EQUILIBRIA IN PRECIPITIN REACTIONS.
COEXIST~.NCE OF A SINGLE FREE ANTIGEN AND ITS ANTIBODY
IN Tmr SAmESER~.*
BY STANHOPE BAYNE-~ONES, M.D.
(From the Department of Bacteriology of the Collegeof Pkysiclans and Surgeons,
Columbia University, New York.)
(Receivedfor publication,January 30, 1917.)
The vagueness of the terms of serology connotes the difficulties
with which the describers of this new knowledge have had to con-
tend. Unable to use simple and definite chemical substances, and
without the guidance of any general theory, they have had to employ
mass adjectives to describe the reactions of a serum as a whole, and
have gathered great numbers of observations without any plan for
their coordination. Whenever, therefore, it is possible to repeat
former experiments by substituting known factors in place of un-
known mixtures, the work should be undertaken. Quantitative
data will be obtained through these studies which wiU eventually
allow serology to find a place in either chemistry or physics, or what
seems almost serf-evident, in physical chemistry. The colloidal
state of matter has afforded so many analogies to the reactions of
serology that the terms of colloid chemistry have already been
applied with uncritical facility by immunologists to reactions which
have not yet been studied with quantitative accuracy sufficient
to indicate their true nature. The systematic attempt, however,
to prove or disprove these colloidal analogies in serology will un-
doubtedly result in a unifying conception of the principles of the
reactions of immunity.
The precipitation of protein by specific sera has many analogies
with colloidal reactions and this phenomenon lends itself readily to
investigation simplified by the use of at least a single pure substance.
* Aided by a grant from The Rockefeller Institute for Medical Research,
837
838 EQUILIBRIA IN PP,_ECIPITIN REACTIONS
Zinsser1has emphasized the colloidal phases of the precipitin reaction,
and recently WeiF has repeated the experiments, using crystallized
egg albumin as the antigen. One of the phases of the precipitiu
reaction which has engaged particular attention is that in which the
precipitinogen and the precipitin occur simultaneously but tmunited
in the same serum. This is easily demonstrable when whole serum
or a mixture of proteins is used as the antigen. Weft, however, using
crystalline egg albumin as the antigen was unable to demonstrate
the coexistence of free antigen and antibody in the same serum. It
is of considerable theoretical importance to confirm this finding.
The results obtained by Weil are so little in accord with expectations
based upon chemical or colloidal theories that it was considered ad-
visable to repeat the experiments on precipitin reactions with crys-
talline egg albumin. The following studies, therefore, were carried
out under the direction of Dr. Zinsser, with some modifications of
recent work on this reaction.
HISTORICAL.
In 1902Linossier and Lemoine,3Ascoli,4and Eisenberg5noticed that when for-
eign serum had been injected into rabbits in large doses the sera of these animals
contained both antigen and antibody. They used whole sera as antigens. To
demonstrate the simultaneous presence of free precipitable substance and its
precipitant in the same serum, fractions of the serum were tested, first for anti-
gen, by adding another antiserum, and second for precipitin by adding an homo-
logous antigen. Gay and Rusk6 have described the same phenomena, and by
showing that the sera do not fix complement, although they contain antigen and
antibody, they have furnished additional evidence of the ununited state of the
substances.
Zinsser and Young~have pointed out that the association of free antigen and
its antibody in serum, as shown by the failure of this serum to fix complement,
is evidence that the reaction does not take place according to the law of mass
1Zinsser, H., Infection and Resistance, London and New York, 1914, 266.
2Weil, R., J. Immunol., 1916, i, 19.
s Linossier, G., and Lemoine, G.-H., Compt. rend. Soc. biol., 1902, liv, 85.
4Ascoli, M., MUnch. reed. Woch., 1902, xlix, 1409.
5Eisenberg, P., Centr. Bakteriol., 1re Abt., Orig., 1903, xxxi, 773.
e Gay, F. P., and Rusk, G. Y., Univ. California Publications, Pathology, 1912,
ii, 59, 73.
Zinsser, H., and Young, S. W., J. Exp. Med., 1913. xvii, 396.
STANIIOPE BAYN~-JONES 839
action. According to the formula of that law, this serum should not only con-
tain free antigen and antibody, but also a certain quantity of the united complex.
To explain the inhibition of the union of antigen and antibody in such sera Zins-
ser and Young7 have drawn a strict parallel with the protective colloidal reac-
tions. These experiments show that it is not improbable that the serum pro-
teins act as protective colloids, preventing the precipitation of a protein by its
antiserum, just as gum arabic will protect colloidal arsenic sulfide against the
precipitating action of gelatin.
Von Dungern 8confirmed the findings of Linossier and Lemoine but explained
them according to a different hypothesis. He showed that whole serum consists
of a number of protein complexes, all of which cause the formation of precipitin
in the immunized animal. One portion of the protein m~xture may have more
antigenic properties than another, as a result of which partial precipitins of various
strengths are produced for the whole serum. Von Dungern believed that these
complex antigens and multiple antibodies reacted irregularly in such a way that
while the supernatant fluid of a precipitin test would contain free antigens found
in part in the original serum together with precipitins capable of flocculating the
original serum, yet the antisubstances in the supernatant fluid were not strictly
homologous and hence would not react with each other. His extensive experi-
ments have had great influence upon subsequent opinions. He worked, however,
entirely with multiple antigens and gave only inferences and analogies for the
precipitin reactions with pure proteins.
At first glance it would seem as if the experiments in the precipitation of casein
with lactoserum, reported by Miiller,9 were evidence that with purified antigens
the phenomenon of Linossier and Lemoine did not occur. The supernatant
fluids after his precipitin reactions never contained free antigen along with anti-
body. His experiments, however, are not relevant to the results of precipitin
reaction with purified protein. Mfiller used whole milk as his casein, obviously
a .multiple antigen, and the degrees of dilutions used in the tests leave such large
gaps that the zone in which antigen and antibody coexist ununited might have
been overlooked.
Weil2 confirms all the previous observations on the coexistence of free antigen
and antibody in the supernatant fluids of precipitin tests when the whole serum,
raw egg white, or any mixed protein is used. With purified egg albumin as an-
tigen, the results ~were different. He states: "If this antigen is mixed in graded
proportions with the serum of a rabbit immunized thereto, and the resulting pre-
cipitates are removed by centrifugation, the supernatant fluid never contains
both antigen and antibody; either one is present alone." The protocols of Weil's
precipitin tests allow him to draw sharp conclusions between the effect of 0.005
cc. and 0.004 cc. of the antigen, a delicacy of reaction which not all workers have
s yon Dungern, F., Centr. Bakteriol., lie Abt., Orig., 1903, xxxiv, 355.
9 Miiller, P. T., Centr. Bakteriol., 1re Abt., Orig., 1903, xxxiv, 48.
840 EQUILIBRIA IN PRECIPITIN REACTIONS
been able to attain. The crystalline egg albumin, prepared for him by Coca,
was certainly as pure as it is possible to obtain this substance. It is not stated
whether anaphylactic reactions were carried out with this particular preparation
to determine whether it was thereby free from antigen complexes which cause
reactions in animals sensitized with egg globulin. Well,2however, referring to his
earlier work with CocaIstates that he "found that if egg albumin and egg globulin
were separated by chemical means, and different rabbits were immunized to each
substance, the resulting immune sera reacted powerfully each with its own anti-
gen, and very weakly with the other antigen." This anaphylactic interrelation
of egg proteins will be discussed later. It is significant here to note the apparent
discrepancy between the results of these precipitin tests and anaphylactic reac-
tions, although one antigen was used which admittedly gives some anaphylactic
reaction with other proteins.
EXPERIMENTAL.
Sera capable of precipitating egg albumin were prepared by im-
munizing rabbits with intravenous injections of this protein. By
varying the proportions of immune serum and solution of egg albumin
in the precipitin tests, zones of dilution were obtained in which tests
for free antigen and free antibody could be made. In addition, by
testing the sera of immunized rabbits at certain intervals after the
injection of albumin, the proportions and duration of circulating
antigen and antibody could be studied.
The important material for these experiments is the antigen.
Attempts to use casein failed because of the confusing precipitates
caused by the calcium salts required in the tests with this protein.
In some experiments, pure edestin from hemp-seed was used. n The
results with this protein were definite. Because, however, of the
slight solubility of edestin, work with it is somewhat difficult. Crys-
talline egg albumin was found to be most suitable for these studies.
This substance was prepared 1'~according to the method of Hopkins
and PinkusY The whites of freshly laid eggs were carefully pre-
cipitated with ammonium sulfate and the albumin from these was
~0Weil, R., and Coca, A. F., Z. ImmunitCitsforsch., Orig., 1913, xvii, 141.
ixThe preparation was made by Dr. Harris of the Connecticut Agricultural
Experiment Station.
12The substance was prepared under the direction of Mr. Balls of the Depart-
ment of Physiological Chemistry of Columbia University.
13Hopkins, F. G., and Pinkus, S. N., Y. _Physiol., 1898-99, xxiii, 130.
STANHOPE BAYNE-JONES 841
recrystallized three times. The solution of protein was dialyzed
under toluene in the ice box to free it as much as possible from am-
monium sulfate and acetic acid. The albumin was then dried in
vacuo to a white glistening powder. The Zsigmondy "gold number''14
for the preparation was 8, showing the usual protective value of pure
egg albumin as a colloid. This preparation met the requirements
for pure crystalline egg albumin and was probably as pure as it is
possible to obtain a protein of this nature.
Anaphylactic tests were made in the hope of showing that this
preparation of egg albumin was free from globulin or substances
which had antigenic properties like globulin. Egg globulin was
prepared from the first fraction of egg white precipitated by half
saturation with ammonium sulfate. A solution of the precipitate
was dissolved in distilled water and dialyzed against distilled water
until the globulin began to flocculate. This material was centri-
fuged and the precipitate taken up in 10 per cent sodium chloride.
This formed a turbid emulsion, free, however, from large particles.
One series of guinea pigs weighing 250 gm. was sensitized by intra-
venous injections of varying amounts of crystalline egg albumin;
another series by intravenous injections of egg globulin. After
an interval of 26 days the animals were reinjected with protein
(Table I).
The results of the experiments summarized in Table I show that
preparations of egg albumin and egg globulin have a common anti-
gemic factor, capable of producing mutual anaphylactic reactions.
They confirm the statements of Wellsis and Wells and Osborne,16who,
after a long series of experiments concluded that: "In spite of the
most careful separation of these two portions of egg white by means
of ammonium sulfate precipitation, the resulting preparations each
react almost as well against the other as against itself." Weil and
Coca1 report a somewhat similar experience with crystalline egg
albumin when injected into animals sensitized to egg globulin. These
reactions have been considered in detail in order to.determine whether
or not the armphylactic test is a valid criterion of the purity of a pro-
14Schulz, F. N., and Zsigmondy, R., Beitr. chem. Phys. u. Path., 1903, iii, 137.
15Wells, H. G., J. Infect. Dis., 1911, ix, 147.
J5Wells, H. G., and Osborne, T. B., J. Infect. Dis., 1913, xii, 341.
842 EQUILIBRIA IN PRECIPITIN REACTIONS
TABLE I.
Anaphylactic Reactions with Solutions of Ovalbumin and Ovoglobulin from the
Same Eggs. Guinea Pigs, Weight 200 to 250 Gin., Injected Intravenously
on July 13, 1916. Second Injection on August 7, after an Interval
of 26 Days.*
Int**-I I
pig I Sensitizing dose. val. Second injection. Result.
No. I
Animals sensitized to crystalline ovalbumin.
6
7
8
9
[}.002 gm. ofalbumin.
0.002 " " "
[}.001 " " "
[}.001 " " "
[}.0005 " " "
[}.002 gin. ofalbumin.
[}.002 " " "
9.001 " " "
[}.001 " " "
days
26
26
26
26
26
26
26
26
26
0.1 gm. ofalbumin.
0.2 " " "
0.1 " " "
0.1 " " "
0.1 " " "
0.1 gin. of globulin.
0.1 " " "
[}.002 " " "
0.002 " " "
Convulsions. Died.
" " in 2 hrs
Typical anaphylaxis. Died
in 3 rain.
Convulsions. Died in 3
min.
Violent convulsions. Died
in 2 min.
Convulsions. Died in 3 rain.
Died in 30 min.
Diarrhea. Survived.
Sneezing, coughing. Sur-
vived.
Animals sensitized to ovoglobulin.
10
11
12
13
14
15
16
17
18
19
20
[}.004 gm. of globulin.
3.002 " " "
3.002 " " "
3.001 " " "
[}.001 " " "
3.0005 " " "
3.002 gin. of globulin.
[}.002 " " "
[}.001 " " "
3.001 " " "
[}.0005 " . . . .
26
26
26
26
26
26
26
26
26
26
26
0.1 gin. of globulin.
0.002 " " "
}.002 " " "
D.O02 " " "
0.002 " " "
D.O02 " " "
Convulsions. Died in 1 min,
" Paralysis in
hind legs. Died in 2 rain
Cough. Survived.
Convulsions. Died in 2 min
~c ~c cc ~g
D. 1 gm. of albumin.
0.1 " " "
D.1 " " "
9.1 " " "
[}.1 " " "
Cough. Survived.
Sneezing, dyspnea. Sur-
vived.
Sprawling, twitching. Sur-
vived,
Cough. Survived.
* Separate syringes and needles were used for each protein.
STANHOPE BAYN-E-JONES 843
tein, and in particular whether the anaphylactic interreactions of
egg globulin and egg albumin exclude the latter from use as a single
pure antigen. Crystalline egg albumin is acknowledged to be a
pure protein.~s.17 Egg globulin, however, probably consists of ovo-
mucin and ovalbumiu admixed in indefinite proportions,t5 The
impurity of the globulin renders the anaphylactic reactions without
significance. In the protocols of Table I, it is seen that the albumin-
globulin reactions are much weaker than those caused by the injec-
tion of the specific protein into a sensitized animal. This undoubt-
EXPERtM~NT I.
Edestin as Antigen.
This experiment was done to show coexistence of edestin and its precipitin in
the serum of an immunized rabbit.
Apr. 15-18, 1916. Serum from Rabbit 21, immunized to edestin, obtained 1
hour after an intravenous injection of 12 cc. of 0.027 per cent solution of edestin.
This serum was kept for 3 days in the ice box before being tested for its content of
antigen and antibody. Serum from Rabbit 22 also immunized to edestin, having
a precipitin titer of 1:4,000. These sera were mixed in various proportions.
The diluent was a mixture containing 0.9 per cent sodium chloride and 0.105 per
cent sodium carbonate in order to keep the edestin in solution.
Tube.
For precipitin.
!
2
For antigen.
3:
4
5
Controls.
6
7
8
9
Serum21.
Undiluted.
1:5
Undiluted.
1:100
1 : 1,000
Undiluted.
CC. CC.
0.2 0.:
0.2 0.2
0.2 i
0.2
0.2 q
q
q
0.~ q
0.2 q
Serum22. Precipitate.
~ 40
CC.
=~ +
o ++
Undiluted 0.2+ ++
" 0.2 ++ +
" 0.2  -"
0 0
" 0.2 o o
O O
1:4,000 0.2 :e +
o.-
+
++
++
+
0
0
0
+
17Schryver, S. B., General Characters of the Proteins, New York, 1909, 20.
844 EQUILIBRIA IN PRECIPITIN REACTIONS
edly indicates that the globulin preparation contained ovalbumin.
Belief in the individuality of the crystalline egg albumin as an antigen
rests upon its chemical characteristics.
Accepting, therefore, purified edestin and ovalbumin as single
antigens, the experiments were conducted upon their specific precipi-
tin reactions.
EXPERIMENT II.
Edestin as Antigen.
Apr. 20, 1916. Various quantities of edestin were added to its antiserum to
find the zone of dilution in vitro in which antigen and antibody occur in super-
natant fluids after the primary precipitin reaction.
Serum from Rabbit 22, immunized to edestin, having a titer of 1:2,000.
0.02 per cent solution of edestin mixed in various proportions with undiluted
Serum 22.
Tube.
I
2
3
4
5
6
7
8
9
Controls.
10
11
Supernatant fluids.
Serum 22,
undiluted.
co.
0.2
0.02 per cent
solution of
edestin.
2
Chloride-
carbonate
mixture.
Precipitate
after 24 hrs.
in ice box.
++
For aut~en.
0.2 ce. of
supernatant
fluid plus 0.3
cc. of Serum
22.
+- ++
Forprecipldn.
0.2 cc. of
supernatant
fluid plus 0.3
cc. of edestin
solution.
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
1
0.5
0.2
0.1
0.01
0.001
0.0001
0.00001
0.2
0.3
0.6
0.7
0.8
0.8
0.8
0.8
0.3
0.3
++
+++
+++
++
+
+
0
+=
+-
+
Tr.
?
0
o
0
O
++
-4-
+
+
+
++
++
++
++
O
This experiment demonstrates the phase in the precipitin reaction
in which a single antigen occurs simultaneously but tmunited with
its homologous antibody. In the table, the shaded columns overlap
each other through the zone in which edestin and antiedestin wer~
found free in the same fluid.
STANHOPE BAYNE-JONES 845
EXPERTM~NT HI.
Crystalline Egg Albumin as Antigen.
June 15-16, 1916. Various quantities of crystalline ovalbumin were added
to its antiserum to find the zone of dilution in vitro in which antigen and anti-
body occur in the supernatant fluids after the primary precipitin reaction. Serum
from Rabbit 23 immunized to crystalline egg albumin, having a titer of 1:10,000.
5 per cent crystalline egg albumin in normal salt solution was added in dimin-
isking amounts to undiluted Serum 23 according to the following table.
Tube.
1
2
3
4
5
6
7
8
Controls.
9
10
11
Serum 23,
undiluted.
t:C
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
5 per cent
crystalline
egg albumin.
cc.
0.05
0.01
0.005
0.001
O.00075
O.0005
0.0001
0.00005
0.5
0.5
Normal salt
solution.
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
Precipitate
after 18 hrs.
in ice box.
++
+++
+++
++++
+++
++
+
Supernatant fluids.
For ~t~em Fcrprec~piti~.
0.2 cc. of
0.2 cc, of I superaatant
superna~an~
* fired plus 0.2
flmd plus 02] of cr st "
" cc. y al-
oe. of Serum [llne albumin,
.o. Idiluted I:i0.
I
+++
++
++
+
O
O
O
O
+
+
+
+
++
++
++
++
+++
+++
++++
O
O
The overlapping shaded areas in the above table show the zone in the super-
natant fluids in which free antigen and antibody (crystalline egg albumin and
its precipitin) definitely coexist.
EXPEP ~NT IV.
Crystalline Egg Albumin as Antigen.
This experiment was done to show coexistence of antigen and antibody in the
circulation of an immune animal.
June 2, 1916. Serum from Rabbit 23, immunized to crystalline egg albumin,
having a titer of 1:5,000 before the last injection of antigen. 10.45 a.m. 3 cc.
of 5 per cent solution of crystalline egg albumin were injected intravenously into
Rabbit 23. 11.30 a.m. Bled from carotid 15 ce. 3 p.m. Tests were carried
out as follows.
846 EQUILIBRIA IN PRECIPITIN REACTIONS
Tube.
For antigen in Se-
rum23 B.
1
Forprecipitinin Se-
rum 23 B.
2
3
4
Controls.
5
6
7
8
9
Serum 23A'Serum 23B
before after
injection, injection.
$ per cent solution
of crystalline egg
albumin.
Normal
salt solu-
tion.
Dilution. Amount.
cc. cc. cG. C6.
0.2 0.2
0.2 Undilut- 0.2
ed.
0.2 1 : 10 0.2
0.2 1 : 100 0.2
Normal
rabbit
scru~.
0.2 o
0.2 o
0.2 0.2 0.2 o
0.2 0.2 0.2 o
Undilut- 0.2 0.2 o
ed.
10 " 0.2 0.2 o
11 0.2 0.2 o
Precipi-
tate after
i hr. at
37C.
++++
/++
//
o
Tube 1 shows that the serum of Rabbit 23 contained antigen un-
united with its antibody 45 minutes after an injection of crystalline
egg albumin. Tubes 2, 3, and 4 show that during this time free
precipitin was still present in the blood of the immune animal.
In several experiments conducted like this one, similar results were
obtained, showing that the sera of rabbits immunized to egg albumin
contain uncombined antigen and antibody in the circulation for ap-
proximately 48 hours after the last injection of the antigen.
In many of the precipitin reactions with crystalline egg a]bumin
the maximum of precipitation occurred only when the albumin so-
lution was diluted from 500 to 1,000 times. This phase of the pre-
cipitin reaction is commonly known as the prozone, and is most read-
ily explained as a phenomenon of colloidal relationship. In this
particular case it seemed desirable to investigate the inhibition of
precipitation effected by the albumin preparation, to relate it, if
possible, to the protective influence exerted by this emulsoid on a
sensitive gold sol. To show this effect, Experiment V was per-
formed.
STANHOPE BAYNE-JONES 847
EXPERIMENT V.
Protective Action of Egg Albumin.
June 1, 1916. This experiment was done to show the protective action of a
solution of crystalline egg albumin upon the precipitation of human serum by
its specific antiserum. To the serum from Rabbit 24, having a titer of 1:10,000
against human serum, various quantities of a 5 per cent solution of crystalline egg
albumin were mixed with human serum. After 1/2hour, anti-human serum from
Rabbit 24 was added to these mixtures, as follows. The gold number of this
prepaxation of egg albumin was 8.
Tube.
Human serum. Anti-human
Serum24.
Dilution. Amount. Dilution. ] Amount.
1 Undilut- 0.2 Undilut- 0.2
ed. ed.
2 " 0.2 1:5 0.2
3 I :100 0.2 1:5 0.2
4 1 : 1,000 0.2 1:5 0.2
5 1:5,000 0.2 1:5 0.2
6 1 : 10,000 O.2 1 :5 O.2
7 Undilut- 0.2 0.2
ed.
8 " 0.2 1:5 0.2
9 1:100 0.2 1:5 0.2
10 1 :100 0.1 1:5 0.1
11 i1:100 0.1 1:5 0.1
Controls.
12 Undilut- O.2
ed.
13 " 0.2
14 0.2
15 0.2
I$per cent
solution of Normal Precipitate
crystalline salt solu- afte~ 1/2hr.
egg s.lbu- tion. st 37"C.
rain.
0.2 ++++
0.2 +++
0.2 ++
0.2 +
0.2 +
0.2 o
0.2 ++++
0.2 ++
0.2 -4-+
0.2 0.2 "-
0.4 o
0.2 o
0.2 o
0.2 o
0.2 o
These tests (Tubes 7 to 11) show that the presence of sufficient egg albumin
prevents the flocculation of human serum by its antiserum. This effect of the
solution of egg albumin is in accord with its well known pro'tective action on other
colloids, particularly its protection of a sensitive gold sol. against precipita-
tion by electrolytes. The protective action in the above tests is obviously
stronger than that exerted by the sera and may explain the long prozones com-
monly found in precipitin reactions with solutions of crystalline egg albumin.
The hypothesis of yon Dungem s is insufficient to explain the fact
that a serum containing both antigen and antibody, though clear at
first, undergoes spontaneous precipitation upon long standing. 1 As
848 EQUILIBRIA IN PRECIPITIN REACTIONS
this precipitation progresses the amount of antigen and antibody
gradually diminishes. This phenomenon is a matter of common ob-
servation when the sera of animals immunized to a mixture of protein
are studied at successive intervals after an injection of the proteins
used as antigen. Since it was shown by Experiments I and IV that
a simple antigen such as edestin or crystalline egg albumin remains
ununited with its homologous precipitin in the circulation of an im-
mune rabbit for at least 24 hours after the last injection of the pro-
rein, a serum was readily obtained in which the slow spontaneous
union of a single antigen with its antibody could be observed. This
series of tests was carried out as follows.
EXPER~ENT VI.
Spontaneous Precipitation in a Serum Containing Antigen and Antibody (Crys-
talline Egg Albumin and Its Precipitin).
July 27, 1916. Rabbit 25 immunized to crystalline egg albumin, having a
serum with a titer of 1:25,000, was injected intravenously with 3 ce. of a 5 per
cent solution of crystalline egg albumin. 1 hour later blood was drawn asepti-
cally into sterile tubes. Serum obtained from this was divided into several lots
and stored in sterile tubes in the ice box. Titrations for antigen and antibody
were made as follows.
5 per cent solution Precipitate.
Serum Serum of crystallineegg Normal
Tube. 25 after 25 be- albumin, salt Normal
injec- fore in- solu- rabbit After !
tion. jection, tion. serum. 1/2hr. at i After 24 hrs. in
Dilution. Amount. 37C" / ice box.
CC. CC. CC. CO.. CO.
For antigen, i
1 0.2 0.2 +++ ++++
For precipitin.
2 0.2. 1 : 10 0.2 ++ ++
3 0.2 1 100 0.2 + +
4 0.2 1:500 0.2 + +
5 0.2 1 : 1,000 0.2 + -{-
6 0.2 1 : 5,000 0.2 + +
7 0.2 1 : 10,000 0.2 o ~-
Controls.
8 0.2 I 0.2 o Slight precil>
Rate (spon.
taneous).
9 1 : 10 0.2 0.2 o o
10 1 : 10 0.2 0.2 o o
11 0.2 0.2 o o
12 0.2 I 0.2 o o
13 0.2 1:25,000 il 0.2 -b +
STANHOPE BAYNE-JONES 849
Tube 1 shows that the serum contained a considerable amount of antigen,
while Tubes 6 and 7 show that the precipitin titer of the same serum equalled
1:10,000. Tubes 8 to 13 were used to show that no non-specific precipitation
occurred in any of the components of the primary reaction. With the subsequent
tests, this series of controls was repeated invariably. Their results were always
as anticipated and they will be omitted in detail from the following protocols.
The batches of Serum 25, kept in the ice box, showed definite spontaneous
precipitation at the end of 24 hours.
July 28. Titration of the antigen and antibody content of this serum was
carried out as follows.
Tube.
Serum 25 after injection.
For antigen.
1
2
3
4
5
6
For precipitin.
7
8
9
Dilution.
cG.
Undiluted. 0.1
1:10 0.1
1:100 0.1
1:250 0.1
* 1:500 0.1
1 : 1,000 011
Undiluted. 0.1
" 0.1
" 0.1
0.1
0.1
0.1
0.1
0.1
0.1
5 per cent solution
of crystalline egg _ _
albumin.
Dilution.
1 : 1,000 0.1
1:5,000 0.1
1:10,000 0.1
Precipitate.
After 24
,*,~ hrs. in ice
~o
~ box.
++ ++++
+ +++
+ ++
+ +
o -~-
O o
+ +
O O
This series of tests shows that the quantity of antigen present in Serum 25
(after injection) was sufficient to give a visible reaction when diluted 500 times.
At the same time the predpitin titer of this serum containing free antigen was
1:1,000. In the previous 24 hours, however, while spontaneous precipitation
had occurred in the serum, the precipitin titer decreased from 1 *5,000 to 1 : 1,000.
Spontaneous precipitation continued, and at the end of 48 hours the following
proportions of egg albumin and its precipitin were found to be present.
July 29. Serum 25 titrated as follows:
850 EQUILIBRIA IN PRECIPITIN REACTIONS
Tube.
For antigen.
1
2
3
4
5
For precipitin.
6
Serum 25
after injection.
Dilution. I Amount.
$$.
Undilut- 0.1
eel.
1:10 0.1
1:100 0.1
1:250 0.1
1:500 0.1
Undilut- 0.1
ed.
" 0.1
" 0.1
Serum 25
before in-
jection.
~.
0.1
0.1
0.1
0.1
0.1
5 per cent solution of
crystalfine egg
albumin.
Dilution. Amount.
cc.
1:500 0.1
1 : 1,000 0.1
1 : 5,000 0.1
Precipitate.
After Mter
1/2hr. at 24 hrs. in
370C. ice box.
++ +++
+ ++
 +
o +
O o
+ +
O o
O O
July 31. 96 hours after the serum had been obtained, the tubes were turbid
with considerable flocculated material in the sediment. The turbidity disap-
peared on warming the serum, but the sediment remained, giving evidence of the
specific nature of this spontaneous precipitate. Titration of the serum at the end
of 96 hours gave the following values for its content of free antigen and antibody.
Tube.
For antigen.
1
2
3
4
5
For precipitin.
6
Serum 25 after in-
jection.
Dilution. Amount.
#.
Undilut- 0.1
J ed.
1:10 0.1
1 : 150 0.1
1:200 0.1
1:300 0.1
Undilut- 0.1
edl o. 1
Serum 25
before in-
jection.
C.
0.1
0.1
0.1
0.1
0.1
5 per cent solution ot
crystalline eggalbumin.
Dilution. I Amount,
cc.
1:100 0.1
1:350 0.1
Precipitate.
After ~ hr. Mter 24
hrs. in ice
at 37C. box.
+++ +++
++ ++
+ +
+ +
o o
 +
o o
This shows that during the course ~f spontaneous precipitation n the serum,
both egg albumin and its precipitin became decreased. At the end of 96 hours
the antigen titer had fallen to 1 : 200 and the precipitin titer to 1 : 100.
STANttOPE BAYNE-JONES 851
Subsequent titrafions were carried out to follow the decrease of these anti-
substances coincident with progress of spontaneous precipitation in their serum.
At the end of 144 hours the serum no longer contained precipitin, while the
antigen titer had decreased to 1 : 100.
At the end of 196 hours the precipitin was still zero, while the antigen titet
remained constant at 1: i00.
The results of these titrations of the supernatant serum during spontaneous
precipitation are summarized in Table II, and represented graphically in the curve
of the reaction (Text-fig. 1).
TABLE II.
Summary of Experiment VI.
Hours. Precipltln. Antigen.
2
24
48
96
144
196
1 : 5,000
1:1,000
1:500
1:100
0
0
?
1:500
1 :250
1:200
1:100
1:100
The chart of the observed amounts of antigen and antibody remain.
ing dissociated in the supernatant fluid of the precipitin reaction
suggests at once that the process has taken place according to a defi-
nite law. As the amounts of antigen and precipitln are stated only in
terms of the dilution of mixtures whose protein content was not accu,
rarely estimated it is not possible to analyze this curve in detail. Ex-
periments are being undertaken, therefore, to determine quantitatively
the mutual relationship of the factors of this reaction. While, how-
ever, calculations are not possible which would reveal the nature of
this reaction, it is permissible to point out that the curve is not un-
like that of some colloidal precipitations. In this case, the amount
of precipitate together with the amount of precipitin remaining in
solution in the presence of the precipitable substance is apparently a
function of the concentration of the precipitable substance; e.g.,
crystalline egg albumin, the colloidal properties of which are in-
disputable. Although the experiment was not conducted upon a
quantitative basis, the regularity of the curve of its results is striking.
From this feature it is a fair assumption to suppose that the reacting
852 EQUILIBRIA IN PRECIPITIN REACTIONS
antibodies were strictly homologous, and that the absence of irregu-
larity in the curve is evidence that the antigen was a relatively
simple substance. This confirms the chemical evidence of the purity
of the preparation of crystalline egg albumin used as antigen.
~IITIB[N PRECIPITIN
500 5,0O0
I
i,.
t
t
t
I
t
I
I
400 4,000
300 3,000
zoo ZDoo
!oo t,ooo -..
x..
HOUR5 20 40 60 80 I00 IZO 140 160 I~0 200
Antigen (eggalbumin).
..... Precipitin.
TExT-FIa. 1. Graphic representation of the results of Experiment VI, show-
ing the changes in the quantities of antigen and antibody during spontaneous
precipitation in a serum containing these substances. The quantities are ex-
pressed by titration values.
SUMMARY.
1. In these studies several phases of the precipitin reactions were
investigated by the use of purified proteins as antigens. These prep,
arations were edestin from hemp-seed and crystalline ovalbumin
from fresh eggs. The ovalbumin, isolated by the method of Hopkins
and Pinkus~ was apparently as pure as is obtainable by chemical
means. This albumin, however, produced moderately severe ana-
phylactic reactions in animals sensitized with ovoglobulin. Anaphy-
lactic tests of the individuality of a protein cannot be any longer
regarded as the criterion of the purity of the substance as an antigen.
STANIIOPE BAYN-E-~ONES 853
Wells and Osborne~ehave shown that proteins of considerable chem-
ical difference may have a common antigenic group which causes
mutual anaphylactic reactions in animals sensitized to these proteins.
In particular, as egg globulin is a mixture of proteins, one of which
is undoubtedly egg albumin, almphylaxis produced by injections of
albumin into animals sensitized to the so called globulin offers no
evidence for or against the purity of the albumin. The character
of the curves shown in Text-fig. 1 confirms the assumption, based
upon chemical data, that crystalline egg albumin is a single protein.
2. With edestin and crystalline egg albumin as antigens, phases
in the precipitin reaction were found in which these substances and
their specific precipitins could be demonstrated to be coexistent
but ununited in the same serum.
3. When edestin or crystalline egg albumin is injected into a rabbit
immunized thereto, the antigen may be found in the circulating blood
during 48 hours after its injection, while at the same time the animal
maintains a high titer of free precipitin in its blood.
4. When the pure protein antigen is mixed in proper proportions
with the serum of a specifically immunized rabbit and the resulting
precipitate removed by centrifugation, the supernatant fluid con-
tains both antigen and antibody.
5. The serum drawn from a rabbit during the period in which free
antigen and antibody are coexistent in the circulation undergoes
slow spontaneous precipitation when kept in sterile tubes in the ice
box. The curve of this reaction is reproduced as Text-fig. 1. The
relationships of the parabola indicate that the interaction of antigen
and antibody takes place according to a definite law. When suffi-
cient quantitative data are obtained to allow an analysis of this
curve, the formulas for this reaction will undoubtedly throw light
upon the chemical or physical nature of the process.
6. The protective action of the solution of egg albumin as a third
colloid preventing precipitation in a reaction between human
serum and its antibody was readily demonstrated. This observa-
tion and the constancy of the long prozone in precipitin test with
egg albumin are in accord with the protective action of ovalbumin
upon colloidal gold.