. In some cases the serum of the
blood of hogs injected with the F. 26 or with the G. P. 4692 strain
seemed to have retained bactericidal power for the Crawford strain
after having lost it for the other two.

The variations which have been already alluded to as occurring in
the serum of animals from time to time, and which are due to causes
as yet obscure, must be borne in mind in drawing conclusions from
the results of experiments with the serum from animals injected with
cultures. Of course it is impossible to determine in the present case
the extent to which this spontaneous variation, as it is at present con-
venient to call it, may have affected the results.

In view of the many conflicting results with the serum of the blood
both of normal and of injected animals, in so far as the bactericidal
properties are concerned t/ it is evident that there are some obscure fac-
tors in the process which the methods at present in use are incapa-
ble of determining. If the blood serum from one and the same animal
is found to be bactericidal at one time and not at another, where the
animal is kept under apparently the same conditions, and if, further-
more, the injection of cultures is found by certain observers to have
no effect upon the bactericidal potency of the serum, by others to
increase the potency, and by others again to lessen the potency, there
must be gome very subtle factors which enter into the problem, else
they would scarcely be overlooked by so many investigators.

EFFECTS OF DILUTION UPON THE BACTERIOLYTIC POWER OF
NORMAL AND IMMUNE SERUM.

DILUTION OF NORMAL SERUM.

The effects of dilution upon the bactericidal power of hog serum were
tested in the manner usually employed in such cases. The method for
obtaining the blood was to cut oft a piece of the tail and allow the
blood to run into sterile test tubes or flasks, which were put into the

i RIOL1 ii' r\vn:

HI..M.M -i 1:1 M

muss.

refrigerator immediatelx . Tin- >ermn wa- <li-;i\\n <>!l a- -i><>n a- i: had
-eparated. and 1 c.c. of i( \\ a ^ put into each of three tube... Yarioii-
dilution-.. a> mentioned in tin- taMe>. were al>o prepared with ph\ )0
logical salt solution, and |>ortion> of these dilution-* \\eiv di-t riluited
into te>t tube*, three tiil>r- from each dilution, each containing 1
In most casrs o. 1 or t'.o;, ,-.<-. of neutral. \ -pinni/cd U'ef broth was
added to inMiiv>ullicient nutrient material for tin- bacteria which \\en-
subsequently introduced. Tlii- \\a- done in conformity with the
method ordinarily employed in such experiment-. Kach of the tul>e>
was finally inoculated with a carefully niea-mvd amount, usually
C.C., of a -u-penxi.n in phy-iolou-i-al salt solution from a recent agar
culture. One -eiie-. each was inoculated with one of the three -train-
of 11. </!>/, r.i v//v. Inaddition to the tulx's of undiluted and of diluted
xenim. three lulu's of physiologic*] salt solution with the addition of
0.05 c.c. of leef broth were also prepared in each ea-e. and inoculated
at the same time as the serum tule-* with an amount of culture e^ual
to that used for the .-erum tuljes. These tubes constituted the cheek-..
Immediately after the inoculation of the tubes, agar plates were made
from each of the tubes of salt solution with carefully measured
amounts, and after twenty-four hours at room temperature agar plates
were made from all the tubes of a series from lnth the undiluted
and diluted serum, and from the salt solution. In nearly all cases
0.05 c. c. was used for each plate. The colonies were counted after
twenty-four or forty -eight hours in the incubator at body tiwnperature.

TAKI.K X X I V. Bactcriolytie power of normal hog *tnun on tlilnlion.

No. of
hog.

Culture used to test
Lxu-terioljrrifc

Number of
tartfria
Introduced
per 1 c. c.

Number of bacteria in serum after 24
hour*.

Number

.. i..,,
teria in

V.llltlOll.

(MOoted.

lMllltr.1

1 to 50.

Dilut,-,!

i to KM

i.iiut.ii
1106,000.

L7M

1W8
1809

:-
1MO

Crawford

840
2,620

840
S.400

6,700

:.._<>
1,140

32.200
30,400

..,,..,

: :
1,020

1,460
2.100
10.200

1.9HO
1,600

2,040
2,900

5,900
10,900

f>3,MO

f> -:. HOO

' IT. Hi

2,900
3, WO

M
17.4<io

26.400
1.400
2.700

1 I

620

;,.. ,.

2,880

81.200
1,748

74,200

- '
1.9W

: . M

2,2X0

. 1'

900

60,800

p 26

. 1' 4692

Cniwfiinl

'. 1' 4692
Crawford

73,000

OB
OB
0*

!-:'
D ..'

F. 26

O. P. 4692

Crawford . . .

Dilnt. .1
1 > 10.

Iilnt.-.l
1 to 100.

DOvb !

It.. 1.000.

m.fioo

37.200
1.160
: ,000

30, 4*0

ll.4il

808.000

; .

1H7.600

;. ,,-.

..; (00

61.400

OB
OD
B

i;-. Ml
-. .
445.200

F.26

O. P. 4492

Crawford

F.26

O. P. 4092

Crawford . . .

F. 28

O. P. 4M2

oThia <-h*rnrter kigniflea that thvre were too many colonien to count.

* IMluted 1 to 4.

EFFECTS OF DILUTION UPON BACTERIOLYTIC POWER. 37

The results given in Table XXIV show that the serum taken from a
hog before injection i. e., normal serum as a rule, loses in potency
upon dilution; that is to say, a given amount of the diluted serum
kills actually fewer bacteria than the same amount of the undiluted
serum. This was to be expected a priori and from the results of
others with other organisms and with other animals. But the point
which seems specially worthy of note, and which will be discussed in
a different connection in this paper, is that the diminution of potency
on dilution is not in proportion to the degree of dilution. Not only is
this the case, but in at least 4 out of the 18 experiments given in
the table there was apparently an actual increase of power on dilution.
Those familiar with the subject will at once see the bearing of these
results upon the difference between the bacteriolytic properties of
normal serum upon the one hand and of immune serum on the other.
As it is attempted to show below, the difference seems, after all, to be
one of degree only, and not a difference of kind.

In interpreting the above results it must of course be borne in mind
that the number of bacteria found in the serum after twenty-four hours
probably does not represent merely the bacteria which are not killed
by the serum, but it represents in addition the number resulting from
the multiplication of these after bacteriolysis has ceased. Thus, in the
case of the serum from hog 1859, if 1 c. c. killed approximately 4,000
of the Crawford bacilli, 0.1 c. c. of the serum should evidently have
killed 400 and have left 3,600 to multiply, unless in addition to the
bactericidal power the serum also possessed inhibitory power for the
bacteria which survived bacteriolysis. If the serum had exhausted its
bacteriolytic power in ten or twelve hours and exerted no further
influence after this time, there should have been in the 1 to 10 dilution
after twenty-four hours a multiplication of the 3,600 bacilli left,
resulting in some 4,000,000 or 5,000,000 even at the rate of only one
generation an hour, as a simple calculation shows. But even if only
one bacillus were left after twelve hours, there would result over
4,000 in twent\*-four hours, provided there were no hindrance to
multiplication at the rate assumed above. It would seem, therefore,
that either relatively more bacilli were destroyed by the diluted serum
than by the undiluted, or that, aside from the bacteriolytic action, the
diluted serum exerted an inhibitory effect in the experiment quoted;
for at the end of twenty-four hours there were only about 1,000 bacilli
in the 1 to 10 dilution instead of the 4,000 calculated as the minimum
number possible under the supposition stated.

The dilution of 1 to 1,000 in the same experiment should evidently
have killed only 40 bacilli if it acted with the same relative intensity as
the undiluted, i. e., if it possessed just one one-thousandth of the potency
of the undiluted serum. Supposing that all bacteriolytic action ceased
after fourteen hours, the 3,960 bacilli left alive increasing at the rate
of one generation an hour would be represented by over 4,000,000

HAITI- :i:i"i vi 1- POWEfl "i iti.".|. -KUM ..i HOO8.

bacilli ill ten Inuu^, when -a- as a matter of fact tin-re \\eiv --nly about
:;-JJMN in this .liluti.'H aft, -r t\\ent\ four hour-.

Tin- time at \\hich bactei io|\ -is i- exhausted in serum at room t.-in-
p.-ratures \\as not detei mined, hut tests made at tin- end of three or
four hour- showed that up to thi- tiuif there was little or no appreci-
able effect Upon the number of bacteria introduced, -o that thcaltove

a uiiiption that tin- proce-- i- exhausted in twelve or fourteen hours
limy or may not be correct. Hut whether it i- or i- not. the relative
etlcct of the various dilution-, would be shown ly assuming the \<
6BI to come to an end simultaneously and thi- i^ prohahly the c-ae
at any time within the twenty four hours, the oliject of the ealcula-
tioi. IMMIIJT merely to show that no delinite rule of proportion < oiild be
oliM-rved, and that the diluted serum in all ea-e> d,-xt roved relatively
more Kaeteria than the undiluted -.enim. While dilution of normal
-erum eau-e^ a marked diminution in the haeterieidal power, it doe*
not >eem to dimini-h it to the extent that would be exj)ected 11 priori.
In other words, a dilution of 1 to 1< doe- not >eem to de-troy just
one-t*nth the number of bacilli that the undiluted -fimn destroy-, but
a much larger number than this, or at least if it does not destroy a
larger number it prevents a> relatively rivat an inerea-e as that
whi-h takes place in the undiluted serum.

Still another method of compari>on which apjx'ars permi->il'le. and
which leads to the same conclusion, is to calculate the number of
bacteria which should result in a dilution of 1 to KM) if the same rela-
tive increa-e took place as in the 1 to 1<> dilution; or to compare the
relative increase in any of the dilutions with that in the undiluted
serum in the experiment ^i\en above. For such a comparison it may
lie assumed that the organisms which were not destroyed in fourteen
hour- multiplied in the undiluted and in the dilated serum at the same
rate, or at least that if the serum retarded multiplication this effect wa>
more marked in the more concentrated serum than in the more dilute.
It would s. Tin evident, then, that the relatively smaller number of
bacteria in the diluted -eriiin can lie accounted for only on the >up-
po-ition that the diluted serum retains more bacteriolytic power than
would be exacted from the decree of potency shown by the un-
diluted scrum, according to the simple rule of proportion. This point
-eeins no t to have }x>en alluded to in the literature, but it would seem
nevertheless not without significance perhaps, since it tends ap-
parentlmto show that the difference U-tween normal serum on the one
hand ana immune serum on the other is nierelv a diflerence of degree
and not a difference of kind. The matter of the actual identity in the
behavior of normal sera and of immune sera on dilution, in spite of
the apparent difference shown by them, will be discussed more fully
in u different connection below . where the attempt is made to -how
that the law of complement diversion applies to normal serum as well
as to immune scrum.

EFFECTS OF DILUTION UPON BACTERIOLYTIC POWER. 39

DILUTION OF IMMUNE SERUM.

The results obtained by tests made to determine the effect of
dilution upon the bacteriolytic power of immune hog serum that is to
say, of serum from hogs injected with cultures in the main agree
with those observed in all recent investigations in this direction
with the serum of other animals injected with t3 T phoid, paratyphoid,
Asiatic cholera, and other bacteria. The behavior of the serum under
such conditions is very peculiar ; it seems to be in conflict with all
laws of proportion which are applicable to other chemical reagents,
particularly in regard to germicidal substances in general. With ger-
micidal substances in general, as it is hardly necessary to state, the
more concentrated the germicide the more powerful the destructive
power. A strong solution of carbolic acid is a more powerful germi-
cide than a weaker solution. But not so with immune blood serum.
Paradoxical as it ma} 7 seem, the bacteriolytic potency of the immune
serum increases upon dilution. The undiluted immune serum may be
weak or wanting in bactericidal properties, and yet become more and
more potent upon dilution. In other words, 1 c. c. of undiluted
immune serum may be found to kill few, if anj T , bacteria ; it may be
found even to furnish a good nutrient medium in which the bacteria
grow ; whereas 1 c. c. of a dilution consisting of 1 part of the same
serum to 10 or 100 or 1,000 parts of dilute salt solution may be found
to kill many hundred bacteria introduced into it. It goes without
saying, of course, that there is a limit to this increase of potency upon
dilution. Where the dilution goes beyond a certain point the potency
becomes weakened. There is for each immune serum an optimum
dilution for bacteriolysis, above and below which there is less and
less potency. Thus, as will be seen in some of the experiments
described below, the greatest potency in one case was in a dilution of
1 part of serum to 100 parts of salt solution. In the case of a dif-
ferent serum the optimum dilution for bacteriolysis was 1 part of
serum to 1,000 parts of salt solution.

Buxton very appropriately speaks of the phenomenon just described
as the formation of bacteriolytic zones, of which three may be recog-
nized in every immune serum a pro-killing zone, a killing zone, and
a post-killing zone. The action of the undiluted immune serum lies
in the pro-killing zone, that of the optimum dilution lies in the killing
zone; that of dilutions in which bacteriolysis becomes again weakened
lies in the post-killing zone. Of course it is to be understood that these
zones are not sharply marked off from one another, but that they grad-
ually merge into each other. The pro-killing zone, for example, may
extend over dilutions of several strengths, and the same is true of the
killing and of the post-killing zones. Thus in some of the experiments
given below there appeared to be no marked difference between the
bactericidal power of the undiluted serum and the same serum diluted

BAOTERIOLYTIO I'-'WKK <K KMIOD SKKl'M OK HOGS.

in the proportion of 1 to lo<>r 1 to lo<>; tin- pro-killing /one in these
QMe* would I... - :l id |,, ,.\ti ml t<> the dilution- o| 1 t.> 1" or 1 to I '".
The /one- a IT therefore spoken of a- longer or shorter, accordin
whether they apply t<> many or to few dilution-.

The jMViiliar behavior of immune serum ju-l de-ei -il.ed \\a- iii--t
ol-erved l>\ Nei--er and Weehsberg," and is called tin- Neisser-
\Yechsl>erg phenomenon. While it is true that tin- phenomenon was
ol.-er\ed (j ui to frequently in immune hog serum, a- will be -ho\\n
later. it was not as n<>ti<val>l<> in sonic ca-e- a- it wa- in other-. :md
moreover the three /ones varied in length, not only after the injee
tion of ditTerent strain^ of II. </////,// .v///.v in ditl'eient hog, but also
after the injection of the >ame strain in ditlernit hoj^s.

It ha> already been shown that normal serum of hogs gains rela-
tively in potency on dilution, and. as has just Keen Mated, immune
-eimn al-o puns in potency. The two behave alike, and at mo-t then-
is only a difference of degree between the amount of gain in either
a-e. But even a difference of degree between tiie two sorts of serum
i- not always noticed in the present experiments, for the normal serum
acted in some cases just like immune serum. It wa- not only relatively
more potent on dilution, but it killed actually more bacteria per 1
c. c. of serum in certain dilutions than in dilutions containing more of
the serum.

TABI.K XXV. /.'//. <t ./' <lihitin >i/>n immune hoy nmtin UK <<;///.// //'/// /
(liliilinii (//;</ //</ iinniinl vrniiii. Hmj ;.- / /;,/// ,-nlhii. <,. /'.

Culture

ll4-<) Illtl-t

Imi-tiTin-
ly>iv

Tinn- blood wan ilniun
relative In injeetiini.

Niimlier

of bacteria

inlnxlueed
per 1 e. e.

NiiintNTof luieteria in MTIIIII niter ?!

Illlllpi.

Number

terin in
elieckmll

Dilution.

:. 1
L.0001

2,700?

200?

I n.lililte.1.

Hilllte.1
1 to IIP.

iMhK <1

Diluted
1101,000.

(twwfiinl ..
F. '
<i.l'.40n..

l^'iUri 1 iniei'iiiin

3,700
3.760
600
600
1,080

i.oao

M
2.S40

200

80
1,400

,.,, L,,

l-.HKI

2,900
80

3..HKI

:i... li NI
8,800

M BOO
12.400

1,100

200
4.600
700
i ,000

11. IM,

27. HOD
l.NOO

-'7. li i
22.400

ll'.t.XW
4.900
HO
4.MO
1K7.600

l-.,
9.440
16. ON)

19.200
400

OCO

520
12,400
/
i
6,500
900

00
00

1.600

.' lioiirx lifter iiijfi-tiini
1 day after inje<-ti(.n. .
' (lavs after injeetioli.
Hi'fure injeetiiin

'2 hours after injtftiini
1 iliy after inji-^tioii ..
J ilays after Injcetinn.
Befi>ri' injection

'2 IKIIIIN niter injeetioli
1 day after injceiion..
2 days after Injection.

a Thin ehnrnetcr xiKnlne* that then- were too many colonies to munt.

An example of the difference often to be seen between the bac-
tericidal power of the senim from a hog hefore and aft<>r injection
with culture- of 11. ,-l,,l, // .v///X insofar as the .'tied of dilution is con-
cerned, is given in Table XXV. The animal in this case was injected
with a culture of the (J. I'. 4t'.:-j -train, and the blood was drawn
before and at several interval.- after injection. Before the injection
the serum from the blood of this hog showed inhibitory, but no l>ac-
tericidal power for the organism with which tin' animal was injected.

EFFECTS OF DILUTION UPON BACTERIOLYTIC POWER.

In this case there were 2,840 of these organisms introduced per 1 c. c.
into the undiluted serum, and after twenty-four hours there were
3,800 per 1 c. c. present. In the dilution of 1 to 10 the same number
of organisms introduced about doubled in number in twenty-four
hours; in the dilution of 1 to 100 the s%me number of organisms
increased about fivefold; in the 1 to 1,000 dilution there was a count-
less increase. In the undiluted serum from the blood of the same hog
two days after inoculation there was a sixty-two fold increase of the
200 bacilli per 1 c. c. which were introduced, while in the 1 to 10 dilu-
tion with the same number of bacilli introduced there were only 140
alive after twenty-four hours, in the 1 to 100 dilution only 400, and
in the 1 to 1,000 there were 1,600. In other words, the normal serum
diminished in potency on dilution while the immune serum increased
in potency on dilution. It should be noticed, however, that the dimi-
nution of potency of the normal serum is an actual and not a relative
decrease on dilution. In this case, as in others, the normal serum
killed more bacteria, relatively, when it was diluted than it did when
undiluted.

TABLE XXVI. Effect of dilution upon immune hog serum as compared with the effect
of dilution upon the normal serum. Hog 1798, injected urith culture F. 26.

Culture
used to test
bacterio-
lysis.

Time blood was drawn
relative to injection.

Number of
bacteria
introduced
per 1 c. c.

Number of bacteria in serum after 24
hours.

Number
of bac-
teria in
check salt
solution.

Undiluted.

Diluted
Ito50.

Diluted
1 to 500.

Diluted
1 to 5,000.

F. 26

Before injection

2,620
8,400

5,920
2,400
840
5,700
13,000
2,660
1,140
7,300

1,980
7,200

38,000
36,600

160
359,600
1,680
1,600
5,100

1,400
106, 400

38,400
686,800
26,400
400
6 12, 000
2,700
2,000
69,100

1,740
103,600

84,000
681,200
81,200
27,600
620,400
32,800
5,500
67,500

2,280
6,900

rt 00

6112,000
140

6175,000
5,200
26,600
boo

Crawford . .
G. P. 4C92..

10 davs after first injec-
tion.
9 days after second in-
jection.
14 days after third in-
jection.
10 days after first injec-
tion.
9 days after second in-
jection.
14 days after third in-
jection.
10 days after first injec-
tion.
9 days after second in-
jection.
14 days after third in-
jection.

00
CO

73,000

This character signifies that there were too many colonies to count.
6Dilutions used were 1 to 10, I to 100, 1 to 1,000.

That there is a difference in the behavior of the blood serum of dif-
ferent hogs injected with the same strain of B. ckolerse suis is shown
by comparing the results given in Table XXVI, hog 1798, with those
shown in Table XXVII, hog 1860. Both of the animals were injected
with culture F. 26. Before injection the blood of hog 1798 showed
considerably more bacteriolytic power on dilution for the organism
with which the animal was afterwards injected than the blood serum
of hogs did in other cases, though, as has been seen, normal hog

l-J

I: \ I ) Kl-'I.N IK I'.'U IK "| ULoop -1 Kl M oK II-

scrum frequently retains :i c.msider.ihle dcgr ,f potency upon dilu-
tion. This serum ap|>ears to have possessed about an equal d<
potency in all tin- dilutions which were tried, and to have Keen al>.ut
as potent wlu'ii diluted :is it was when undiluted. Ten day- nl'tei in
jection its behavior was v^ry peculiar, for it was apparently
potent a- the nuriual serum when il \\a- undiluted and in the I to
'dilution, luit it had ls| potency in t he dilut ions of 1 to .".o ;md
1 to 500. Nine days after a second injection it had apparently lost in
bactericidal potency, both when diluted and undiluted. Fourteen
days after a third injection it seems to have gained in power in all
strengths. If this is eoinj)ared with the behavior of the serum from
the blood of hog IM'.H, Table XXVII, a hog also injected with the F.
J'. -train, it will be noticed that there is a marked difference.

TAHI.E XX VII. Efferi of dilution </*// immune hog nerum at compared with the effect of

ililiiliii iifHDi thr normal ternm. ll<><) JSi',n, injfrl,</ intnn nniixlii n four MUXCtrive
days ii'ilh culture F. 2(S, nlmni .' <-. <. lirntxl ir tliirt;/ minnti* nt '>8-^5S C.

culture
uned to teat

Time l>loo<l ua- ilra \vn

Number ol
bacteria

NunilMT of bacteria in wnim after 24
boon.

Number

.,; ', ..
fi.rlu In

bacterio-

1VM.

relative to inj.-etion.

Introduced

|ier 1 '. c.

Dndltatad,

l>ilntc<l
1 to 10.

Dilated

1 to 100.

Dilute,]

1 to 1,000.

icrin in
ahaekai
Milutlon.

('niwforil

Hefore injection

1,460

12,500

118 400

1M Till

1 day after injection. .

1/JI"

ft. 200

2 ilaysafter injection.

140

a 30

19, INK)

, . ,

460

:t ilaysafter infection.

I i

6 30

1 clnys after injection.

1,00

I M

14,400

1 Ml

F.26

Before injection

J.I (Ml

504. C0

V .

1 ilay after injection..

880

70,000

10.000

.' <la>> after injection.

1,100

OD I /

200.200

A4.400

:t 'lays after Injection.
I ilays after injirtion.

1,000

i ' ,

L.OO

/ ft 30

2,860 6.900

IS. 600

<i. 1'. 4692..

Hefore injection

10,200

17. lii 11 IIKI

'1. HO

MB m

a ,..,

1 ilay after injection..

1,100

<X) H, MOO

10,600

., v ,..

J ilaysafler injection.

1,600

OB I /

, l. ;,.,

i i ,,

4.400

:< 'laysaftcr injection.

9.900

a | Ax

4 ilny* after injection.

:,o.>o

2,100

2,4*)

10.800

oThU character MKniiie* tliat there were too many colonies to count.
6 Dilution- 11-.. 1 W.TC 1 to :*i, I to. '-in, 1 to 5,000.

In the experiment with the blood serum of hog I860, Table XXVII,
the animal was injected on four successive days with <|tiite den-e MM-
pen-ions, considerably denser than a twenty-four-hour culture >f
typhoid. The >uspensi(, n^ \\ere heated for thirty minutes ; ( t ;,s ;,: ( '.
Iwfore injection, and the injections were made intravenously. On
each day before injection, and previous to any injection, blood was
drawn from the tail of the animal, and after the separation of the
serum this \\as used on the three strains of //. , }n>l< r;> .-///*. Doubtless
the fact that the test- were made while the animal was being immu-
ni/ed had some effect upon the results obtained, and for this reason
they are different in some measure from those obtained in other
experiments in which the serum was tested only after the animal had

been immunized.

The undiluted serum from this ho<_r before injection was quite
strongly bactericidal for the Crawford and the F. _'', -trains, but it

EFFECTS OF DILUTION UPON BACTERIOLYTIC POWER.

WHS at most mereh 7 inhibitory for the G. P. 4692 strain. After the
second and third injections the serum seems to have lost all potency
for all three organisms, at least when undiluted. But after the fourth
injection the serum seems to have regained its potency both when
undiluted and in certain of the dilutions.

TABLE XXVIII. Effect of dilution upon immune hog serum as compared with the effect
of dilution upon the normal serum. Hog 1885, injected with Crawford culture.

Culture

used to test
bacterio-
lysis.

Time blood was drawn
relative to injection.

Number of

bacteria

introduced
per 1 c. c.

Number of bacteria in serum after 24
hours.

Number
of bac-
teria in
check salt
solution.

Undiluted.

Diluted
ItolO.

Diluted
1 to 100.

Diluted
1 to 1,000.

Crawford . .
F. 26.

1 hour after injection
1 day after injection. .
1 hour after injection
1 day after injection . .
1 hour after injection
1 day after injection. .

5,860
3,200
16,600
6,900
9,600
4,600

18,200
71,100
oo

31,000
3,600

5,200
110, 400
40,400
7,200
9t>,000

33.600
11,000

28,000
10,700

a oo

CO
00

00
00
00

oo
102,200

1,280?

G. P. 4692..

a This character signifies that there were too many colonies to count.

In the experiment of which Table XXVIII gives the summary the
hog was injected intravenously with a dilute suspension of the Craw-
ford culture. Two injections of the organism were given on two con-
secutive days, and blood was drawn from the animal one hour and
twenty-four hours after the last injection. No blood was drawn
before injection in this case;

There seems to have been no difference between the behavior of the
blood on dilution and that of ordinary normal blood on dilution.
Omitting one or two unimportant discrepancies, the serum lost more
and more in potency upon further and further dilution.

TABLE XXIX. Effect of dilution upon Immune hog serum as compared with the effect of
dilution upon the normal serun,. Hog 1809, injected with culture G. P. 4692.

Culture

used to test
bacterio-
lysis.

Time blood was drawn
relative to injection.

Number of
bacteria
introduced
per 1 c. c.

Number of bacteria in serum after 24
hours.

Number
of bac-
teria in
checksalt

solution.

Undiluted.

Diluted
Ito60.

Diluted
1 to 500.

Diluted
1 to 5,000.

F. 26

Before injection

5,920
2,400
' 5,700
13,000
1,140
7,300

5,900
1,800

56, 600
10,900
7,900

109,200
l> 1<>8, 000
520
ft-1,000
2,880
&9.200

<l 00

''224,000

f'6,l>00
123,200
b 25, 900

oo
b oe
oo
'< v>
60,800
b oo

oo
OO

oo
oo
73,000

Crawford . .
G. P. 4692 ..

14 days after injection.
Before injection
14 days after injection.
Before injection
14 days after injection.

a This character signifies that there were too many colonies to count,
b Diluted in the proportion of 1 to 10, 1 to 100, 1 to 1,000.

In the experiment summarized in Table XXIX the hog was injected
subcutaneously with 5 c. c. of a suspension of G. P. 4692, but there
seems to have been little if any effect produced upon the bacteriolytic
power of the serum. It is true that with the Crawford strain there is
some evidence of loss of potency in the undiluted serum, and of
increased potency upon dilution in the serum after the injection of the
animal, but the effects on the whole are not striking.

SKKI'M

TMUI \\\ /-on mi in unt hog itfrn in >

ul ililittinn 11/1,11 lh> nnrmtll Hfritm. //</ IMfll. ni/V.7.--/ //. ' 'run -fnl >

Culture

twclerlo

K |

'I 1. I.H.I .i.,lniw n relative
In injection.

Number

Intro

|.er 1 c c

Nmnherof Uicleriii IIIMTIIIII i,
hour-

tin In
-alt

1 M.ll

luted.

IMI.ilMl
Ito.V).

i io 000

lMlule.1

Hi. la\ suit, r tir-t injection..
a d*v after M-comi injection
1 1 ilnv- after thinl Injection.
Hi la\- after tint Injection. .
.' il:i\ ai'icr ccoui injection
11 .lay* after thlnl injection.

10 day* after ilrst injection.
9 tUvK nfier M-cotnl injection
14 day* after thinl inj.-ctlon.

8.400
MO

ton

1.1 in

7 BO

,..,

' 60

(1
1-11

' -J"

3. .-WO
7!00

|9 MOO

oo MM

640

19,000

:i 000

not

: 000

"I'.l -1l

0, 1 1

fc8,100

172.000

I ,-.

. n

7^600
'21.000

..nl ..

., ] |I.T;

78.000

"This cluirnctcr ^ik'iiilic- tliat there \\rn- I.... many coloni.
'Hilule.1 1 to 1(1. I to 1IKI. I to 1,000.

The hog (1801) in (he experiment abown in Table XXX wa- injeeted

on three different occasions with the Crawford organism. No te-t-
were made of the serum before the inoculation of the animal, but there
-ems little or no evidence of the reaction usually seen in the serum of
animals injected with bacteria in the tests which were made after each
injection. The serum in most of the test- showed !<>-- of bacteriolytic
power on dilution, behaving in this respect like normal serum, ll i-
true that with the homologous -train Crawford -the -enini ten <lay>
after the first injection and fourteen day- after the third injection
appeared to be somewhat more strongly bacteriolytic in the dilution
of ltx>5,<HX)and ltol,(X><>. respectively, than in the le diluted -erum:
but on the whole the serum .seemed to act more like normal than like
immune serum.

TABI.K XXX I. KtJ'i-i-t <>f >Hlntii,n HJXHI innmine hog xerinn nx n.ii,j,,ij,<l mil tin ,f
Man tin- normal writ in. Jim/ Jfi.iH, injn-t,<l mil, ni/liii, /'. S6.

Culiun-

UXtKl '

(it-tiTiii-

Crawfonl . .
K. 26

Tinii- l.l. .-! waxdrawn
rt-lntivi- to injection.

Number of

luicieriii

hatrodooed

|HT 1 c. c.

NiiinlMTof liHctcrin in M-riiin after Jl

bam.

NiiiiiU-r
of 1-jirtc-
ri in

cheek silt

Dilution.

...

Dtodttmad.

. 1840
<i. r t.S40

JI7.MO

a an

"47.400

, | ...

lilllte.l
1 to Id.

I>illlte,l

l>i iiinl

1 to 1. (111.

Before injii-tfon

oo m

i M
00 (00
... (00
ft. 000

a, ooo

, ...

240

117,600

0| ...

- I'-
ll MM)

; . 000

OB
1 .1 Ml

00
OB

30K.OOO

1". i

7 m

'"
OB

1 hour iifli-r iiiji-<-tloii ..
Sdayx iifti-r injcctiini .

Mfforr illjrctioll

' I' MW2..

1 hour nfii-r injection ..
- after injection. . .
Before Injection
1 hour after injection . .
Sdaynaftcrinji tion. . .

IMlilteil in the |.ro|Nirtion of 1 | art of MTIIIII to :: of |-l.\ !..:. cu nl -nit solution.
*>Thi chiinicier 'iuiiitli^ Hint there were !> iiiiiny colunjc* to count.

In the experiment with the serum fr<m the blond of hor \W*. Table
XXXI. the animal was ujven intrax enoii>ly ." c.c. of a dilute beef-broth
>u|>fn>ion of an ajjar culture of F. _'' -train. < )ne da\ after the first

SUMMARY OF DILUTION EXPERIMENTS. 45

injection the animal was given a second injection of the same amount
of the same strain. The blood was drawn before any injection was
made, one hour after the first injection, and three days after the second
injection.

The results oDtained with the serum from the blood of this hog, at
least in so far as the strains G. P. 4692 and Crawford are concerned,
are in striking contrast with the results obtained with the serum from
the blood of hog 1801 (Table XXX). In the case of hog 1836 the ani-
mal was injected with a more virulent strain than was the case with
hog 1836, and it is possible that this may have had some influence
upon the results in the two cases. But aside from this the two sera
must have had different potency originally, for even the normal serum
of the blood of hog 1836, before the animal was injected, will be seen
to have shown increase of potency upon dilution; the dilution of 1 to 10
was more potent than the dilution of 1 to 4. There was no test made of
the undiluted normal serum for the reason that not enough of the blood
was obtained to use for the purpose. This increase of potency on
dilution of the normal serum of hog 1836 is to be seen particularly in
the tests with the Crawford strain, in which the dilution of 1 to 4
showed 3,840 bacilli per 1 c. c. whereas the dilution of 1 to 10 showed
only 320 bacilli after the same exposure one day in each case with
the same number of organisms 32,200 per 1 c. c. introduced. This
serum in other words seems to have behaved somewhat like immune
serum to start with, and this behavior seems to have been intensified
by the injections, as will be seen by examining the table.

SUMMARY OF THE DILUTION EXPERIMENTS.

On summing up the results of the experiments made with normal
and immune hog serum on dilution, it will be noticed that the normal
serum seems to differ greatty in its behavior in different cases; that
while it sometimes shows marked diminution of potency upon dilution
it sometimes shows increased potency on dilution. The immune hog
serum also shows variation in behavior.

But the results which would seem to be of especial interest are those
in which the immune serum showed strong bacteriolytic power when
undiluted, and lost in power on dilution in certain proportions, but
gained potency upon further dilution. In these cases there appeared
to be a combination of the behavior of normal and of immune serum,
as this is usually stated to occur.

As already stated above, the characteristic behavior of immune
serum is said to consist of the formation of bacteriolytic zones. A
typical immune serum reaction consists of a pro-killing zone, a killing
zone, and a post-killing zone on dilution. In the pro-killing zone,
represented by the undiluted immune serum, there is no bacteriolysis.
In the killing zone, represented by one or more dilutions, there is

I'. n\< IKUIOI.Y n< I-IIXVH: >K in. ...... NEBUM 01 HOO8.

bacterioh -i-. In tin- p<>-t killing zone, represented by further dilu-
tion than tin- killing /one. there is again lack of bacteriolysis. Bui in
tin- re-ult- at pre-ent under con-iderat ion tin- serum showed a killing
/one to begin within the undiluted serum, thus imitating to tin- extent
the normal -erum.

The peculiar behavior ju-t alluded to was to he >een in sexn.il
instances given in the preceding tallies. Perhaps the mo-i -trikm-_r
example- are furni-hed by the lehavior of the immune >era from I
171'San.l l^iix ,.,,.,. Tablr \\IYi. I,, the-.- -era with the Crawford
culture there wa- strong bacteriolysis shown in both ca-e- undiluted.
The Ixos \\a.- still trough potent in the 1 to :> dilution, but the 17'.^
serum was much less potei-t in this dilution than when undiluted.
lioth sera seem to have lo-t all potency when diluted in the proj>ortioii
of I to 500. But the jx)int to be specially noted is that they both
-.mi to have Lfained ^-really in the dilution of 1 to .".JUKI; particularly
is this the case with the 17!* scrum.

For the present it is merely nece ary to cull attention to the-e
re>ult-: they will l>e taken up and discussed in another connection fur-
ther along in the present pajx-r.

GENERAL DISCUSSION.

Judging from the results of the above experiments as a whole, it
seems apparent that the blood serum of hojr- ha^ but feeble bactericiilal
potency for K. cl>!< r;> A-///.V. at least for the more strongly pathogenic
strains. This would naturally lead to the presumption that the known
susceptibility of hogs to infection by intravenous injection was due to
this cause, bat it doe- not -eem a- yet >ettled that the sii-cept ibilitv
of an animal in general is in direct proportion to the want of bacteri-
cidal power of its blood. In fact, in some instances, as is well known.
quite the contrary has been shown to be the ea-e. Nuttall, Huchner.
and many others have shown that rabbit serum in te-t tubes will kill
large numbers of anthrax bacilli, and yet a comparatively small num-
ber of these organisms introduced beneath the skin of a rabbit i- . r
tainly fatal to the animal, (iuinea-pig serum, as recently >hown by
Buxton,' may be expected to kill approximately I,OOU.<MMI typhoid
bacilli per 1 c. c., and yet the injection of a relatively small amount
of this organism intraperitooeally ifl certainly fatal for guinea pig-.
In spite of this, however, a- \VoltF' : " ha- pointed out. it is probably all
a matter of dovigr. and bacterioly-i- i- an ellicient means of defense
where the number of bacteria concerned is not large. On the con-
trary, where the number of bacteria exceeds certain limits, bacteri
o|\-is i- not only not a mean- of defen-e. but it actually lil>erate- the
di -ease-producing -ub-tance> the endotoxin- from the Ixxiies of the
bacteria.

GENERAL DISCUSSION. 47

Pfeiffer" was the first to give this name, endotoxins, to the poison-
ous substances which are present in the bodies of the bacteria of
certain species, and it is now quite generally the opinion that it is the
liberation of these that brings about disease in infection with these
bacteria. So long as bacteria of this kind remain intact, they are
thought by those holding this view to be incapable of producing dis-
ease. Mechanical injury and other harmful effects formerly attributed
to the action of bacteria are now thought by many not to exist.

The effects of inoculations of hogs with cultures of B. choleras suis,
and the conclusions to be drawn from these experiments are on the
whole in harmony with those derived from experiments with various
bacteria injected into rabbits, guinea pigs, and other animals. Thus
Loeffler and Abel 12 found that undiluted immune blood serum from
dogs failed to protect guinea pigs from injections of virulent colon
bacilli. Neisser and Wechsberg 14 obtained analogous results with
immune rabbit serum in test tubes, the undiluted immune serum failing
to destroy the bacteria of the kind used to produce the immune serum.
Buxton,* as already quoted, found that the undiluted serum of rabbits
injected with cultures of typhoid or paratyphoid bacilli behaved in a
similar manner. It was to be expected, therefore, that the serum of
hogs injected with cultures of B. cJiolerae suis would show loss of
bacteriolytic power in undiluted serum. The behavior of the diluted
serum will be considered below.

In the foregoing experiments with the serum from the blood of
hogs injected with cultures of B. cholerse suis, it appears that
the efiect of the h'rst injection was sometimes, as stated, to deprive
the undiluted serum of its power to kill B. cholerae suis. It is true
that this behavior was frequently lacking. But if the results after
several injections are examined, it will be found that the undiluted
serum frequently regains its power. This is noticeable in a number
of cases if not in all, and it perhaps accounts for the experience of so
many observers that animals frequently die from injections of com-
paratively small doses after having been treated with carefully graded
amounts of culture. Thus Wolff 20 found that strong, healthy rabbits,
instead of becoming more and more accustomed to injections of
cultures, withstood these less and less readily, even without increasing
the dose; that after remaining apparently unaffected by previous
injections, they often die suddenly on the inoculation being repeated
with the same sized dose. This supersensitiveness on repeated in-
jections has often been observed in various animals and with various
bacteria. It has also been met with in the course of the present
experiments with hogs. In fact, it does not seem so difficult to grade
the first dose so as to avoid killing the animal as It is to strike upon
the proper subsequent doses. The return of bacteriolytic properties
in the serum following further injection of animals may account for

48 BA< TKKH-I.VI K POWXB 'i BLOOD 9BBUN 01 H"<

(hi- result, lli- lir-t injection- depriving the -ermn of the l.a.-tn i.|\ ti.
DOWer Mkd thttt preventing the liberation <f endotoxin and it- <-on
M'.|IICMCC-. and the subsequent injection- producing :i ivturii of
bacteriolylic |HWT and causing the liberation of the cndotoxin.

It would seem porhap-. difficult to reconcile the fai-i (hat tin- -mini
becomes more |)oteut for tin- protection of animals other than the
individual which produces it. and that \ rt fnun tin; very fart of it-
bacteriolytic i>ower it should U> injurious to the animal furni-hing it.
It would appear paradoxical to lind that -mini obtained from an an!
nial jMissessed the property of protecting other animal-, and, at the
same time, not only failing to protect tin- animal furni-liin^ it. hut
actually increasing the >u>ceptihility of this animal. Here again
WoltF offers the explanation that it is a matter of dosage. If the
numher of Imctoria injected into the animal furnishing the >ermn, or
into the animal which has been ^iven a protective dose of the -erurn.
i- not so large that on In'ing disintegrated the amount of toxin is >uf-
ficiently large to cause disea-e. the animal escapes. There are appar-
ent difficulties in the way of this explanation, but it is, perhaps,
nevertheless not without weight.

In this connection the results obtained by Elischer and Kentxler*
on comparing the bactericidal power of unheated immune serum
i. e., immune serum containing its own proper complement with the
same serum heated and reactivated by the addition of fresh rabbit
serum would seem to apply. The latter was found to be more potent
than the former, and this would seem in a way to offer an explanation
of the protective power of immune serum for an animal into which it
is injected on the one hand and of the failure of such protective prop-
erties of the same serum for the animal from which it was obtained
on the other. Elischer and Kentzler interpret their results as indi-
cating that immune serum contains an insufficient amount of com-
plement for the full development of its bactericidal ]>ro]M>rties. So
in other ca>e-. the amount of complement in immune serum may be
insufficient to enable the serum to protect the animal from which it
was obtained, while it would find sufficient complement in the serum
of a fresh animal. To state this view briefly, the excess of aml>ocep-
tors in immune serum finds the requisite amount of complement in
the normal serum of uninfected animals.

When the behavior of hog serum on dilution i> considered, the
explanation afforded by the theory of complement diversion seems
applicable, not only for the behavoirof immune >erum in which there
was shown often to be increased bactericidal |>otency on dilution, but
also in the case of normal serum, as it will be attempted to show.

It will be recalled from the statement^ in regard to the theory of
complement diversion given on page 1:5 that in accordance with the

GENERAL DISCUSSION. 49

views of the Ehrlich school generally it is held by Neisser and
Wechsberg, the originators of the theory, that bacteriolysis is due
to the action of a substance called complement, which is present in
normal and in immune serum, but that this substance can act on the
bacteria only when it becomes united to these through the medium of
certain bodies called amboceptors. The complements are held to be
incapable of uniting directly with the bacteria. Now, according to
the theory of complement diverson, if there are more amboceptors
than complements present in any given serum, all the complements
which are present unite with a corresponding number of amboceptors,
and if any bacteria are introduced into the serum they unite with the
free amboceptors in preference to uniting with those having comple-
ments attached. In this way the complements become diverted from
the bacteria by the excess of amboceptors. It will also be recalled
that while there are amboceptors in normal serum, these become
increased when the animal from which the serum is obtained is
injected with bacteria. The amboceptors formed by these injections
are specific; that is to say, an animal injected with B, choleras suis
develops amboceptors in the serum which are capable of uniting with
B. choleras suis only. Reference to figure 4 (page 51) will serve
to elucidate whatever may be obscure in the above explanation of the
theory of complement diversion.

Bearing in mind that normal serum contains a relatively large
amount of complement and a relatively small number of ambocep-
tors, it is evident that by introducing normal serum into immune serum
it is possible to increase the amount of complement without adding
materially to the number amboceptors. In fact, by making a series
of dilutions of immune serum and adding to each lot in the series a
definite amount of normal serum, it is possible to obtain a set of dilu-
tions containing various amounts of amboceptors and approximately
the same amount of complement. Evidently if the immune serum is
merely diluted and no normal serum is added, the complements
present will be diluted in the same proportion as the amboceptors.

If the amount of complement is kept constant in the way above
described, by the addition of a constant amount of normal serum, it is
evident that in some of the dilutions of the immune serum the ambo-
ceptors will have enough or more than enough complements to com-
bine with the amboceptors, and that in such cases the conditions
necessary for bacteriolysis will be fulfilled the amboceptors with the
complements attached will unite with the bacteria. In other words,
the undiluted immune serum, having an excess of amboceptors over
complements, fails to show bacteriolysis, while the diluted immune
serum with complements added, having no excess of amboceptors
over complements, causes bacteriolysis.

f'H HA I I Kh'l.'i I l< I'.'WIIJ <>|. MI..i| -I i;i \| .., BOOB.

Buxton' ha- raised the objection to tin- theory that it can account
for those cases only in which the an:, unt of complement is kept . ,,u
slant in the dilutions of immune serum, while the amboceptoi - alone

re diminished. HA makes the point that in experiments such as arc
recorded in the present pa;>er, in which amlioceptoj-s and comple-
ments arc diluted to an equal extent, the theory of complement (Inci-
sion is inapplicable; and >ince the pr<t-e-- i- probably the -anie in loth
cases, both where the complement is kept to a on-taut amount by
heating the serum and adding a definite quantity of normal scrum on
the one hand and where the serum is not heated and ha- no normal
serum added on the other hand, the theory a- a whole i- not tenable.
Valid as this objection appears from Buxton's presentation, it appear-
on further consideration a- scarcely to be insuperable, for the theory
seems to afford a satisfactory explanation in every case \\liere the
ambocepton are in excess of the complements, and, granting A cer-
tain proportion between these in the undiluted serum, it i- not n-
sary for the complement to be present in an equal amount in the undi-
luted serum and in the dilutions for the occurrence of the diversion
of the complement to take place. All that the theory would seem to
demand is that the undiluted serum contain an insufficient amount of
complement to satisfy the amboceptois which are free and those which
are attached to the bacteria. Bacteriolysis will take place just in pro-
]>ortion to the paucity of the free amboceptors, for, according to the
theory, those bacteria which are not coupled to free amboceptors
are liable to bacteriolysis by becoming attached to the amboceptor-
complement combinations.

In figure 4 it is attempted to show that comp.ement diversion may
well account for those cases in which the complement is diluted in
tin 1 same proportion as the amboceptors. This point could be more
strikingly shown if a larger number of amboceptors and complements
could be taken, and a much greater relative difference could l>e made
in the diagram between the number of amlxxjeptors and of comple-
ments, but for obvious reasons it is impracticable to do this. Thus, if
it were feasible to show 100 amboceptors and 100,000 amboceptor-
complement combinations in 1 c. c. of undiluted serum, it could be
demonstrated that 100 bacteria introduced would all esea|>e bacterio-
lysis in this serum, only 10 bacteria would escape in 1 c. c. of a dilu-
tion of 1 to 10 of the same serum, only 1 bacterium would escape in 1
c. c. of a dilution of 1 to 100, and none would escape in 1 c. c. of a
dilution of 1 to 1,000. In other words, theoretically in serum in which
the above conditions prevailed, taking 1 c. c. in every case and intr<>
ducing 100 bacteria, there would be no bacteriolysis in the undiluted
serum, but complete bacteriolysis in the same serum in the 1 to 1,000
dilution.

GENERAL DISCUSSION.

In figure 4 the amboceptors are represented by the parts marked
#, the complements by the parts marked #, and the bacteria by the
parts marked b. No. 1 is intended to represent the undiluted serum,
and all except two of the amboceptors are shown as having diverted
a corresponding number of complements, while the two amboceptors,
which are not united to complements, and which represent the excess
of these bodies over complements, are united to a corresponding num-
ber of bacteria. Evidently in such a case there would be no bacte-
riolysis, since all the complements have been diverted from the
bacteria. In No. 2 half the amount of the serum, as in the first case

FIG. 4. Diversion of complement in Immune serum, not heated, and \rithout the addition of

foreign complement.

that is to say, 1 c. c. of a 1 to 1 dilution is represented, with the
same number of bacteria added as in the undiluted serum represented
by No. 1. Under the conditions represented by No. 2 it is evident
that one-half of the bacteria would be killed. In No. 3 is represented
1 c. c. of a dilution in which 0.25 c. c. of the original serum is contained.
In this case all the bacteria would be killed, since there are no longer
any free. amboceptors to unite with the bacteria, or, at any rate, there
would not be enough of the amboceptors to divert the complement in
all of the four portions into which it is assumed the serum is divided.
If the amboceptors are to be regarded as chemical substances in solu-
tion, then there would not be enough in &ny one of the four portions
to cause diversion of the complement. If, on the other hand, the

.'_' BA<TKK1I.YTI< I-..W1K <>K HI IM'M "I MIN!

amlo,-ept,.r- are l>>lie- in tin- >hirt -MX,., thru it i* e\ idmt that two
of the four portion.- would show liict-i i>|\ -i- ,.f <>ne half f tin- la-
teria, while the other two would -li>\\ ...mplrte ta<-teriol\ -i-. a inn
ing uniform distribution of (lie amboceptors. In rithrr cast- the
diluted scrum would be more powerfully bactericidal than the undi-
luted serum. It will be observed that the complement is not k< pt
at u constant amount in the dilutions represented by the diagram-,
but that it is diluted simultaneously and to a like degree with the
amlK>ceptors.

I-'rom what has been said above it would appear that diven&n of
complement seems a rea>oimble explanation of the phenomena whieh
are observed in unheated un reactivated serum, and it therefore >emis
to overcome Buxton's objection to the application of the theory to the
phenomena observed under such condition-.

It is aside from the purpose of the present article to try to meet in
general objections to the theory of complementdiversion, or todi-cu--
this theory apart from whatever bearing it may have upon the pre-ent
experiments; but incidentally the objection raised by Buxtonof bring-
ing in accord with the theory of complement diversion the fact that
Asiatic-cholera spirilla are killed about as readily by typhoid and para-
typhoid immune serum as by normal serum may perhaps be overcome
by consideration of the fact that the injection of an animal with any
one kind of Imcterium probably has no effect upon the aml>oceptor-
complement combinations for any other kinds of bacteria which are
present in the serum of the animal undergoing inoculation. The effect
is probably very sharply specific, even going so far in some cases as to
affect different strains of the same organism, as is shown by the results
of some of the experiments in the present paper. It seems not at all
inconceivable, at least, that an animal injected with any Driven kind of
bacterium might well be deprived of the power of killing the kind
of bacterium with which it was injected, owing to the excessive pro-
duction of specific amboceptors, without interfering with the ambo-
ceptor-complement combinations for any other bacteria which may be
originally present.

In view of the very specific reaction which takes place on injecting
an animal with bacterial cultures, there would really seem to be no
good reason why immune serum should act differently from normal
serum for any but the homologous organism. The amboceptors which
are newly formed in consequence of the injections could not reasonably
be supposed to have the power of depriving the amboceptors originally
present of the complements to which they are attached. If it is true, as
assumed, that in normal serum there are am boceptor-coniplement com-
binations which will destroy D. cholera? #//*, and in the same serum
aml)oceptor-complement combinations which destroy //. typhoawi, the
injection of the animal with cultures of B. cholerxauis would evidently

GENERAL DISCUSSION, 53

increase the amboceptors for this organism, and these newly formed
amboceptors would have the power of uniting with any free comple-
ment, but not with any complement already united to the B. typhoxux
amboceptors. So that while the newly formed amboceptors would
and probably do find free complements with which to unite, the} 7 could
hardly be expected to tear loose the bonds between the original ambo-
ceptors and complements.

As has been pointed out by Neisser and Wechsberg in their experi-
ments, the increase of bacteriolysis usually observed on the dilution of
immune serum does not appear to be explicable by the conception of
Bordet that the immune body the amboceptor is merely a sensitizing
agent. For if the amboceptor were in reality merely a "substance
sensibilisatrice " it is not likely that immune serum would be more
active when dilute than when concentrated. At least it would be only
reasonable to presume that a reagent possessing the properties ascribed
to the sensibilisatrice by Bordet and his followers would lose rather
than gain in power upon dilution, and that a given amount of the
undiluted sensibilisatrice would be more potent than the same amount
of the diluted substance, as is the case with other reagents. On the
other hand, Ehrlich's conception of the amboceptor affords a satis-
factory explanation of the behavior of immune serum upon dilution, as
has been shown in the above discussion of the theory of complement
diversion.

The results obtained in the present investigation in which the bog
serum appeared to increase in potency upon dilution seem best ex-
plained by the theory of complement diversion. But in these experi-
ments, as in all others of like nature, it would seem possible that the
process of bacteriolysis or the failure of bacteriolysis may at least be
complicated by the formation of antibacteriolysins on the injection of
the hogs or by the presence of antilysins normally present in the
serum. Opinions a re. divided upon the question as to whether there
are such antilysins present in normal or in immune serum. But if
there should be such antibodies present, it is reasonable to suppose
that these would be weakened in power on diluting the serum, and
that consequently such serum would be more potent when diluted than
when undiluted, unless the lysins and antilysins should be diluted in
like proportions.

There seems to be no difference of opinion in regard to the develop-
ment of qualities antagonistic to bacteriolysis in serum under certain
circumstances, but the matter in dispute is whether this antagonistic
property is due to special antilysins or whether it is due to the pres-
ence of nonspecific, normal amboceptors or of superabundance of
immune amboceptors which absorb the complement, and in this way
prevent bacteriolysis.
Pfeifl'er and Friedberger, 16 who were the first to detect this antago-

54 HMTKKI>M I I' P.'UIK <>K IU.""I> -I KIM Of H<<8.

ni-tic action on the part ( normal scrum. hold that under certain cir-
cumstances specific antilxxlies tire formed, which coinliiin> with the
complement. In other word-, they regard the antih -in- a-anticom
plements. Their 0zperime&tl < % <nsi>t of triturating normal serum
with bacteria ami centrifugali/.ing. I'lider -uch cir.-um-tance- the
-iipernatani serum acquires the proj>erty of neutnili/ing the bacieri-
lytic jxwer of immune serum. Thus normal scrum, which in it-e|f
exerts noantihactcriolytic power, becomes EBtiDBCterfolytic for cholera
immune scrum when it is triturated with cholera bacilli and centrifu-
galixed or filtered through a bacteria-proof filter.

Gay" attributes antilytic pn>i>erlics developed in this way to the
formation of specific precipitines. He finds that only when then- i^ a
specific precipitation is there any antilytic property shown by henio
lytic sera, and since hemolysis and bacteriolysis are probably identical
processes, the same statement would apply to the latter a- well as to
the former.

Gay's view has been called into question by Sachs, 11 who attribute-
the antilytic properties of serum to normal, nonspecific amlnjceptors
which he regards as acting as neutralizes of the complement. The
only difference l>etween the explanation offered by Pfeiffer and Fried-
bergeron the one hand and Sachs on the other is that while the former
look upon the antilysins formed in their experiments as newly formed
bodies derived in part from the bacteria, Sachs contends that the
antagonistic bodies are present in the serum from the start. He
states, at least, that this is the only point in which he take- i uc with
Pfeiffer and Fricdlwrger on the matter of the formation of antily>in-
for hemolytic scrum.

Lipstein,"on comparing strongly agglutinating immune sera, found
that while l>oth the specimens examined possessed equal agglutinating
power only one of them showed the phenomenon of complement diver-
sion that is to say, antibaetcriolytie properties. He was also unable
to discover any constituent of the normal serum which showed diver-
sion of complement, but on the contrary he found this phenomenon
to take place in immune serum only. In other words, the antagonistic
action of serum to Iwcteriolysis he attributes solely to the diversion of
complement by the specific, immune amlxx-eptors.

In spite of this conflict of authority, it would seem at least that the
Neisser-Wechsberg phenomenon permits of easiest explanation by the
theory of complement diversion. For if the failure of the undiluted
immune serum to produce Iwicteriolysis were due to the presence of
antilysins, it would U less potent upon diluting the serum, it is true,
and consequently exert less power u|>on the Imcteriolysins when diluted
than when undiluted, but the relative effect would remain the same.
It is nevertheless prolmble that if Imcteriolysins are present in immune
-enini they complicate the process.

GENERAL DIWCUSSION. 55

While the theory of complement diversion seems to have been used
by all investigators to explain the bacteriolytic phenomena of immune
sera only, it would not seem inapplicable to those which take place in
normal serum as well. The theory seems at least to explain the
behavior of the normal serum of hogs upon dilution in the present
investigation. Even in those cases where the normal serum showed
diminution of potency on dilution, the diminution was not in direct
proportion to the dilution, as has been previously noted; but on the
contrary the diluted normal serum acts more powerfully, relatively,
than the undiluted normal serum. This relatively greater potency of
the diluted normal serum (>tr the undiluted may be accounted for
by the theory of complement diversion, if it is assumed that there is
an excess of amboceptors over complements in the normal serum of
hogs to start with, as certain considerations will show.

Difficulties in the way of the acceptance of the explanation just
offered are of course very obvious. In the first place, normal serum
generally has been found to contain an excess of complement over
amboceptors; in fact, in adding normal serum to heated immune
serum, this is done for the purpose of increasing the amount of com-
plement without adding to the amboceptors. In heating immune
serum it is the complement which is rendered for the time inactive, as
it is called, and the addition of fresh serum reactivates the immune
serum by means of the new complement contained. The addition of
fresh serum to the heated t}^phoid immune serum of a rabbit reacti-
vates the immune serum. On account of the large amount, or of the
great number, of complements contained in the serum from the blood
of a normal horse this can be used to reactivate several different kinds
of immune sera. Then again, as stated above, in the process of im-
munization, the newly formed amboceptors probably find free comple-
ments with which to unite in the serum of the animal undergoing
injection.

While the force of these and of possibly other objections can not be
denied, it would seem evident from the experiments herein reported
that the only difference between the behavior of immune serum on
the one hand and of normal serum on the other, upon dilution, was a
difference of degree and not of kind. If this is true, the theory of
complement diversion must be applicable to both or to neither.

Evidently there are only three possibilities in regard to the relative
number of amboceptors and of complements in an} 7 given serum either
there is just the proper proportion of amboceptors to complements or
there is an excess of the one or of the other of these. If the ambo-
ceptors and complements are present in just the proper proportion,
then there would be a simple relation between the undiluted and the
diluted serum. The serum would become relatively and actually less
and less potent in exact ratio to the extent of the dilution. If a given

.">', BAt n Ki"i.vn< I-MWKR OF -BI.< ..... IKRDV "i- HOO8.

number of bacteria were killed by 11 certain amount of tin- undiluted
- ruin, a certain f motion of this numtier would be destroyed hy the
same fraction of the serum. If tin- complement i" in BXOMt, evidently
the same rule would hold: tin- free compleinents would neither help
nor hinder the process. Hut if there were mi excess of ainboceptoifi,
a part of the bacteria corresponding to the amount of tlii- excess
would escape bacteriolysis, according to the theory of complen
diversion. On dilution, however, there would le relatively more l.ac-
teria killed where there are free aml>oceptor present in the undilutrd
serum, as already explained in the description of the Nei--er \Ve<-|isl>erg

theory. Aft

* . ^^

Since none of the experiments in the pre-ent investigation have IK in

made to determine whether the serum from normal hog's blood con-
tains free auilxx-eptors, and no experiments of the kind with hog's
blood seem to have been so far reported, the matter must for the
present l>e left undecided. But the as>imiption of such ti condition in
the normal serum would seem necessary to account for the results
obtained.

A specific example taken from one of the experiments already
described may serve to show the application of the theory of complement
diversion to the phenomena observed on diluting normal hog sera in so
far as the effect of dilution upon the bacterioly tic potency i- concerned.
Into the undiluted serum from the blood of hog 1851>, and into each
of the various dilutions, there were introduced before injection of the
animal 3,760 bacilli of the Crawford strain per 1 c. c. of the serum.
The number of bacilli surviving after twent3'-four hours in the undi-
luted serum was 80 per 1 c. c. There are two conditions which may
be assumed to exist in the serum, either of which would account for
the escape of the 80 bacilli, disregarding for the present the fact -Ue-
where discussed that there is a multiplication of the bacteria surviving
Inicteriolysis and assuming that the 80 Imcteria in the present c*M
represent merely those bacteria which escaped bacterioly>is. Hither
there were not enough amlweptor-coinpleiiients for all the bacteria
introduced in other words, then- were >o short of these or, if there
were enough or more than enough aml>oceptor-coniplements. t lien-
were also 80 free amlMn-eptors. Hither of these two conditions would
account for the escape of the 80 bacteria in the experiment. A simple
calculation from the result obtained on diluting the serum will show
which of these conditions is most likely to have prevailed. If there
had been a paucity of amboceptor-complenients if there had ln-en.
in other words, just 3,680 of these instead of over H.7'.n. which was
the number of bacteria added then a dilution of 1 part of the >ei um
to 9 parts of salt solution should have killed only :'.',^ bacilli per 1
\\hen these were introduced; whereas, as a matter of fact, this dilu-
tion killed :i,60o \trr 1 c. c. There would appear, therefore, no ground

CONDITIONS AFFECTING BACTERIOLYSIS. 57

for assuming the failure of bacteriolysis in the immune serum to have
been due to a paucity of amboceptor-complements in the undiluted
serum.

Suppose, now, that the other condition mentioned above existed in
the serum, namely, that there was an excess of amboceptors over and
above the amboceptor-complements, and that there were more ambo-
ceptor-complements than the bacteria introduced. A simple calcula-
tion will show that such a supposition will readily account for the
results obtained in the diluted serum as compared with the results in
the undiluted. If there were in the undiluted serum 26,000 ambocep-
tor-complements and 80 free amboceptors per 1 c. c. then on dilution of
1 to 10 there would evidently be 2,600 amboceptor-complements and 8
free amboceptors per 1 c. c. of the diluted serum. Consequently 8
bacteria would escape bacteriolysis on account of being attached to the
free amboceptors, but 2,600 would be destroyed.

The above calculations are, of course, not intended to indicate pre-
cisely the processes involved, for these are probably complicated by
the multiplication of the surviving bacteria, as mentioned above.
Still, even if such a complication is admitted, the probabilities in favor
of the condition last named become even greater, as a moment's con-
sideration will show.

A further complication may also arise from the possibility, however
remote, that the amboceptors and complements are not uniform!}" dis-
tributed in the serum that while they may perhaps usually be so
distributed, they may be found more abundantly in one part of the
serum than in another.

SUMMARY OF THE CONDITIONS AFFECTING BACTERIOLYSIS.

The conditions under which bacteriolysis would take place, and
those under which no bacteriolysis would take place in any given
serum, may be summarized as follows:

1. Complete bacteriolysis could take place only where there were
no free amboceptors and where there were at the same time a number
of amboceptor-complements equal to or greater than the number of
bacteria introduced.

On dilution in a serum of this kind there would be a loss of bac-
teriolytic power in proportion to the degree of dilution if the ambo-
ceptor-complements were exactly equal in number to the bacteria
introduced. If there were more amboceptor-complements originally
in the undiluted serum than the bacteria introduced, then on dilution
there would be relatively more bacteria destroyed. If the excess of
amboceptor-complements is large enough, there could of course be
enough present in the diluted serum to kill as many bacteria as were
killed by the serum before dilution.

'.^ BA TKKInM I I. POWJEB 'K IU."'I -MUM "I- II".

J. Partial kwetoriolysift would follow when there were fewer ainbo
ceptor- pn -cut in tin- -eruin than the bacteria int rod need ami when
at the -ame I line there were any Rmbooeptor-OOOlpleineilt0 pre-ent.

The extent of bacterioly-i- upon dilution would depend upon the
number of amlxweptor-complemcnt- present originally.

H. No hacterio|y-i- could take place if the free amboceptor- wen-
e(|iial in number to the bacteria introduced, or if they were in cxce-.-
of this number, either in the undiluted or the diluted -erum.

Tin: im.i;i . ,i DMPI.KMKNT itivi:i;.-ioN.

In spite of the serious difficulties alluded to it \\ould neverthele--
appear that the theon of coiiipleiuent diversion give- an insight into
the proce-x of bacteriolysis OOCnring in normal serum in ><> far a- thi-
i- allected by dilution in tho-e ca-e- at lea-t \\here the experiiu. ni -
>how that the serum U actually or relal ively more potent upon dilution
than when undiluted irrespective of the nature of the serum, whether
thix ix normal or immune. But this would leave the result- obtained
in certain other experiments still unexplained. It will be recalled
that in certain experiments the M-riim was found to be bactericidal
when undiluted, that it lost in bactericidal potency on dilution, but
that it irained in potency upon further dilution. It is the.e result-
which would he left unexplained by the theory of complement diver-
sion as thus far dUcuxsed. One example of this behavior amon^ other-
is to t>e seen in Table XXVI. in the serum of hog 1708, ten day-
after the inoculation of the animal with the F. :.''. strain of 11. rJ,t,1eraB
. I none of the tests there were introduced into the undiluted -ermn
and into the various dilutions in round numbers .v>o Crawford bacilli.
The result after twenty-four hour.- wa- that all the bacilli \\eiv
de-t roved in the undiluted serum, while in the dilutions of 1 to ."><' and
1 to .".uo there \\a- an increa-e of about thirty-three fold and one
hundredfold, respectively. Hut in the dilution of 1 to 5,(X>" there
were only II" bacilli renmining. In other word-, the undiluted
serum showed strongly bactericidal power, and the dilution of
1 to :..i'oo ;d-o showed considerable potency, while the dilution-
bet ween the-e two showed at Ilio-t but feeble power. If this re-lilt is

analyxed. it will be seen that the -erum appeared to act like normal
serum at the -tart when undiluted and in the lir-t dilution-, but that
on further dilution it appeared to act like immune serum. This siig-
~ that po ibly the normal ambyeeptor-comph-nient- per-i-t ing in
tin- -erum may have been more potent than the immune amboceptor-
formed in consequence of the injection, that the tir.-t eflect of dilution
wa- to reduce the potency of the normal arnU.ceptor complement-, and
that further dilution enabled the immune amlxjceptors to produce the
phenomenon of complement diver-ion. This would le ijuite in accord

with certain view- e\p|V ed elsewhere in this |Ml|>er to the ellect that

THEORY OF COMPLEMENT DIVERSION. 59

the injections of cultures or their products while they cause the for-
mation of new amboceptors leave the normal amboceptor- complements
unaffected.

In the above discussion it is assumed that the two bodies concerned
the amboceptor and the complement of Ehrlich, the sensibilisatrice
and the alexin of Bordet are capable of uniting and do actually
unite independently of the presence of bacteria or of other cells. But
Bordet 3 has recently published a series of investigations tending to
show that the experiments of Ehrlich and Sachs 5 , which constitute the
chief evidence in favor of this view, are capable of quite a different
interpi-etation from this, and that this interpretation is in fact not justi-
fiable from the results of the experiments which consisted in the
demonstration of the fact, not denied by Bordet, that ox serum will
produce cytolysis only when the serum has in it amboceptors and com-
plements simultaneously. It is not possible, as in some other cases, to
produce cytolysis by sensitizing cells with ox amboceptors that is,
with heated ox serum and, after washing these sensitized cells, add-
ing complement that is, fresh serum. Cytolysis with ox serum
takes place only when the heated ox serum and some unheated fresh
serum (horse serum was the kind used in the experiments) are
employed at the same time. This is interpreted by Ehrlich and Sachs
as showing that while the free amboceptors present in the ox serum
will not unite with the cells they will and do so unite when the}' are
previously attached to complements. But Bordet's results appear to
show quite plainly that in this case the horse serum which was used
as complement produces cytolysis quite independently of the ox serum,
and that while cytolysis takes place more promptly when heated ox
serum ox amboceptors are added, the ox serum is not necessaiy.
Bordet therefore regards the experiments as showing- that the heated
ox serum acted merely as an accelerator of cytolysis.

Bordet summarizes his conclusions as follows:

We see but one rational explanation of the peculiar action of ox serum that there
exists in the serum a peculiar substance capable of resisting heat of 56 C. and
which remains unaltered for many months in this heated serum. The substance
is probably of an albuminous or colloid character, and does not adhere to the
normal corpuscles, but is precipitated upon the corpuscles which are previously
charged with sensibilisatrice and alexin. We believe that it is a veritable process
of glueing of absorption depending upon molecular adhesion. In con-

formity with the statements of Ehrlich and Sachs, experiments show that the
corpuscles of guinea pigs become hemolyzed in a mixture of fresh horse serum and
of ox serum, the latter having been heated to 56 C., while they resist hemoly-
sis if they are first subjected to the action of the heated ox serum and have the horse
serum added subsequently. But the interpretation offered by Ehrlich and Sachs,
according to which the sensibilisatrice furnished by the ox serum does not unite
with the corpuscles unless it (the sensibilisatrice) is previously connected with alexin
derived from the horse serum is n(h correct. In the first place, the sensibilisatrice,
which plays a preponderating and most essential role, is not contained in the ox

i'l I HACTr Kl.'I.VHc 1'oUKR OF BI I' -I-.WM K H<"

-.nun ut all, hut i- furni-hed by tin- h.ir-e scrum. ( 'on-eo.ncntly theft)

> U-have lik- all <f their conveners, in tin- nonse that they do not require the
pn-scii.-e ! tin- alexin In-fore they are capable of uniting \\ ith tin- corpuscles.

Finally, this interpretation leaves completely in the dark thf MTV -^-.'ial |>ccnliari-

: the cu.ses of heinolysis in (jiutiti.m.

The (teculiarity of ox serum consist* in the prewnce of a certain element which
- ">ti ('.. ami also reeistw tanding, and is of the nature of a >lliil, doubt lew
albuminoid, and which, furthermore, in al^irlKsl by curpuwlff which have IKT..III,-
rharxi^l witli HHibilisatric- and aK-xin, but which remain^ fn-- in tin- presence of
normal OOCjMMlflt OT Oi corpusrlrs imTt'ly sennit i/c. I i , cnrpn>cli-.- tn-atvd \\ith
heated Heriun alunc. Tin* absorption of thi>rolloii| by corpuscle which have btfii
tn-att-d with lx>th wnsihilisatrire and alcxin ha the effect of energetically aggluti-
natiim them and of rendering them more siiM-eptible to liemoly.-i.- except un<l.
tain circumstance*. * The absorption of the sensiti/ed and alcxinized cor-
puscles is very likely due to molecular adhesion, the preliminary treatment h:.
niiMlilieil the corpuscles in so far a* their adhesive properties are concerned. I n-ler
them* conditions the absorption may take place inde|>en<lcntly of the upecieu of
animal from which the corpuscles are obtained; it may even take place with the
corpuscles of the name animal which furnishes the colloid, an in the cane of ox wrum.

N"\\, if the contention of Bonlct a> ><-t forth al>ov is correct, and
if tin- t\v<> lolir> conrrriH-d in cytolysis do not unite, then of cour-e
the tlu-ory of complement diversion must fall, since it requires as the
lir-t condition the union of amboceptors with complements. Never-
theless, there appears to be as yet no other explanation of the phe-
nomena observed on diluting immune serum, and. a^ an attempt ha*
IXMMI made to show in the present paper, there seems to be no other
explanation of the behavior of normal serum on dilution.

CONCLUSIONS.

In summarizing the results of the present investigations it is seen
1. That the bactericidal potency of the serum from the same hog

varies from time to time.

_'. That the serum from one and the same drawing differs in poteiu-y

for different strains of /?. chnl,>rn> xuix.

3. That while the effect of standing is to weaken the potency, this
tl. , t is more marked in some sj)ecimens of serum than in other-
under the same condition-.

4. That the bactericidal power of the serum from the venous blood
of hoefs is not ahvay* more potent than that from the arterial blood.

5. That heating the serum at about 54 C. for thirty minute*, inacti-
vates the serum.

6. That the Neisser-\Veehsl>erg phenomenon is sometimes seen in
and sometimes missed from the Mood of hogs injected with cultures of
//. cAokrM MM.

7. That the theory of complement diversion may not only account
for the behavior of immune serum. l>ut also for that of normal >erum
on dilution, or, at least, that the theory, if applicable in the one case,
is equally so in the other.

BIBLIOGRAPHY.

(1) BORDET, JULES.

Agglutination et dissolution des globules rouges par le serum. Deuxieme
memoire. Ann. de 1'inst. Past., t. 13, no. 4, p. 273-297. Paris, Apr. 25,
1899.

(2) BORDET, JULES.

Sur le mode d'action de scrums cytolytiques et sur 1'unite de Palexine dans un
meme serum. Ann. de 1'inst. Past., t. 15, no. 5, 303-318. Paris, May 25,
1901.

(3) BORDET, JULES, and GAY, FREDERICK P.

Sur les relations des sensibilisatrices avec Palexine. Ann. de 1'inst. Past.,
t. 20, no. 6, p. 467-498. Paris, June 25, 1906.

(4) BUXTON, B. H.

Bacteriolytic power of immune serum and the theory of complement diver-
sion. Jrn. of med. research, v. 13, no. 5 (whole no. 90, n. s. v. 8), p. 431-
485. Bost, Aug., 1905.

(5) EHRLICH, P[AUL], and SACHS, H[ANS].

Ueber den mechanismus der amboceptorenwirkung. Berl. klin. woch.,
jahrg. 39, no. 21, p. 492-496. Berl., May 26, 1902.

(6) ELISCHER, JULIUS v., and KENTZLER, JULIUS.

Ueber die baktericide eigenschaft des typhusserums. Berl. klin. woch.,
jahrg. 42, no. 29, p. 897-900. Berl., July 17, 1905.

(7) FRIEDBERGER, E[UGEN].

Die baktericiden sera. Kolle und Wassermann's Handbuch der pathogenen
mikroorganismen, bd. 4, teil 1, Jena, Gustav Fischer, 1904. See p. 497.

(8) See citation 7, p. 521.

(9) See citation 7, p. 561.

(10) GAY, FREDERICK P.

, The fixation of alexines by specific serum precipitates. Cent. f. bakt., abt. 1,
Orig., bd. 39, hft. 5, p. 603-610. Jena, Oct. 10, 1905.

(11) LlPSTEIN, A.

Die komplement ablenkung bei baktericiden reagenzglasversuchen und ihre
ursache. Cent. f. bakt., abt. 1, Orig., bd. 31, no. 10, p. 460-468. Jena,
Apr. 16, 1902.

(12) LOEFFLER, F[RIEDRICH], and ABEL, R[CDOLF].

Ueber die spezifischen eigenschaften der schutzkorper im blute typhus- und
coli-immuner tiere. Cent. f. bakt., abt. 1, bd. 19 no. 2|3, p. 51-70. Jena,
Jan. 23, 1896.

(13) MORGENROTH, J., and SACHS, H[ANS],

Ueber die completirbarkeit der amboceptoren. Berl. klin. woch., jahrg.
39, no. 27, p. 631-633. Berl., July 7, 1902.

(14) NEISSER, MAX, and WECHSBERG, FRIEDRICK.

Ueber die wirkungsart bactercider sera. Munch, med. woch., jahrg. 48, no.
13, p. 697-700. Munch., Apr. 30. 1901.

(16) I*rm:i;~<>v A i i

I'I-IHT <Iii> natiirlirlir !iiil/.t)r.iii<liiii!iiiiiiilat .lex hnndcfl and d hnhiu>. Gent.
f. bakt., al>t. 1, Orijf.. IN!. :, 11... - 8M, -l.-im. Apr. 4, 1903.

(16) IVRIFKKR, K[n'ii.\iti>], ami KuiKDHEtuiBK, K[I*EN].

l'clH>raiitil>akt*Ti(>Iy tint-he (antagoniti8che)8ultanzen nunnuli-rHera. Dent.
Hied. wn-h., jahty. :U, n>. 1, p. JK-S. It4-rl , .Ian. 5, 1906.

(17) PntirrKR, R[i< HAKI>], and WAMHKKMANN, A[r(;fKr].

('ntcr><iictiunp>n U!HT -hw weeen der choleraiinrnnniUit. /.twlir. f. hyg. n.
inf.-kt.. 1,1. 14, hft. 1, p. 46-63. Leipz., 1893.

(18) SACHS, HANS.

TI-UT .la.- RamnOMmHrinmt nonualer nnd iniiiiiini-atnrisrli eraenjrter amho-
/fpt.-rtMi IKM dt-r hatnnlym-. Deut. tned. woch., jahi^g. 31, no. 18, p. 706-707.
llt-rl.. May 4, 1905.

(19) Tl(iIM>lK)KK. KU IIAKD.

I'cU r di-n ulfxiii^chalt nonnaler und patholo^flcher menschlicher hlatsera.
Cent. f. bakt., abt 1, Orig., bd. 32, no. 6, p. 439-449. Jena, Sept. 25, 1902.

(20) Woi.KK, Al.KKKl).

Ueber grund)?etetze d-r immunitat. Cent. f. bakt., abt. 1, Orig., Ixl. :;7. n-.
4, j). 566-676. Jena, Dec. 12, 1904.