AN MPMSAL GP SEROLOCEQAL WW‘
AS APPLIED T07 ENTOMOLGGiCAL TAXONOMY
WSTH SPECIAL REFERENCE TO 'fHE HYMENOPTERA’

Thai; for tho Dogm of Ph. D.
MICHIGAN STATE UNIVERSITY
George B. Noland
19552

 

as

This is to certify that the

thesis entitled

An Applaisal of Ser010‘ical Technique
as Ayplied to Entomolocical
Taxonomy with Special Reference
to the Hymenoptera

presented bg

George B. Noland

has been accepted towards fulfillment

of the requirements for

#2— degree in m

 

”Ki 7%

Jor professor

Date August 10, 1955

 

0-169

 

Y!..—.~_

 

 

‘ll«ii.llllllliillllilliillllllll J

AN APPRAISAL OF SEROLOGICAL TECHNIQUES AS APPLIED TO

INTOMOLO GICAL TAXONOMY WITH SPECIAL

REFERHVCE 10 THE HWENOPTERA

by
George B. Noland

Jichigan
and Applied Science
in partial fulfillment of th

e requirements
for the degree of

mCTOR OF PHILOSOPHY

Department of Entomology

Year 1955

E 3‘ \ 5
Approved {32.2f(*./Na/’}Kln .-‘]_:'{L1/(rr/ "1

 

[[[[[!l[[[[Il-lllllll_l[' IHII.'III

 

George B. Noland Abstract

An appraisal of the methods of precipitin testing as
applied to entomological systematics was made. The ring
test, turbidity test and Petri plate methods were considered
and their specific application to entomological material de-
termined.

'Ihe Petri plate method of Ouchterlony, as modified by
using glass rings to contain the reactants, was shown to be
the most satisfactory technique, utilizing an agar base. In-
dividual antigens, from extracts of various species of fly-
menoptera, were evaluated as taxonomic aids and their sim-
ilarity or dissimilarity to antigens from other extracts
were studied.

The importance of the various procedures employed in
the preparation of both antigens and antisera was shown. It
was determined that a number of insects, equal to the vol-
ume of five hundred honey-bee thoraces, must be gathered in
order to conduct a successful test. Several rabbits must be
injected with each insect extract in order to insure the
production of a satisfactory volume of antiserum.

The nitrogen content of each antigen was determined so
that corresponding amounts of the different antigens could
be tested.

While the results obtained by the various precipitin

methods could not be considered as final, they did indicate

lllll,l.‘

111'

-2-

George B. Noland Abstract

the possibilities of the techniques. Studies relating to
possible origin of the bee families, relationship within a
genus, relationship within a family, and interrelationship
of various families of Hymenoptera were presented.

It was determined that the collaboration of a systemat-
ist and a serologist would be essential in any extended
study. Application of the ring test, turbidity test, and
glass ring modification of the ouchterlony test were all
considered essential to a satisfactory and signficiant in-
vestigation of any problan involving insect relationships.

x. . ‘. .1. u. 1.. .2 I ‘1 x. ‘11

Ar! 1‘.[!Illdl‘lllilll

AN APPRAISAL OF SEROLOGICAL TECHNIQUES AS APPLIED T0
EN'IOMOLOGICAL TAXONOMY WITH SPECIAL
REFERENCE '10 THE HYMENOPTERA

by
George B‘.“ Noland

A THESIS

Submitted to the School ofAdvanced Graduate Studies of Michigan
State University of Agriculture and Applied Science
in partial fulfillment of the requirenecnts
for the degree of

1300le OF PHILOSOPHY

Department of mitomology

1955

 

"Inc
to Profs
05?, and
and Publ
carried
a “ms:

nu
‘Ai

 

ACKNOWLEDGMENTS

The author wishes to express his sincere appreciation
to Professor Ray Hutson, Head of the Department of Entomol-
ogy, and to Dr. Walter Mack, of the Department of Microbiology
and Public Health, under whose.directions this research was
carried out. To Dr. Herman King is extended the gratitude of
a struggling student, for the advice and good counsel extended
freely during the course of research and graduate studies.

Thanks are also due to Dr. R. L. Fischer for assistance
in the collecting and identification of the insects used in
this study. The assistance and general camaraderie of vari-
ous graduate students in the departments of Entomology and
Microbiology and Public Health aided.materially the conduc-
tion of this study. The helpful criticism of the manuscript
by Professor'Hutson and Dre. Mack, King and Fischer is ac-
knowledged with thanks.

Finally, the author acknowledges the assistance of his
wife in the preparation of the manuscript and also her tol-

eration of a "part time" husband over the past three years.

George Bryan Noland
candidate for the degree of

Doctor of Philosophy

Final examination, August 10, 1955, 10:00 A. M., Room 244,
Natural Science Building

Dissertation: An Appraisal of Serological Techniques as
Applied to Entomological Taxonomy with
Special Reference to the Hymenoptera

Outline of Studies

Major subject: Entomology
Other subjects: Serology, Botany. Zoology

Biographical Items
Born, March 7, 1926, Fort Ogden, Florida
Undergraduate Studies, University of Detroit, 1946-50

Graduate Studies, University of Michigan BiologicalSta—
tion, 1950, University of Detroit, 1950-52, Michigan State

University, 1952-55

Ebcperience: Teaching Fellow, University of Detroit,
1950-52, Graduate Assistant, Michigan State
University, 1953-55, Member United States

Navy, 1945-46
Member of Society of the Sigma Xi

TABLE OF CONTENTS

Page

INTRODUCTION.................... 1
BISTORICALBACKGROUND................ 6
Development of the Precipitin Technique . . . . . . 6
Applications of Precipitin Testing . . . . . . . . 9
Applications of Precipitin Testing to Entomology 12
mmrnsmnrraons................ so

Preparation of Antigen Extracts . . . . . . . . . . 20
PreparationofAntisera.............. 25
InjectionSchedules........... .. .. 26
BleedingTechniques............... 35
TestingMethods.................. 56
TheRingTest.................. 56
Turbidity Measurenents . . . . . . . . . . . . . 59
TheOudinTechnique..........I..... 40
ThemekTechnique............... 42
The Ouchterlony Technique . . . . . . . . . . . . 44
PRESENTATION AND ANALYSIS OF DATA . . . . . . . . . . 49
AntigenExtracts................. 49
AntiseraProduction................ 52
ResultsofTesting................ 55
Reactions in aLiquid Medium . . . . . . . . . . 55

we Ring T9813 0 o o o o 0 o o o o o o o o o o o 55
TurbidityTesting. . . . . . . . . . . .

iv

Reactions in Agar . . . . . . . . .
TheOudinTechnique............
The Elek Petri PlateMethod. . . . . . . . . . . 80
The Ouchterlony Technique . . . . . . . . . . . . 82
DISCUSSION......................101
SUWARYANDOONCLUSIONS................110
LITERATURECITED...................112

 

Illllll .

l_l.[llt .

. 1 All...
Ali—I.[IIIIII‘IIIIIIIIIII I

TABLE
I .

II.

III .

V.

VI.

VII.

VIII .

K.

II.
III.

XIII.

XIV.

LIST OF TABLES

Two Kinds of evidence bearing on animal rela-

tionships
Insects which have been used in serological
studies ,W,
Insects used in this study . . . . . . . . .
Injection schedules . . . . . . . . . . . .
Method of antigen dilutions . . . . . . . .
Typical results of turbidity test . . . . .
Antigensmade
Antisera produced . . . . . . . . . . . . .
Ring test results with antiapis serum 96B .
Turbidity readings of antiapis serum with
homologous antigens . . . . . . . . . . .
Summation of homologous Apis turbidity tests
Turbidity readings of antiserum 26B with
heterologous antigens . . . . . . . ... .
Turbidity readings of antiapis serum and
heterologous antigens . . . . . . . . . .
Heterologous reactions other than with anti-
.apisserum................
Antiserum-antigen systems used with Elek
method..................

Antiserum-antigen systems used with Ouchter-
lonymethod

Page

15

22

29

41

54
5'7

59
66

67

69

'74

81

84

 

LIST OF FIGURES

FIGURE

1.

2.

3.

4.
5.
6.
'7.
8.
9.

10.

11.

12.

13.

14.

15.

Diagram showing the relationships of four Am-
phibia based on ring tests . . . . . . . . . .
Diagram showing the interrelationships of common
mammals, based on the ring tests . . . . . . .
Diagram representing the relationships of feur
families of Orthoptera, based on turbidity
testing . . . . . . . . . . . . . . . . . . .
Rabbit box . . . . . . . . . . . . . . . . . . .
Results of a typical ring test . . . . . . . . .
DiagramofOudinmethod . . . . . . . . . . . .
Smutlnek method . . . . . .... . . . . . . . .
TheOuchterlonymethod . . . . . . . . . . . . .
Glass ring modification of Ouchterlony tech-
nique . . . .... . . . . . . . . . . . ....
Setup of Ouchterlony method using paper discs .
Antiserum BBB x MIL, GHFAo, 4ACF and GHFA5 . . .
Possible positions of reactants in Petri plate .
Ouchterlony method, antiapis serum 56B x homol-
ogous antigens . . . . . . . . . . . . . . . .
Ouchterlony technique, antiserum 96B 3: CHFAS,
CHFA6,Q-IFA’7,CHFA6
Ouchterlony method, antiandrena 52B 1: 2ACP, MIL,
3AOF,aninFA8

Page

11

11

19

2'7

42

45
46

85

86

87

89

91

93

 

FIGURE

16.

1'7.

18.

Ouchterlony method, antiapis 568 x END, 6AGF,

m.and<r1rA7...

Ouchterlony method, antiapis 96B 1: GHFAV,

mFm O O O O O O O

Ouchterlony method, antigen (111MB 1 antisera

2513, 3213, 56B and 96B

711

Page

94

9'7

98

 

_ -—-——— ——.——-

1'1
illi‘
III-sill
{til
{ll
‘1‘“ {ll Ill“
,l
[[‘I‘
[it
[1!

 

INTRO IIICTION

The development of man's knowledge concerning the rela-

tionships of animals has closely paralleled the development

of biology itself. For a very long time any description and

discussion of animals was limited to external morphology

alone. Gradually the information gained by dissection and

observation of internal anatomy was added to the overall

picture. With the development of the microscope a new area

was added to what was essentially comparative anatomy.
Later the findings of the allied sciences, such as
physiology, embryology, paleontology, biogeography, ecology

and biochemistry, were considered in classification and re-

lationships. As each new bit of information appeared, man

added to his knowledge and in many cases altered or modified

his conceptions of animal life.

The development of each new branch of science helped man

to clear up points of misinformation and ignorance.

This is
as it should be.

In most groups of organisms there are

still many gaps in our knowledge. As new sciences and new

applications of old methods are applied, we may expect to
fill at least some of these gaps.

Within the last fifty years a relatively new science

has been established and is beginning to furnish information

.‘VA-fl—l

 

[lllili‘li ‘li [ [tilllnlllii {[[q_ll[.Ivl[‘l[-l I

about organisms and their relationships. This science is
serology, especially as applied in systematic serology.

The theory of systematic serology implies that sero-
logical correspondence depends on biochemical correspond-
ence, and that the biochemical correspondence of pro-
teins is an expression of the genetic natures of the or-
ganisms producing them. . . . The relation between se-
rological correspondence and genetic relationship is
supported by the fact that closely related species al-
'ways ShOW'a high degree of serological correspondence,
and that species hybrids fall in between their parental
species. . . . If the precipitin tests give results in
accord with known genetic relationships then we have a
right to assume that_they apply equally to species whose
genetic and sysiamatic relationships are not sufficient-
ly established.

With the foregoing in mind (see Table I) , the author
began studies on the principles and methods of systematic
serology with reference to their possible applications to
entomological taxonomy with special reference to the Hymenop-
tera.

At this point it mw be well to review some of the basic
concepts and nomenclature of systematic serology.

That branch of biology dealing with serum, its nature
sand complexities is called serology. Serum is the fluid por-
ixion of the blood with the clotting mechanism and particulate
matter removed. When a susceptible animal is injected paren-
terally with an antigen, usually a foreign protein, there
are produced in the blood and other body fluids substances

which will, under suitable conditions, react with the antigen

1Alan A. Boyden, ""Fifty years of Systematic Serology,"
Systanatic Zoology. _2_ (1): 19-30. 1955.

TABLE 13

m KINDS or EVIDENCE BEARING 0N mom. RELATIONSHIPS
-._Basis of Comparison Comparative Serology , Comparative Anatomy

 

1. Correspondence Guaranteed in the re- Relation of structural
actions with protein homology difficult to

 

antigens. delimit.
heterologous .
2. Objectivity Relative 1101,1010 ous Relative Similarities
areas more objec ive. and differences less
. objective.
3. Facility of Relative curve areas Relative morphological
measurement are easily measurable. indices more difficult

to measure and subject
to errors of interpre-
tation.

4. Constancy Relatively constant May be greatly varied
during the life cycle during the life cycle.
(checks and balances Some morphological
in antibody-forming characters remarkably
mechanisms). Moderate- conservative in evolu-

ly conservative in tion.
evolution.
5. Convergence Rare. Common.

6. Relative place- Apparently consistent. Not always consistent.
ment

'7. Primitiveness No direct test for Sometimes more primi-
primitiveness. tive conditions may be
distinguished from more

advanced conditions.

8. Fossil record Practically nonex- Sometimes well-docu-

istent. mented.

9. Requirement for Apparatus for quanti- Sometimes no compli-
special equip- tative precipitin cated apparatus needed,
ment testing needed. but commonly micro-

scopes are required.
10. Testing proce- Dynamic. Static when descriptive
dure static or alone. Dynamic when it
dynamic becomes experimental.

W

2A. A. Boyden. Ibid.

‘II!l[[{{i[{[[_{[t[[[I I I‘ll-Instill: ll

to produce an observable phenomenon. These substances are
referred to as antibodies. Sera containing antibodies are
called immune sera or antisera. The animal in question may
be a rabbit, chicken, guinea pig, horse or other suitable
subject. The animal of choice depends on the situation and
reacting system. The serum of a rabbit injected with honey
bee (A213 mellifera Linnaeus) antigens could be referred to
as an anti-apis rabbit serum.

The presence of antibodies in an antiserum can be de-
tected in various ways. Of these only the modifications of
the precipitin test will be considered in any detail. The
technique of mixing a soluble antigen with an antiserum in a
suitable container and observing for a visible precipitate
is called the precipitation technique. In the ring 5933, a
suitable antigen is carefully layered over an antiserum so
that there is a distinct interface between the two liquids.
The appearance of a cloudy layer at this junction consti-
tutes a positive test. If the precipitate resulting from
the antigen-antibody mixture is resuspended in solution and
measured with a suitable photoelectric device, the method is
called a turbidity test. A modification of the ring test
wherein the antiserum is mixed with agar and allowed to so-
lidify, then overlaid with antigen, is the Oudin technique.
A further modification in the precipitin technique is per-

formed by letting the antigen and antiserum diffuse towards
each other through an agar medium in a Petri dish.

5

If an antiserum is tested against the specific antigen,
the result is a homologous reaction. If a reaction occurs
with a non-specific but related antigen, it is a heterolo-
593g reaction. The property of serological reactions where-
in the homologous reaction is always stronger than the het-
erologous reaction is called"serological specificigy.”

It has been the object of this thesis to investigate
the above mentioned concepts and techniques and to make an
appraisal of these methods with special reference to the
problems involved in their application to entomological sys-

tanathS.

HISTORICAL BACKGROUND

Development of the Precipitin Technique

The application of serology to systematics followed
closely the development and modifications of the precipitin
technique itself. The works of Chester (1957) on plants,
and Boyden (1942) on animals, review and examine the signifi-
cant studies. No attempt will be made here to exhaust the
literature on the subject, but rather the development and
highlights of precipitin testing will be traced.

The precipitin reaction as such was discovered by Kraus
(1897) , who injected goats with sterile mixtures of cholera,
typhoid and plague bacteria. He then mixed the serum from
the goats with a filtrate of the bacteria injected and ob-
served a precipitate in the tubes. Each species of these
bacteria caused antibodies to be formed in the serum, which
when mixed with its homologous antigen gave rise to a visible
precipitate. No reaction was seen when mixed with a heterol-
ogous antigen. Thus in the original observations the anti-
sera were completely specific. This discovery was followed
by numerous studies with bacteria.

That organisms other than bacteria could induce the
formation of antibodies was discovered independently by Bor-
det (1899) and Tschistovich (1899). Bordet found that injec-

tions of fowl serum would cause a rabbit to produce antifowl

antibodies which would also react with pigeon serum. This
was the first indication that the reaction was not absolute-
ly specific. Tschistovich injected a variety of animals
with eel serum and obtained antibody production. He also
injected horse serum into a rabbit and tested it against
both horse and ass serum. In 1901, Uhlenhuth also observed
that the precipitin reaction was not entirely specific, that
a precipitate would form with related organisms. Since the
amount of precipitate differed with the relationship of the
organisms tested, he concluded that the technique could be
applied to biological problans. The precipitation technique
was used in all of these early studies.

The method of layering antigen over the antisera was
used by Ascoli in 1902. This new development, the ring test,
allowed a much finer determination of the end point of reac-
tion to be made.

Nuttall, in a series of studies with a large number of
different sera and antigens, suggested that the precipitin
technique would be useful in determining animal relation-
ships. His book (1904) resulted frOm the observation of
sixteen thousand tests and was the most important work of
its kind in early serology. By 1906 several workers had in-
troduced methods of measuring the amounts of precipitate
formed in the antigen-antibody complex.

The observation by Evans (1922) and others that differ-
ing results were obtained with solutions at different pH led

to the use of physiological saline, 0.85 percent, buffered
at approximately 7.0-7.4. This is still an important factor
in performing precipitin tests.

Boyden (1926) began studies on phylogenetic application
in this country. He emphasized the use of the Micro-
Kjeldahl test for nitrogen content in standardizing the pro-
tein content of antigens. He was the first worker to dis-
cuss percent of error of the method and introduced the prin-
ciple of reciprocal relationships. This consisted of test-
ing all antisera used against both homologous and heterolo-
gous antigens.

The varied results obtained in different laboratories
led to the works of Wolfe (1935, 1936, 1939a, b) and Wolfe
and Baier (1938) in the attempt to standardize the various
methods of injecting animals, performing the tests and in-
terpreting the results. Boyden and Noble (1933) and Boyden
(1934) introduced three dimensional figures to illustrate
relationships, observing that both Darwin and Haeckel had
stated that family trees needed more than two dimensions to
show relationships properly. The measuranent of the amount
of precipitate by a photoelectric beam was introduced by
Libby (1938) with the development of the photronrefleotometer.
This instrument has come to be used almost exclusively by
Boyden and his students.

The next significant development was the work of Oudin
(1948) who developed a method of performing the ring test

9
with an agar antiserum mixture overlaid with the antigen so-
lution. This permitted the actual number of reacting anti-
gens to be established. Elek (1948, 1949) and Ouchterlony
(1949) devised methods of performing the test in an agar
base in a Petri dish. The antigen and antiserum were al-
lowed to diffuse through this agar base towards each other.
This method gave a visible and lasting demonstration of the
number and kinds of reacting systems, showing both the anti-
gens which were common to the species tested and also those

which were peculiar to each one individually.

Applications of Precipitin Testing

The first notable application of precipitin testing was
in the medico-legal differentiation of human from animal
blood. This same test is used today in modern criminological
laboratories. Later the method was used in the diagnosis of
disease. After the work of Nuttall (1904) the technique was
used in an increasing number of studies of both plant and
animal products. The work of Chester in 1937 reviews the
Plant studies and gives special treatment to the work of M62
and his coworkers. In a series of projects from 1912—1937
Mez and his associates performed an amazingly precise anal-
Ysis of plant relationships. This work culminated in the
"KBnigsberg Stammbaum":5 encompassing most of the plant king-

b

____4

5A copy of this ”Stammbaum" is reproduced in "Outlines
of Biochemistry" by R. A. Gortner, p. 542. John Wiley and
Sons, Inc. New York.

10
dom. 0n the whole, this agreed with the existing conception
of phylogeny and was received with favor by plant systema-
tists. Chester relates that their methods and techniques
were very exacting, leaving little room.for error or’misin-
terpretation. A series of exact requirements had to be met
before any test was considered significant. Although Mez
had.many detractors it is pointed out that their failure to.
duplicate results was due to neglecting to follow'an identi-
cal procedure. Accordingly, Chester states:

The main advantages of the method are its sensitiv-
ity, its specificity and its objective character. It is
limited at present [1932] by a number of inadequacies in
technique which are slowly being minimized or eliminated.
The method is significant only when skillfully employed
and when the reactions are very rigidly controlled. Give
an such prerequisites, however, and judging pragmatical—
1y the serological reactions have already served in the
solution of a number of important botanical problems and
give promise of far-reaching value as an adjunct to many
phases of botany.

No similar treatment of animals approaches Mez' "Stemm-
baum" although it is anticipated that the work of Boyden at
the Serological Museum.of Rutgers University will eventually
give some such treatment of animals.

Relationship studies with animals include those of
“blfe (1939) on.mammals, birds, and reptiles. Boyden and
Nbble (1933), in a paper on relative placement of amphibians
(Fig. 1), suggested an answer to a problem caused by the
suppression of adult characters in certain genera. Wilhelmi
(1940, 1942, 1944) worked with interrelationship of mollusks,

helminths, and other invertebrates and suggested an answer

SIREN . 11

12.5
AMPHIUMA 21'8
NECTURUS
42.7
41.5

 

 

CRYP'I'OBRANGHUS

Fig. 1. Diagram showing the relationships of four Amphibia
based on ring tests.

 

PIG . RSE

 

DOG

Fig. 2. Diggrams showing the interrelationships of common
mamm 5, based on the ring test.

1|)... 1 Elli}.

I) :l’lll‘l’l‘til‘lt)

‘1'. I! In) 11.1!

(II .I ‘l .i s

12
to the origin of the chordates. Other works include those
of Eisenbrandt (1936) on helminths, DeFalco (1942) on birds,
Leone (1949, 1950) , and Pryor and Leone (1952) on crusta-
ceans, and many others through the years. Erhardt (1931)
reviews the literature up to that time and includes reports
of studies in most phyla of animals.

Most of the earlier studies have been criticized by
Boyden (1942) on the basis that no attanpt was made to use
standardized methods, so that results could not be compared.
The various works of Boyden and his students (many papers),
mostly on mammals, have been an attempt to standardize all
procedures so that all results would be comparable. Boyden
was largely responsible for the development of three dimen-
sional figures to illustrate relationships (Figs. 1, 2).

Serological techniques have been applied to embryologi-
cal problems by Perlman (1953) in a study of sea urchin anti-
gens during developmental stages. Telfer and Williams (1953)
and Telfer (1954) have studied the blood antigens of the
Cecropia silkworm during its metamorphosis.

Applications of serology to the study of genetics using
Drosophila app. have been made by Cumley (1939, 1940) and
Fox (1949a, 1949b).

Applications of precipitin testing_to mtomolOQ. In
his review-of serological studies up to 1929, Erh‘ardt (1931)
listed fifteen species of insects which have been used in

some way in precipitin testing. Among these were numerous

13
studies on the differentiation of true honey from "Kunst-
honig." Many studies on veterinary problems of myiasis were
listed. The only relationship studies mentioned are those
of Aoki (1915) and Geyer (1915). The work of Aoki was re-
produced in some detail. Using larvae of seven species of
Lepidoptera he succeeded in getting satisfactory antisera
and made reciprocal cross tests of all seven species. The
work agreed in part with the conventional classification and
in part did not agree. Geyer used the lymph of Lepidopteran
pupae as antigens and concluded, "Die serologischen und sys-
tanatischen Verwandtschaftsverhgltnisse geben hier also ein-
ander parallel."

Various studies have also been made on such toxicologi-
cal and immunological topics as effect of bee serum and de-
tection of myiasis. The test has been used to detemine the
host of blood-sucking mosquitoes and flies by Bull and King
(1923), DeFalco (1952), and Eligh (1952). The use of pre-
cipitin testing as an aid in determining the value of preda-
tors has been reported by Downs and West (1953) and Hall,
Downe, West and MacLellan (1953). West (1950) refers to a
number of studies of mosquito blood meals, host preference
and malaria carriers at several laboratories. Brook and
Proske (1946) report on use of the test to detect predators
of mosquito larvae.

The first application of precipitin testing to insect
phylogeny in this country was by Brown and Heffron (1928) .

14
Using several species of butterflies and using guinea pigs
as antibody producers, they were able to detect inter-
subfamily and intergeneric differences. Because their tech-
nique included extracts of both old, dried specimens, and
fresh-killed specimens, which were heated at 58-60 degrees
Centigrade for one hour, their results are considered only
from a historical viewpoint.

Martin and Cotner (1934) made a serological study of a
number of phalaenids (Table II). The moths were captured in
light traps throughout the summer and kept in a dried condi-
tion. Only the thorax was used in the tests. Antigens were
standardized by weight and extracted with physiological sa-
line. They did not mention whether or not the mspension
was buffered. Rabbits were used as the antibody producers
and the ring test as the precipitin method. A sphingid and
nymphalid were also used in the tests. Their results agreed
very closely with the conventional classification in most
points. A. few points of difference were noted. They con-
eluded that the precipitin reaction was useful in determining
phylogenetic relationships between genera, subfamily and fam-
ily groups. On the whole, this appeared to be a very care-
fully conducted study.

In a comparison of morphological and serological rela-
tions of Drosophila spp. Cumley (1940), used a variety of se-
rological techniques, including complement fixation and vari-

ous precipitin tests. In comon with Martin and Cotner, he

l5

 

MELE II
INSECTS WHICH HAVE BEEN USED IN SEROLOGICAL STUDIES .
Source Insect '
Brown and Order Lepidoptera .
Heffron(1928 Pieridae Pieris ra ae (Linn.)
n1 us EIIodice (Latr.)
15 us an theme (Bois.)
Papilionidae Pa filo trohus Linn.
Nymphalidae s czBeIe Godart

Aoki (1915)*

Geyer (1913)*

Martin and
Cotner(1934)

Order Lepidoptera

Order Lepidoptera

Order Lepidoptera
Noctuidae

Sphingidae
Nymphalidae

Bomb mori Linn.

35me xEfia—n‘d'arinq Moor. 1 ,
erase ernzgi yamama Guer.

Cali La laponica Moor.

findroLlnus pin:L_ Linn.

L¥r_uant:§qzlL_a_ di a]; Linn.
D acr Sig an o

Deile hilia eu horbiae Linn.
flinx II strg Linn.
Snorintfius oceIlatus Linn.

PIerIs Brassicae Linn.

 

 

 

 

 

 

 

 

Moe albi ennis Grt.
Euxoa uafiidentata G. and R.
a $ris’EIc'fiIa Morr.
e a venera'fiiIis-arida (2:11.
a a ducens WIE.
Feltia vancouverensis Grt.
rotis sIIon Hoff.
ro s 0%50 onia Morr.
as o is av ae Grt.
or za rotis auxiliaris Grt.
Heliothfisiofisoleta Figfi.
Scoto ram na trifol‘ ‘Rott.
mpgfis Ioda S r.
Sidemia‘ devastator Brace
Protagro't'is nIverenosa Grt.
Hygcaena rufostri' ate Pack.
ra

 

Au 5 p a rass cae ‘iley
nu grag a calii'orni'ca' Speyer

finer nt us ceris K '

Vanessa car 11 L nn

 

 

[{l I} . If" I
It I [II-II.

16

TABLE II ( continued)

 

Source Insect

V

Cumley (1940) Order Diptera

Drosophilidae Droso hilia caribbea Sturt.
5530 EIlIa meIanogaster Meigm

DrosopHIIi a mull eri Sturt.
3011

1a vIrIIIs Sturt.

Leone (1947a, Order Orthoptera ‘ .
1947b) Acrididae Melanoplus femur-rubrum (DeG.~)

Melano lus differentialis Thea

R mEIea micro tera Beauv.

Paroxya a‘EIanEIca Scudd.

Le t sma mar infoolis Serv.

SE arogg E5133 §cudd.
optera Bum.

on
a xan
Tettigoniidae ifinocephalus strictus (Scudi)
@nocephalus fisci‘atus DeG..

 

 

 

 

Mantidae garatenodera sinensis' ('Sauss.)
Gryllidae Cry-nus assimilis Fahr.
Blattidae ‘geriplaneta amerioena (LinnJ

 

cthcelus surinvamensis (Linn)
atta orién is L nn.
We germanicaminn.)

Leuco haea maderae Fahr.
garcoglatta late (Brunn)

*
Reported by Erhardt (1931) .

 

17
used extracts of dried material as antigens and rabbits as
test animals. His work indicated that, in a series of com-
parisons of serological and morphological relations of four
species of Drosophila (melanogastgr, virilis, mulleri, carib-
bea), the first and'se‘cond position ~Were always apparent by '
both methods. The third and fourth places would sometimes
agree in both methods and where they did not agree there was
some doubt in the original morphological ranking. The re-
sults of serologic (testing were apparently more specific than
taxonomic ranking, indicating both species and subgroup rela-
tionship. In cases where morphologic differences are slight
or absent, the serological technique may prove of value in
determining relationships.

Leone (1947a, 1947b) made a series of studies on various
Orthoptera using as antigens, extracts of starved, freshly-
killed insects. These were ground in a meat grinder or with
mortar and pestle and extracted with buffered physiological
saline. The resulting extract was filtered through a sterile
Sietz filter and the protein content standardized according
to the nitrogen content. Rabbits were used as experimental
animals and a definite injection schedule was followed.

Some difficulty was encountered in getting satisfactory
antibody production. Tests were conducted using both the
ring test and turbidity measurement by the photronrefleo-

tomet er.

18
Although not attanpting to show definite relationships
in presenting his findings, he did establish a three dimen-
sional figure (Fig. 3) illustrating relationships among the
Blattidae, Acrididae, Gryllidae, and Mantidae. The figure
indicates the relative position of any one family in rela-
tion to any other family. He stated:

Systematic serology does not necessarily answer ques-
tions on the phylogenetic origin of organisms but rather
associates animals as they exist today into natural
classifications based on their essential biochanical
similarities. . . . Where serological evidence is in
agreement with a paleozoic taxonomic study which corre-
lates phylogenetic ori in and present day classifica-
tion, the biochemical i.e., serological) evidence be-
comes additive to the other two and increases the like-
lihood that such a study presents the true picture of
the evolutionary development of the species examined.

These studies by Leone are the only ones using insects
which fulfilled all of the requirements of present day se-
rology. Namely, the protein content of the antigens was es-
tablished by nitrogen determination, the injection schedules
were standardized, care was taken, by use of buffered saline
solutions and use of freshly killed insects, that antigen de-
naturization did not take place, and finally the insects
were starved to remove any traces of plant protein from the
digestive tract. It is unfortunate for entomology that

Leone has turned his emphasis to crustacean relationships.

19

ACRIDIDAE

 

 

GRYLLIDAE MANTIDAE

Fig. 3. Diagram representing the relationships of four
families of Orthoptera, based on turbidity test-
ing.

 

 

| F L , L'l‘ll‘lllllll Ell-II.!(I’
’ 11]) IIIIII IIII.’ illrll l til-Iii ‘.

MATERIALS AND METHODS

Preparation of Antigen Extracts

With the exception of gig mellifera, Tribolium con-
m and Sitophilus granarius; all insects used in this in-
vest‘igation were gathered in the field. The honey bees were
obtained from.the Entomology Department apiary and the
beetles were reared frmm infested grain.

Unlike many other insects the Hymenoptera are not easi-
ly reared in captivity. Because of this, a satisfactory
method of killing and preserving the specimens had to be
found. It was:necessary that this method do two things sat-
isfactorily. First, the killing agent had to kill without
denaturing the proteins and thus altering the antigenicity
of the specimen. Secondly, the method had to be used with-
out damage to the insects in any way, because of the danger
of destroying taxonomic characters. After consultation with
Dr. Mack it was decided that the use of chloroform as the
killing agent would be justified. The material was stored
and preserved in a deep freeze.

The complete procedure for handling the specimens, then,
was to capture than in the field, kill in a chloroform
bottle, transfer to carrying cans until back at the labora-
tory and then transfer to the deep freeze. Since the number

and kind of species captured on any one occasion was unpre-

21
dictable, all insects were stored until a quantity believed
sufficient for testing was gathered. They were then sorted
and identified. Species identification of all Hymenoptera
used in this stum' (Table III) was made by Dr. R. L. Fischer
of the Entomology Department.

When sufficient numbers of specimens of any one species
were available, an antigen extract was made. The procedure
used in this preparation followed a fairly uniform pattern.
Since it was believed that there might be a certain similarb
ity in eye proteins, the specimens were decapitated. In or-
der to eliminate side reactions due to digestive tract cone
tents and also due to sexual structures, the abdomen was re-
moved. This also eliminated any danger to the rabbit, due
to accmmulation of poison. Thus only the thorax was used,
and the legs and wings were removed before extraction. The
thoraces of several specimens were placed in a mortar, mixed
with a tissue-grinding medium (Alundum R R, Fisher Scientific
Campany) and ground with the pestle. ‘Varied amounts of 0.85
percent saline solution, depending on the bulk of the in-
sects, were added as an extracting:mediumi The saline was
held at pH 7.0 by means of phosphate buffer. The resulting
extraction mass was transferred to a suitable container and
placed in the refrigerator overnight. The next day the
flask was shaken to resuspend the settled.matter and again

refrigerated overnight .

TABLE III
INSECTS USED IN THIS STUDY

22

 

..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Family - . Scientifichame Code
Formicidae Formica ('2) sp. FPI
Tiphiidae M zinum quinquecintum (Fabr.) TCF

Myzi'num quinquécintum (Fabr.) TACF
Vespidae Polistes fuscatus variatus (Fabr.) PCF'
\ PoIIsfleg fuscatus variatus (Fabr.) PPCF
Sphecidae fihexm‘phex) ichneumoneus ichneumoneus SPH
on 0111mm TRY
cygtes quadrifasciata (Say) BO,
Andrenidae Andrena sp. No. l lACF
gidrena No. 2 ZACF
An rena No. 3 3ACF
An rena No. 4 4ACF
Andrena No.’ 5 SAGE
rena No. 6 6ACF
one No. '7 VACF
An rena No. 8 8ACF
Magachilidae Me achile (S a is u ate a MP?
MegacEIIe fieIome acEIIe) ganufia F. Snith MDG
Megachile ( an asarus latimanus Say MXL
Bombidae Bombus Fervidobombus) americanorum BN1
EmBus Eervi do-bombTus amerfcanorum BN2
EEmEus (Fervidobombufi amerioano’rum (Fahr) BND
Emfius Fervido‘bombus americanorum (Fabr. BM
m us FratoBSmBfis) Wth BV
m us (Fratohombua) aculatus Cresson BBI
Apidae A is mellifera L. CHFA
Is me era L. CHFA2
s m 1 era L. CHFA3
i213 me era L. CHFA4
3 me era L. .CHFAS
18 me era L. CHFAG
Apis me liars L. CHFA?
s mellifera L. CHFAB
his me era L. CHFAQ
£313 mellifera L. CHFAIO

TABLE III ( continued)

 

25

 

 

 

 

 

 

 

 

rs; —=% "‘ _ t: a
Family Scientific Name , Code
is mellifera L. (cyanide killed, frozen)“ CYFA
A 3 me 1fera L. (ether killed, frozen) - EPA
A Is meIlIfera L. (cyanide killed, dry) 1 CYDA
I Is are L. (starved killed, dry) SDA
:8 Is me era L. (ether killed, dry) EDA-
Is meIlII‘era L. (carbon tetrachloride
- ' killed, frozen) CHFA
Apis mellifera 1.. (Freeze killed, frozen) FKA
Tenebrionidae Tribolium confusum Drval ore
Curculionidae Sitophilus granarius (L) G?!

 

 

24

At the end of this forty-eight hour extraction period,
the entire mass was spun down in a clinical centrifuge at
2,000 revolutions per minute for twenty minutes. The superna-
tant fluid.was then filtered through a sterile Bflchner fil-
ter to remove any particulate matter. The resulting fil-
trate was then placed in a refrigerated centrifuge at 0°
Centigrade and spun at 15,000 revolutions per minute for
thirty minutes. This process served to remove bacteria and
other contaminants so that the supernatant fluid was now a
clear fluid. This was removed with a capillary pipette,
placed in sterile serum vials and refrigerated. Needless to
say, attempts were made at all times to observe sterile
technique. On occasion antigen samples were streaked on a
blood agar plate to check for contamination. In some cases
a final filtration through a sterile Seitz filter was:made.

After the antigens had been processed they were tested
for protein content, based on nitrogen determination, by
means of a.modified.Nessler's non-protein nitrogen tech-
nique. Readings were made with a Bauseh and Lomb'Monoehro-
Inatic Colorimeter, using a 630 millimieron filter. 'Samples
of material sent to the Agricultural Chemistry Department to
check on the method agreed. Standardization of the beetles
used in preliminary tests was based on total weight of in-

sects used.

25
Preparation of Antisera

The animals selected for antibody production were do-
mestic rabbits. These were obtained from a local commercial
breeder and housed in conventional cages. A numbered metal
tag was put in the ear of each rabbit. For the actual in-
jections several sizes of needles and syringes were used.
Impending on the route of injeetion and amount of solution,
1, 2, and 5 milliliter syringes and 23 or 26 gauge needles
were needed.

Even though serologists have been injecting animals for
over fifty years, there is still no general agreement as to
the injection route, number and time between injections,
amount of antigen to be used and time to bleed the animals.
Each laboratory has its own procedure. Boyd (1947) states
in his book that no tests involving sufficient numbers of
animals have ever been conducted to determine the best pro-
cedure. In view of this, several different schedules were
used in an attempt to get satisfactory antibody production.

When a rabbit is injected with an antigen complex, it
may produce antibodies to one, ‘two or all of the individual
antigens or it may not produce antibodies to any of then.
There scans to be no method of predicting results.

The conventional routes of injection in the rabbit are
intravenous in the marginal vein of the ear, subcutaneous

intramuscular in the leg muscles, intradermal and intra—

26

peritoneal. In this study the main route was intravenous
with a few injections subcutaneous and intramuscular.

In order for one person to inject a rabbit conveniently,
it is necessary to use some method of restraining it. Many
workers suggest that the rabbit be rolled in a towel with on-
ly the head protruding. In practice the author found that
this method.limited all movement except a slight motion of
the head which was sufficient to dislodge the needle from
the ear. To one unskilled in animal work, this can be quite
nerve-racking. After much anguish a rabbit box (Fig. 4) was
constructed and used. This confined the animal very satisfac-
torily and at the same time caused the ear veins to become
distended, thus facilitating the work. From.his experience
the author highly recommends this method.

Once the rabbit is in position, the ear is cleaned with
70 percent ethyl alcohol. A little xylene may be used to
aid in dilating the vein. In the first injection a volume
of 0.25 milliliter was used to minimize the danger of killing
the animal. Later the dosage was increased.

Injection schedules. During the preliminary studies

 

with the stored grain pests, two rabbits were injected with
each antigen, using the same schedule. All four received
0.2 milliliter of suspension, intravenously, on the first,
third, and fifth days. The dosage was increased to 0.4 mil-
liliter on the eighth, tenth, and twelfth days, and on the
fifteenth day each received 0.2 milliliter. Rabbits "A” and

 

X

 

 

 

\_

 

 

 

 

 

 

B

 

 

Fig. 4. Rabbit box.

A.

Detail of box.

 

B. Box in use.

27

“C" were trial bled by heart puncture on the seventeenth
day. At that time no antibodies were danonstrable. By the
eighteenth day rabbit "C" had died, apparently from damage
due to heart puncture. After a.week's rest the remaining
animals received.0.2 milliliter of antigen on the twenty-
third and twenty-fifth days. The dosage was increased to
0.4 milliliter on the twenty-seventh, twenty-ninth, thirty-
first, thirty-third and thirtyesixth days. Rabbit TA” was
given subcutaneous injections on the thirty-first, thirty-
third and thirty-sixth days. A final intramuscular injection
of 0.5 milliliter of antigen solution in 0.5 milliliter of '
adjuvant was given on the fortieth day. Serum from.a trial
bleeding, by heart puncture, two weeks later showed anti-
bodies. After a day had passed, the rabbits were exsanguin-
ated as completely as possible by heart puncture (see Table
IV).

The adjuvant used was that of Salk and Laurent (1952)
and consisted of a.mixture of one part Arlacel A.(Mannide
monooleate, an anulsifying agent manufactured by‘Atlas Pow-
der Company) and nine parts Bayol F Unineral oil, Standard
Oil Company, Phila., Pa.) . In use'a mixture of adjuvant and
antigen was injected intramuscularly into the hind leg of
the rabbits. This technique permitted the antigen to be ab-
sorbed slowly into the blood. The antigen therefore was

active over a relatively long time.

29

TABLE IV
INJECTION SCHEDULES

WW

 

 

 

Rabbit , .
Number 5 8 25 26
Antigen FPI BNI TCF FKA
Adjuvant used None With last With all in- With all in-
injection jections jections
Injection Intravenous Intravenous Intramuscular mtranuscular
route Intrannlswlar
l, 0.25 1, 0.25 l, 1.0 l, 1.0
3, 0.25 3, 0.25 15, 1.0 15, 1.0
5, 0.25 5, 0.25 27, trial 27, trial
7, dead 8, 0.50 bled bled
10, 0.50 35, 2.0 35, 2.0
15, 0.50 46, exsan- 46, exsan-
17, 1.00 guinated guinated
Day and 19, 1.00
amount in- 22, 1.00
jected (ml) 26, trial
bled
29, 0.5
, trial
bled
62, exsan-
guinated
Total
solution 0.75 ml. 5.75 ml. 4.0 ml. 4.0 ml.
Total

protein 8.1 mgm. 3.45mgn. 0.64mgm. 8.4mgm.

 

 

TABLE IV ( continued)

W

 

 

 

Rabbit
Number 32 41 56 72
Antigen 4ACF FPI CHFA4 CHFA
Adjuvant used With all in- With last None With last
jections injection injection
Inj ection Intramuscular Intravenous Intravenou s Intrav enou s
route mtramlsellar Intramuscular
l, 1.0 1, 0.25 l, 1.0 1, 0.25
15, 1.0 3, 0.25 3, 2.0 3, 0.25
27, trial 5, 0.25 5, 2.0 5, 0.25
bled 8, 0.50 7, 2.0 8, 0.50
35, 2.0 10, 0.50 14, trial 10, 0.50
46, exsan- 15, 0.50 bled 15,, 0.50
guinated 17, 0.50 15, exsan- 17, 1.0
Day and 19, 0.50 guinated 19, 1.0
amount in- 22, 0.50 22, 1.0
:lected (1111.) 26, trial 26, t ial
bled bled
29, 0.50 29, 0.50
60, gfigl 60, trial
6 bled
62, exsan- 62, exsan-
guinated guinated
Total .
T01escalilJltion 4.0 ml. 4.25 ml. 7.0 ml. 5.75 ml.

protein 0.64mgn. 7.65 mgm. 9.8 mgm. 5.52 mgm.

W

Rabbit

TABLE IV ( continued)

31

 

L
I‘

 

_J__

 

 

 

 

8
Number 75 78 87 8
Antigen CHFA BN1 CHFAZ CHFA2
Adjuvantused With last With last With all in- With all in-
injection injection jections jections
Injection Intravenous Intravenous mtramlsollar Intramuscular
route Intramuscular Intramuscular
1, 0.25 l, 0.25 1, 1.0 ‘l, 1.0
3, 0.25 3, 0.25 12, trial 12, trial
5, 0.25 5, 0.25 bled bled
8, 0.5 8, 0.5 20, 1.0 20, 1.0
10, 0.5 10, 0.5 31, trial 31, trial
' bled bled
Day and 15, 0.5 15, 0.5 35, exsan- 35, exsan-
17, 1.0 17, 1.0 guinated guinated
am°unt in” 19 1.0 19 1.0
Jected (“11" 22: 1.0 22’, 1.0
26, trial 26, trial
bled bled
29, 0.5 29, 0.5
60, trial 60, trial
bled bled
62, exsan- 62, exsan-
guinated guinated
Total
solution 5.75 ml. 5.75 ml. 2.0 ml. 2.0 ml.
Total
protein 5.52 mgm. 3.45 mgm. 1.8 mgm. 1.8 mgm.

 

W

TABLE IV ( continued)

32

I)

 

 

 

Rabbit Number 94 96 A
Antigen PCF cam ‘ CFB-
Adjuvant used “nth all in- NOne With last in-
jections jection
Injection route Intramuscular Intravenous Intravenous,

Slbcutaneous,

Intramuscular
1, 1.0 l, 1.0 l, 0.2
18, 1.0 3, 2.0 3, 0.2
30, trial bled 5, 2.0 5, 0.2
58, 2.0 7’ 2.0 8, 0.4
48, exsan- l4, trial bled 10, 0.4
guinated l5, exsan- 12, 0.4
guinated 15, 0.2

17, trial bled

Day and amount SE’ 8':

injected (m1.) 27: 0.4
29, 0.4
31, 0.4
33, 0.4
36, 0.4
40, 0.5
55, trial bled
57, exsan-
guinated
Total solution 4.0 ml. 7.0 ml. 4.9 ml.
1.52 mgm. 9.8 mgm. Not determined

Total protein

 

33
TABLE IV ( continued)

Wmurmw-tW...
.

 

 

 

Rabbit Number B C D
Antigen CFB GW GW
Adjuvant used With last in- None With last in-

jection jection
Injection route Intravenous, Intravenous Intravenous,

Intramuscular Intramuscular

l, 0.2 l, 0.2 l, 0.2

3, 0.2 3, 0.2 3, 0.2

5, 0.2 5, 0.2 , 0.2

8, 0.4 8, 0.4 8, 0.4

10, 0.4 10, 0.4 10, 0.4

12, 0.4 12, 0.4 12, 0.4

15, 0.2 15, 0.2 15, 0.2

23, 0.2 17, trial bled 23, 0.2
Day and amount 25, 0.2 18, dead 25, 0.2
injected (m1.) 27, 0.4 27, 0.4

29, 0.4 29, 0.4

31, 0.4 31, 0.4

33, 0.4 33, 0.4

36, 0.4 36, 0.4

40, 0.5 40, 0.5

55, trial bled 55, trial bled

57, exsan- 57, exsan-

guinated guinated

Total solution 4.9 ml. 1.8 ml. . 4.9 ml.

Total protein Not determined Not determined Not determined

——:

 

 

34

Rabbits five, eight, forty-one, seventy-two, seventy-
five, and seventy-eight were injected in a similar manner.
0n alternate days during the first week, three intravenous
injections of 0.25 milliliter of solution were given. Num-
ber five was dead on the seventh day. Dosage was increased
to 0.5 milliliter on the eighth, tenth, and fifteenth days.
With the exception of rabbit forty-one, which received 0.5
milliliter, all rabbits were injected with 1.0 milliliter of
antigen on the seventeenth, nineteenth, and twenty-second
days. Test of a trial bleeding from the ear vein on the
twenty- sixth day showed a low antibody content. An intra-
muscular injection of 0.5 milliliter of antigen in 0.5 mil-
liliter of adjuvant was given on the twenty-ninth day. A
trial bleeding on the sixtieth day revealed a satisfactory
antibody production and on the sixty-second day all rabbits
were bled out by heart puncture.

A series of intramuscular injections using the adjuvant
technique was made using rabbits twenty-five, twenty-six,
thirty-three, and ninety-four. All antigens were diluted in
half with adjuvant. One milliliter of solution was injected
in each animal on the first and fifteenth days. A test of
serum from a trial bleeding on the twenty-seventh day showed
little antibody production. An additional injection of 2.0
milliliter was given on the thirty-fifth day and the animals
were bled out on the forty-sixth day.

35

Rabbits number eighty-seven and eighty- eight were also
injected with adjuvant. They received 1.0 milliliter of
suspension in the initial injection. A trial bleeding on
the twelfth day was negative. They were again injected with
1.0 milliliter on the twentieth day and trial bled on the
thirty-first day. Serum showed the presence of antibodies
' and the rabbits were bled out on the thirty-fifth day.

A series of exclusively intravenous injections was giv-
en to rabbits fifty-six and ninety-six. On alternate days
1.0 milliliter and then three 2.0 milliliter injections of
antigen solutions were given. A test of serum on the four-
teenth day showed satisfactory antibody production and the
animals were bled out on the fifteenth day.

Antigen extracts of Apis mellifera were injected in
sevm rabbits, Bombus americanorum' in two, Formica antigen in
two, Andrena #4 in one, Polistes fuscatus in one, Myzinum
Muscintlm in one, Tribolium confusum in two, and §_i_--
tophilus granarius in two. Two rabbits. died during the in-
jections and two mere were bled out for control serum.

Bleeding techniques. In the course of injecting rab—
bits there were occasions when the withdrawal of a small
amount of blood was necessary. This was done by heart punc-
ture using a 20-milliliter syringe with an lS-gauge needle
or by puncture of the central artery of the ear using a 22-
gauge needle. An alternate method when sterile blood was

36
not essential, was the slitting of the marginal ear vein,
and allowing the blood to drip into a test tube.

When it was desired to completely exsanguinate the an-
imal, blood was obtained from the anesthetized rabbit by
heart puncture, or by cannulation of the carotid or femoral
arteries. The blood was collected in sterile 25-milliliter
test tubes and allowed to clot overnight in the icebox. The
serum was then drawn off by sterile pipette and stored in the

icebox in sterile tubes until needed.

Testing Methods

Several precipitin techniques were used in this investi-
gation with varying degrees of success. These methods in-
cluded: the ring test or interfacial technique; the turbid-
ity test or flocculation test as read by a galvanometer; the
Oudin modification of the ring test; and the Petri plate
methods of Elek and Ouchterlony.

The ring test. This method consists of the overlaying
of one solution (antiserum) with another (antigen) and watch-
ing for the formation of a cloudy ring at the interface.
Materials necessary for the test include: 0.85 percent sa-
line solution, buffered at pH 7.0; ten to fifteen small test
tubes (Kahn or Wasserman type); ten to fifteen precipitin
tubes (small bore tubes of the type used in fermentation

studies); several capillary pipettes; and a rubber nipple.

37

In practice ten or more Kahn tubes are set up in a rack
and saline solution is added to each. Two milliliters of
saline solution is put in the first tube and one milliliter
in each of the renaining tubes. A snall amount of antigen
(0.04 - 0.1 milliliter) is placed in the first tube. Then
this is mixed with the‘isaline solution by drawing the solu-
tion up and down in a one-milliliter pipette. One milli-
liter of this solution is now transferred to the next tube,
and mixed as before. One milliliter of this is transferred
to the third tube and mixed, and so on until all tubes have
been mixed. One milliliter from the last tube is discarded
(Table V) .

The series of precipitin tubes is now set up. Into the
bottom is placed, by a capillary pipette, a small amount
(0.05 - 0.1 milliliter) of a suitable antiserum. By means of
a capillary pipette using a rubber nipple and beginning with
the highest dilution, an equal amount of antigen is careful-
21y layered over the antiserum in the last tube. Antigen of
the next to highest dilution is put over the antiserum in the
next to last tube, and so on. Controls consisting of anti-
serum and saline solution, and saline solution and antigen
are also set up (Fig. 5). The tubes are now placed in an
air incubator at thirty-seven degrees Centigrade and ob-
served at twenty, forty and sixty minutes. In some cases

they may then be placed in a cold room for twenty-four hours.

38

TABLEV

METHOD OF ANTIGEN DILUTIONS

 

 

 

Tube Nmnber 1_ 2 1 3 4 5 6 7 8 9 10
Saline(mlJ 23 l. l l l l 1 .l l. l
Antigen .04 0 0 0 0 0 0 0 0 0
Ant of mixt.

transferred l—bl—~1->-l—*-l—> 1—>1—*-1 i1 41
Antigen .1 1 l 1 1 1 l 1 1 l;
dil. 50 100 200 400 800 1,600 3,200 6,400 12,800 25,600

WW

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANTIGEN N901
w H m
va— — wvvh‘w’w H
“a a ,. «.9 .. .7
Naci NORMAL
SERUM
eaJUeJls/ooeane

ANTISERUM CONTROL TUBES
Fig. 5. Results of a typical Ring test

39

A cloudy ring will appear at the interface of all tubes
which are positive, thus indicating the smallest amount of
antigen which can be detected by the particular antiserum.
This is the conventional test for measuring the potency of
antiserum. In relationship studies the antisera are over-
laid with both homologous and heterologous antigens.

Turbidity measuranents. Using this method an antigen
is directly mixed with an- antiserum, incubated for a suit-
able length of time and then shaken to resuspend any pre-
cipitate. The amount of precipitate is measured by use of a
suitable photoelectric device and recorded as amount of tur-
bidity or percent of light transmission. Materials neces-
sary for this technique are: buffered physiological saline
solution; ten or more special eolorimeter tubes; the color-
imeter (Bausch and Lomb Mono chromatic Colorimeter or other
photoelectric device); a variety of pipettes; and suitable
antigens and antisera. In practice the tubes are set up in
a rack with an added tube for control. In the first tube is
placed a specified amount of saline solution and in each of
the rest is placed two milliliters of saline solution. An
anount of antigen to bring up the volume to four milliliters
is added to the first tube. The actual amount will depend
on its protein content. The solution in the first tube is
now mixed by drawing it up and down in a two-milliliter
pipette. Two milliliters of the solution is transferred to

"" "I-'HI--Iv—--—‘—---y-.—--v¢:;-r . Wu...“

, ___,-«-.—__ .fi 5..

40

the next tube, mixed, and two milliliters transferred to the
next tube until all tubes are mixed. The final two milli-
liters is discarded flom the last tube. No antigen is put in
the control tube. The transmission of light as 100 percent
is made with the control tube as a base. Each of the other
tubes is compared with the control and any turbidity due to
the antigen is recorded. Then to all tubes (including the
control) is added 0.3 milliliters (or other mount) of anti-
serum. The tubes are now put in a hot air incubator at
thirty-seven degrees Centigrade. After a specified time
they are removed and again read with the colorimeter. The
control is again used as a transmission standard. The tur-
bidity of all tubes is now recorded. The difference between
the original and final readings is the net turbidity (Table
VI) . The net readings are then plotted on a graph or ex-
pressed as total units for the ten tubes. In relationship
studies each antiserum is reacted with each antigen.

The Oudin techniqug. This modification of the ring
test consists of the overlaying of an antiserum-agar mixture
with an antigen suspension. Each antigen in the antigen
complex will form a ring and diffuse through the agar. The
number of rings formed. depends on the number of antigens
present (Fig. 6). A mathematical formula is used to compare
rings in different tubes. Materials needed are essentially
the same as for the ring test. Snell bore glass tubes five

41

ThBLE‘VI
TYPICAL RESULTS or TURBIDITY TEST

 

Tube number 1 2 3 4 5 6 7 8 9., 10 Control

 

Homologous

antigen alone 63 76 83 82 85 89 89 87 86 89 90
Final reading 63 '74 79 '79 83 88 89 87 86 89 90
Net turbidity 0 2 4 3 2 l 0 ‘0 O 0 0
Heterologous

antigen 75 78 78 82 85 86 88 88 90 90 90
Final reading 75 78 77 so as 85 as as 90 90 90

Net turbidity O O 1 Z 2 1 O O O 0 O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

v La L J \4 v
‘ NaCl SOLUTION
Gm v v v v u
ANT ISE
AGAR MIX
TUBE
L“ L—J NJ L"
PLAIN
A AGAR
CONTROL
v H a.» v
v VA v V V
m ~31, I o. o. 'n (“m “d
five. . N.
( ~-
oh“!
L? L—v L. in} n,
B CONTROL

Fig. 6. Diagram of Oudin method. A. Before incubation
B. After incubation

43
to seven centimeters in length may be substituted for the
precipitin tubes. The capillary pipettes must be long
enough to reach the bottom of the tubes. In practice the
agar suspension is made with buffered saline solution using
from 0.3 to 1.5 percent agar (Bacto Agar, Difco). The agar
is dissolved in saline solution and sterilized in an auto-
clave. After the suspension has cooled below 45° Centi-
grade, enough antiserum to make up one-third of the total
volume is added. Then, before the mixture can cool and
harden, the precipitin tubes are half filled using the cap-
illary pipettes. After the agar has hardened, the antigen
suspension is placed over it, again using capillary pipettes.
The tubes are kept at 37° Centigrade and read at twenty-four
hour intervals. The number and kind of rings formed, and
the rate of migration down the tube is used to determine
identity of the various antigen-antibody reactions.

The Eek techniLue. This method uses agar in a Petri
dish as a medium for the formation of the antigen-antibody
complex. Antigen and antibody diffuse at right angles to
each other and form a visible white precipitate where they
meet. The number and identity of antigens is determined by
the number of lines of precipitate formed and by the join-
ing or crossing of lines from adjacent antigens. Materials
needed are: clear, unscratched Petri dishes; one percent

agar made with buffered saline solution; filter paper

44
strips; and the various antigen solutions and antisera. In
practice a layer of agar is poured in the Petri plate and
allowed to harden. A strip of filter paper is soaked in
antiserum and placed on the agar surface. Similar strips
are soaked in antigen and placed at right angles to the en-
tiserum strip (Fig. 7). A second layer of agar, cooled al-
most to the hardening point, is now poured over the strips.
The plates are kept at 37° Centigrade and observed at
twenty-four hour intervals until all activity ceases. In
relationship studies both homologous and heterologous anti-
gens are used on the strips.

The Ouchterlony techniqug. This is similar to the above
in that the reaction takes place in an agar base in a Petri
plate. The agar suspension is made as above and poured into
the plate. As originally described a trench is made in the
agar and filled with an antiserum agar mixture. The anti-
gens are then streaked over the surface at right angles to
the antiserum. A suggested modification is the making of
"wells" in the agar at the apices of a triangle and filling
these with antisera agar mixture and antigen agar mixture
(Fig. 8). The author used these and further modified the
method by using absorbent filter discs in place of wells.
These discs are soaked in the antigen or antiserum to be
tested. Still another modification was made and found very

useful. This is the use of glass rings, such as are used

45

ANTIGEN ANTIGEN !
b .

ISM—[J— ’/
ANTI

 

 

 

 

 

, /”\ b \
/\ )

 

 

 

 

 

 

B \\/

Fig. 7. The Elek method. A. Initial setup of Petri plate.
B. After incubation, antigen-complex "a" has two
antigens in common with "b".

   
      
  

ANTIGEN ANTIGEN
abf

 

 

 

 

 

 

 

 

 

 

 

.f/
!

 

a ': ..‘:'-;3'::F:W ‘:
. In“ t.‘ 0
en'-

\ f 4"... \‘w‘\ ?.
\ .f h .‘l.

\

\ (_ /

\ A /

 

Fig. 8. Ouchterlony method. .A. Initial setup. B, C, D,
possible results. B. Antigen complexes are iden-
tical. C. One antigen is present in different con-
centrations. D. Antigen-complex "a" has two anti-
gens in common with "b", and one not present in "b".

l.‘l!1.lul‘l|’||‘{xltl‘|l'|lll|i'

47
for whole mount in micro-technique, in place of the wells.
In practice the bottom of a Petri plate is covered with agar
suspension which was allowed to harden. The glass rings are
then placed in position and another lmer of agar poured
around them (Fig. 9). The antigen and antiserum to be test-
ed are then placed in the wells formed by the rings and the
plate is put in the incubator at 37° Centigrade. Plates are
observed at twenty-four hour intervals until all reaction
has ceased. This last modification is very useful and gave

much information .

 

 

 

 

 

 

 

 

 

 

 

v-.n..‘..a\\-\~ A
' ‘ ")/')//'/////////a #7 ///// , rm

1", .
.14

 

 

 

 

 

Fig. 9. Glass ring modification of Ouchterlony technique.
A. Petri plate setup. B. Section of plate to show
relation of agar and solutions.

 

 

 

  

 

 

 

 

 

 

 

Ltj
. Ir..vo...l.. Il.'. ' 0“. .QA. .3 \ -\' \ ‘ A I .‘ “o. -
B %/I///////fl/////} z/l/l/Z/ , A»

Fig. 9. Glass ring modification of Ouchterlony technique.
A. Petri plate setup. B. Section of plate to show
relation of agar and solutions.

PRESENTATION AND ANALYSIS OF DATA

Antigen Extracts

A total of forty-six insect extractions was prepared.
To facilitate handling, each was given a code number which
was used with all techniques and on all containers of ex-
tract. Because of the variance in numbers and bulk of in-
dividual species, preparation of uniform mounts of extract
was not attempted. The number of specimens used ranged from
two each in _S_p_h_e_1_c_ and W, to Apia with a total of
1,545. . When only a few specimens wereavailable only one
extract was prepared. Where more specimens were on hand, a
number of extracts were prepared as needed. Seventeen dif-
ferent groups of Ap_i_s_ were used (Table VII).

The volume of buffered saline solution added as the ex-
tracting medium ranged from three milliliters with Diem“
to thirty milliliters with an 521.2 group. The protein con-
tent of these antigen complexes ranged from 0.8 milligrams
per milliliter, in the extract of two Andrena species, up to
16.5 milligrms per milliliter in the Bombus bimaculvatus
suspension. It was interesting to note the difference in
protein extract due to bulk'of specimens. There were 15
specimens of one of the Andrena sp. mentioned above and 17
specimens of How bimaculatus. The bulk of insects was an

important factor in antigen preparation. Antigens FPI, BN1,

 

 

aTetrachloride

How ‘

SP°°m°n -‘ 0°59 Killed stored
Formica sp. FPI Chloroform Frozen
.Myz inum quinqu ec intum TCF " "
‘Myzinum " TACF " "
Polistes fuscatus PCF " "
Polistes " PPCF " "
sphex ichneumoneus SPH " "
Trypoxylon politum TRY " "
Bicyrtes quadrifasciata BO " ”
Andrena 1 1ACF " "
Andrena 2 2ACF " "
Andrena 3 3ACF " "
Andrena 4 4ACF " "
Andrena 5 SAGE " »"
Andrena 6 6AC‘FI " "
Andrena 7 7ACF ” "
Andrena 8 SAGE "' "
,Megachile pugnata MPP " "
Megachile gmula MDG " "
Megachile latimanus 1m. " "
Bombus mericanorum m1 " "
Bombus " m2 " "
Bombus " BND " "
B. mericanorum BAM " "
B. vegans ~ BV « " "
B. bimaculatus BBI " "
Apis mellifera (FIFA " "
Apia " came a u

tn 1! my!“ on n
" " CHFA4 " "
t! n CHE“ W I!
" " CHFA6 " "
" " CHFA? " "
0! " CHFAB fl 0!
" " CHFA9 " "
" " CHFAlO " "
z : 00g: Cyanide "
131'sr
" " EPA Ether Frozen
” " EDA " Dry
" 9' CAFA Carbon Frozen

l

 

 

 

 

VII
:NsluAnE
Saline Nitrogen Protein . Injected
“Egggr solution in in in
Added mgn. /ml . mgm./m1. Rabbit .
? 10 ml. 1.8 11.25 5, 41
19 10 7 0.16 1.0 25
19 5 n 0.53 3.31
15 10 u 0.38 2.37 94
9 7 n 0.65 4.06
2 5 , 0.23 1.44
2 5 n 0.16 1.0
3 ' 3 I 0 o 27 l e 69
6 3. 7 0.27 1.69
20 5 7 0.20 1.25
47 10 n 0.15 0.94
30 10 n 0.16 1.0 32
15 10 n 0.13 0.8
8 3 n 0.70 4.37
5 3 u 0.15 0.94
5 3 n 0.13 0.8
14 5 n 0.25 1.56
3 3 n 0.28 1.75
16 10 n 0.42 2.62
10 20 n 0.60 3.75 8, 78
6 10 n 0.29 1.87
5 5 n 0.64 4.0
12 10 n 1.60 10.0
25 15 n 2.34 14.62
17 10 n 2.64 16.50
85 20 n 0.96 6.0 72, 75
4% 1g " 0.90 5.62 87, 88
n 1.30 8.12
I88 38 3 1.40 8.75 56, 96
.40 8.75
100 15 " 1.90 11.87
100 20 n 1.20 7.50
100 20 v 2.40 15.0
100 20 n 1.40 8.75
100 20 2.10 13.12
125 20 " 1.80 1.25
100 30 n 0.72 4.40
125 20 s 0.32 2.20
50 20 " 0.80 5.0
120 20 n 1.40 8.75

 

TABLE VII

W

‘ How

 

' How
Specimen Code Killed Stored
Apis mellifera sDA Starved - Dry
" " FKA Frozen Frozen
Sitophilus granarius GW Crushed --
Tribolium confusum CFB n --

W

 

 

 

 

 

 

51

 

 

 

 

(continued) _
Saline Nitrogen Protein Injected
Nfiggr Solution in in in
. Added mgm./m1 . mgm.,fnl . Rabbit
100 25 ml. 0.45 2.81
'75 15 " 2.10 13.12 26
2.83 gns. - 35 --- --- A, B
3.25 gms. 25 --- --- C D

W

 

52
CHFA, Q—IFAZ, CHEN}, GW, and CFB were each used to immunize
tm rabbits. One rabbit was injected with TCF, PCF, 4ACF,
and FKA.

In many cases the volume of suspension remaining, after
the entire process of extraction and filtractidn, was not
sufficient to be used in an immunization schedule. These
snaller amounts of extract were used only as antigens. The
relative volume of suspension from any one insect extraction
was an important, and often limiting, factor in this study.

Because of the possibility of death of the rabbit or
non-production of antibodies, a relatively large volume of
extract should have been on hand. In only one case, that of
the honey bee, was this possible. In any future work it was
felt that, in order to successfully conduct a significant
study, there must be available sufficient numbers of each
species involved, so as to equal the bulk of approximately
five hundred honey bees. This bulk would insure a sufficient
amount of antigen solution, and leave a margin for laboratory
accidents, death of rabbits and non-production of anti-

bodies.

Antisera Production
Sera from nineteen rabbits were used during this inves-
tigation. With the exception of control animals, at least
two bleedings were made from each. The first blood drawn

was used to determine the presence of antibodies and whether

53
the rabbit should be exsanguinated or reinjected. In all
cases serum from the initial bleeding was designated by the
letter "A" (e. 3., serum from the first bleeding of rabbit
96 was 96A). When a large amount of blood was obtained in
the trial bleeding, there was usually sufficient serum to
run several other tests. If the rabbit was bled out after
the trial bleeding, this blood was designated as "B". If
the animal required a second series of injections the letter
"B” was used for the second trial bleeding, and "C" for the
final exsanguination. With this system it was possible that
there would be three lots of serum from any one animal
(Table VIII). Such was the case with rabbits 8, 41, '72, '75
and '78. With many of these antisera there was a change in
antibody content between the initial and final bleeding.

Since there was no guarantee that an individual rabbit
would produce antibodies, it was necessary that each antigen
be used to inject several animals. The volume of blood ob-
tained from the rabbits also pointed out the necessity of
using more than one animal for each antigen. The volumes of
blood actually obtained varied from 20 milliliters to 160
milliliters per animal. Here again there was no way of pre-
dicting the volume of blood which would be obtained. With
only one insect, Apia mellifera, was it possible to provide

sufficient serum for all tests.

54

TABLE VIII
ANTISERA PR) DUCED

 

Rabbit Number Antigen I ' Anti sera

 

5 FPI None, Animal dead
8 311 8A, B, c
25 TCF 25A, B
26 FKA 26A, B
32 4ACF 32A, B
41 FPI 41A, B, c
56 CHFA4 56A, B
60 Control 60A
65 Control 65A
'72 CHFA 72A, B, C
'75 CHFA 75A, B, C
'78 BM 78A, B, 0
8'7 CHFAZ 8'7, A, B
88 CHFAZ 88, A, B
94 PCF 94, A, B
96 CHFA4 96A, B
A CFB CFB, A, B
B CFB Animal died
0 GW GW, A, B
D G!!! GWC, D

is; allllv [alibi Iluwall I

55
Because of this unpredictability of serum volume it was
felt that from three to five rabbits should be used. with

each available antigen complex.

Results of Testing
For convenience the various methods of testing are
treated with reference to the medium in which the reaction

was conducted .

Reactions in a Liquid Medium

The ring test. In preliminary testing with the grain
beetles, only the interfacial method was used. Each of the
antigens, GW and CFB, formed a ring in all tubes when reacted
with homologous antiserum. With heterologous antisera the
reactions were relatively fine and involved no more than
three dilution tubes. Antisera against confused flour
beetle reacted with CFB in two dilution tubes. These crude
preliminary tests demonstrated the possibilities of serologi-
cal methods and pointed out several modifications which had
to be made. Serum from all later bleedings of the other rab-
bits was first tested by this method to determine the pres-
ence of antibodies. In all cases serum from at least one
bleeding was positive in every tube. While this seaned very
encouraging, it was later discovered that many of the reac-

tions were due to the presence of fatty matter in the serum.

56
After this discovery food was withheld from all animals for
forty-eight hours before bleeding.

Antiserum 96B proved fat-free and was used in a rela-
tionship trial. The animal had been injected with £1.13 ex-
tract using the intravenous route alone. Serial dilutions
(i.e., involving transfer of part of tube 1 to tube 2, then
to tube 3, etc.) were made with each of the followdng anti-
gen solutions: CHFAB, BN2, m, SAGE, TCF, FPI, and BQ.
Then part of each dilution of each antigen was placed over a
small amount of antiserum 96B in a precipitation tube. After
incubation at 37° Centigrade, in a hot air chamber, the fol-
lowing reactions were observed: 96B x CHFAB, positive to a
dilution of 1:256,000; 96B x are, positive to a dilution of
l:16,000; 96B 2: MIL, positive to a dilution of 1:8,000; 96B
x SACF, no reaction in any tube; 96B x TCF, no reaction; 963
and EPI, no reaction; and 96B x BQ, no reaction. The con-
trols were negative in all cases.(Table II). The notation
96B x BN2 means that antiserum 96B was tested with antigen
BNz.

This series of tests involved a total of 78 tubes and
gave an example of the type of results to be expected wifih
homologous and a variety of heterologous antigens. Interpre-
tation of these reactions seaned to indicate that the Bombi-
dae are closest to the Apidae, followed by the Megachilidae,

and that the other species are farther removed. These crude

TABLE II

RING TEST RESULTS WITH ANTIAPIS SERUM 96B

57

 

 

 

 

Tube:Number
.Antigan .
' 1 2 3 4 5 6 7 8 9 Control

(HFAB* § Q 4- 4- o c- o 4 o ..
m2 0- 0- 0- 4» o - - - - -
MIL .- o o- 0 - - - - - -
SACF - - - - - - - — .. ..
TCF - - - - - - - - - -
FPI - - - - - — - - - -
m - " - - III I- - In I- .-

 

 

 

= homologous reaction.

'

 

 

results agree in part with conventional rankings.

Turbidity testing. Results of the tests below were

 

treated as total turbidity units based on net turbidity as
determined by readings with the colorimeter. For conveni-
ence all homologous reactions were grouped together.

Antiapis sera tested included 26A, 56B, 720, 75B, 750,
87B, 88A, 88B, 96A, and 96B. Homologous antigen extracts
used included CHFAZ, CHFAS, CHFAS, CHFA'7, CHFAB, and CHFAlO
(Table X). Using antiserum 26A against antigen (IiFAB, a
total turbidity of five units was observed. Antiserum 56B
gave 11 units against CHFAB using two milliliters of antigen
in the initial dilution tube. When a smaller amount of anti-
gen was used the total turbidity was lowered to eight units.

Antiserum 720, when reacted with CHFAS, gave 15 units;
with CHFA5, 10 units and with CHFAB, 16 units. When the anti-
serum was diluted in half, the total with CHFAB fell to 6
units. This series of reactions showed very well the neces-
sity of using identical amounts of protein in the test. Ex-
tracts of the same species used at different protein concen-
tration gave a range of reaction from 11 to 16 units. A di-
lution of the antiserum caused a drop in reaction from 16 to
6 units.

When antiserum 75B was tested against CHFAB a low total
of two units resulted. The homologous turbidity with 750

against CHFAZ was 22 units, one of the highest reactions. At

A . III! {I} ’1‘ gall-1.. ‘I‘ [a I'll“ :ilf'llll I'll

 

TABLE I

TURBIDITY READINGS 0F ANTIAPIS SERUM
WITH HOMOLOGOUS ANTIGENS

 
  

 
 
  
 

—_—_- ____,*_~_.~ w.—- ..‘m - .0- *“.-..——'~_- -—

‘ Antiserum 26A, Antigen 'CHFAB

59

 

 

 

 

 

 

 

Net turbidity 2 2 5

Tube 1 2 3 4 5 6 - '7 ‘ 8 ' 9 10' Control
Antigen 88 88 96 99 99 98 100 94 98 95 100
Final 88 88 9 5 98 9'7 9'7 100, 9 4 98 95 100
Net Turbidity O 0 l 1 2 l 0 0 O 0 0
Antiserum 56B, Antigen CHFAB _
Tube ‘ 1 2 5 4 5 a '7 ' a '9 10 Control
Antigen '70 52 so as as 89 90 90 as a? 90
Final '70 81 7 9 80 86 88 90 9O 86 8'7 90
Net Turbidity O l 1 5 2 l O 0 0 O 0_
Antiserum 56B, Antigen CHFAB (dilute)
Tube ’ ' 1 2 ’ 3 4 5 6 '7 8 9 10' Control
Antigen 81 82 9O 95 96 96 98 99 98 9'7 100
Final 81 80 88 95 96 95 9'7 96 98 9'7 100
Net turbidity o 2 2 2 o 1 1 5 o o o
Antiserum 72C, Antigen CHFAS
Tube 1 2 3 4 5 6 7 8 9 10 Control
Antigen 84 9O 94 95 95 95 95 95 95 95 100
Final 82 88 89 94: 95 94 95 95 95 95 100
1 2 l 0 O 0 O 0

 

 

TABLE I ( continued)

60

W

‘ 3

Antiserum '7 20,' Antigen QFAS

 

 

 

 

 

 

 

Tube 1 2 5 4 5 6 7 a “a 10' Control
Antigen 62 '76 86 9O 93 9'5 96 9'7 98 98 100
Final 60 '75 85 88 92 94 95 96 98 98 100
Net turbidity 2 l l 2 l l l l 0 0 O
Antiserum 720, Antigen CHFAB (dilute) *
Tube 1 2 5 4 5 6 '7 8 9 ‘10 Control
Antigen 86 93 9'7 9'7 98 99 99 99 99 100 100
Final 86 90 94 94 9'7 9'7 98 98 9.7 100 100
Net turbidity 0 5 5 5 l 2 l l 2 O O _“
Antiserum 720 (dilute), Antigen QIFAB
Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 83 91 93 9'7 97 99 99 98 100 100 100
Final 85 90 92 95 96 98 99 98 100 100 100
Net turbidity o l l 2 1 1 o o o o o
Antiserum 75B, Antigen CHFAB
Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 88 94 96 9'7 98 9'7 98 97 100 98 100
Final 88 94 94 9'7 98 9'7 98 9'7 100 98 100
O O

Netturbidity O O 2 0 O 0 O 0 O

 

61

TABLE I ( continued)

W

Antiserum '750, Antigen CHFA2

 

Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 84 94 96 9'7 100 100 100 100 100 100 100
Final '79 93 95 93 96 93 100 100 100 100 100
Net turbidity 5 1 l 4 4 '7 O O O 0 O

 

Antiserum 87B, Antigen CHFAB

 

 

 

 

 

Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 8'7 94 97 98 99 98 100 98 100 99 100
Final 8'7 94 9'7 98 99 98 100 98 100 99 100
Net turbidity O O 0 O O 0 O O O O O
Antiserum 88A, Antigen CHFAlO
Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen '70 79 84 8'7 87 88 89 88 8'7 89 90
Final 69 '7'7 81 86 8'7 88 89 88 8'7 89 90
Net turbidity 1 2 5 l O 0 0 0 O 0 O
Antiserum 88B, Antigen CHFAB
Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 89 89 96 98 99 98 98 100 94 94 100
Final 89 89 96 9'7 98 9'7 9'7 99 94 94 100

Netturbidityo o o l l l l l o o o

TABLE I ( continued)

62

 

Antiserum 96A, Antigen CHFAB

 

 

 

 

 

 

 

Tube 1 2 3 4 5 6 7 8 9 10 Control
Antigen 74 82 84 87 88 89 89 87 88 89 90
Final 74 '79 81 84 87 87 89 87 88 89 90
Net turbidity o 5 5' 5 l 2 o o o o o
Antiserum 96B, Antigen GHFAS
Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 88 89 9O 94 98 98 98 99 99 99 100
Final 88 89 86 89 95 97 96 98 99 99 100
Net turbidity O O 4 5 3 1 2 1 O O 0
Antiserum 96B, Antigen (EFAV
Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 91 94 96 98 98 99 99 97 99 99 100
Final 90 93 94 97 98 98 99 97 99 99 100
Net turbidity l 1 2 l 0 1 0 0 O 0 0
Antiserum 96B, Antigen CHFA? (dilute)
Tlibe l 2 3 4 5 6 7 8 9 10 Control
Antigen 93 96 9‘7 96 97 99 97 99 99 99 100
Final 92 96 96 95 96 98 96 99 99 99 100
Net turbidity 1 0 l 1 1 1 l 0 0 O O

 

TABLE I ( continued)

Antiserum 96B, Antigen EFAB

 

 

 

Tube 1 2 5 4 5 6 '7 8 9 10 Control
Antigen 89 9'7 98 94 99 100 100 98 100 100 100
Final 87 95 96 91 97 99 99 9'7 100 100 100
Net turbidity 2 2 2 5 2 1 1 l 0 0 0
Antiserum 96B, Antigen CHFAB (dilute)
Tube 1 2 5 4 5 6 "7 8 9 10 Control
Antigen 74 81 84 87 88 88 86 90 88 89 90
Final 72 '79 82 84 85 86 85 89 88 89 90

Net turbidity 2 2 2 5 3 2- 1 1 0 O O

J -:__i.7___-_

 

64
the other extreme was antiserum 8'7B which failed to produce
turbidity at all when reacted with CHFAB. This was an ex-
ample of non-production of antibodies by the rabbit.

Rabbit 88, which was injected according to the same
schedule and using the same antigen a587, provided serum
with a satisfactory antibody content. Antiserum 88A reacted
with CHFAlO to give 8 units of turbidity. Antiserum 88B, ob-
tained from the same rabbit two days later showed only 5
units with CHFAlO. It appeared that in the short interval
between bleedings the antibody content had dropped. The
reason for this drop is not known. This pointed out the ne-
cessity of trying to bleed the animals at the peak of anti-
body production. In practice this exact day was difficult to
determine.

The serum from rabbit 96 proved to be the best reacting
serum produced. A large amount of blood was obtained from
the animal and the antiserum was used in a series of tests.
Antiserum 96A produced 12 turbidity units with CHFAB. Serum
96B gave a range of turbidity against different homologous
antigens. With CHFA5 the total was 16 units; with CHFA'7, 6
units; and with came it gave l4zand 15 units (Table X) .

The reaction of antiapis sera against the various solu-
tions of homologous antigen showed several phenomena of an-
tibody testing. Different animals reacted in different ways

to the same antigen. Some were good antibody producers ('72

65
and 96); some were fair producers (26, 56, 88), and 87 was
refractory. The effect of dilution of the reactants was
shown by 56B, in which a dilution of the serum caused the
turbidity to be reduced to one-half. An increase in antibody
concentration was shown by 96A.x CHFAB with 12 units, followed
by 96B x CHFAB with 14 and 15 units.

. While the homologous reactions (Table XI) were interest-
ing in that they showed the variation in response to injec-
tion of different animals, it was the heterologous reaction
which gave promise in determining taxonomic problems. These
reactions involved.most of the antisera which were prepared.
In all cases the protein content of the antigens was adjust-
ed so that identical amounts were involved in corresponding
antigen-antibody systems.

One of the earlier procedures involved reacting anti-
apis 26B against a homologous antigen CHFAS, and five heter-
ologous antigens. The heterologous antigens were BN2, 2ACF,
MIL, SPH and PCF (Table XII). A distinct turbidity was re-
corded in each case with the following results: GHFAS, 28
units; BNZ, 14 units; ZACF, 10 units; MXL, 10 units; SPH, 18
units; and POP, 12 units. The results seemed promising to a
certain extent, although they differed to some degree frem
morphological rankings. It was only after discovery'that an
essential control had not been used that the results were

seen in their true light. As they were, they served only to

demonstrate the mechanics of the test.

66

TABLE XI

SUMMATION OF HOMOLOGOUS APIS TURBIDITY TESTS

Tube Nunib er

 

Total

Systen

10

 

26A, aims
56B, cams
555, CHFAB

, 0

O
0
2

11
13

CHFA3

720, CHFA5
720, CHFAB
720, CHFAB
751-3, Cards
750, was
87B, (HFAB
sen, CEFAlO
ass, was
96A, (mus
96B, earns
96B, CHFA?
9513, CHFA?
955, earns
9513, came

720,

10

1

2

16

O

O

0

22

5

O
l
O
O

2

12

16

O

l

1
1

14

2

15

2

TABLE XII

‘I'URBIDITY READINGS OF ANTISERUM 26B
WI'JIi HE‘I‘EMLOGOUS ANTIGMS

67

 

 

 

Antigen Tube Numb er 1.
1 2 5 4 5 6 7 8 9 10 Total
earns? 91 95 95 98 98 99 99 99 1095 100 28
BN2 8 9e 97 98 97 99 99 100 100 100 100 14
2ACF 97 98 99 99 100 99 99 99 100 100 10
MIL 97 98 99 99 99 99 99 100 100 100 10
sea 95 97 97 98 98 98 99 100 100 100 18
PCF 94 97 98 99 100 100 100 100 100 100 12

 

d:
: homologous reaction.

! |\|!..| 1'! Illil

 

 

68

Antiserum 568 was used diluted (to increase the amount
of serum available) against CHFAB, BAM, 8ACF and MIL. When
the testS'were checked the only visible reaction was with the
homologous antigen. None of the others showed any turbidity.
Reports in the literature (e.g., Wblfe, 1959) indicated that
dilution may have eliminated all tendencies to react with
heterologous antigens.

The ”C" serum from rabbit '72 was used with several Apig
antigens and also with 5ACF and MDG. Strong reactions were
present against the homologous antigen but no reaction was
visible with the heterologous antigens.

A different case was that of antiapis serum 750. When
tested with its homologous antigen CHFAZ, BN2 and FPl, strong
turbidities were observed. Against CHFAZ the total was 25
units, against BN2 it was 16 units and dropped to 2 units
against EPl.(Table XIII). Relationship based on these tur~
bidity reactions place Apidae, Bombidae and Formicidae in the
order of 100 percent, 70 percent, 9 percent. These results
agree with morphological treatment of the group. The reac-
tion on the whole showed a strong turbidity ndth the homolo-
gous and closely related antigens, but reacted very little
with the:more distantly related antigen FPI.

Turbidity reactions involving antiserum 88A.and antigens
CHFAlO, TRY and PCF (Table XIII) illustrated the potential of

the method in adding to the information concerning relation-

M
Antiserum 56B, Antigens Q-IFAB, BAM, 8ACF and MJCL

TABLE XIII

69

TURBIDITY READINGS OF ANTIAPIS SERUM
AND HETEROLOGOUS ANTIGENS

 

 

m
___;*

 

 

 

 

 

 

 

 

 

 

 

 

Tube
Reading 2 '5 4 5 6 — 7 8 - 9‘ 10 control
CHFA8* 7o 82 80 83 88 89 90 90 87 87 90
Final 70 81 79 80 86 88 90 90 87 87 90
Net 0 1 l ‘_ 5 2 1 0 0 0 0 0
BAM 66 75 80 ‘88 85 90 90 90 ‘ 90 88 90
Final 66 75 80 88 85 90 90 90 90 88 90
Net 0 0 0 0 0 0 0 0 0 0 0
ancr 98 95 100 100 100 97 100 100 100 100 100
Final 98 95 100 100 100 97 100 100 100 100 100
net 0 0 o 0 0 0 0 0 o 0 o
MXL 87 94 97 98 99 99 98 96 99 100 100
Final 87 94 97 98 99 99 98 96 99 100 100
Net 0 o 0 0 o 0 0 0 ‘o 0 0
Antiserum 750, Antigens CHFA2, BN2, FPI
Tube

Reading _

1 2 5 4 5 6 7 8 9 10 Control
CHFA2* 84 94 96 97 100 100 100 100 100 100 100
Final 79 95 95 95 96 95 100 100 100 100 100
:Net 5 1 1 4 4 7 0 0 0 o 0
BN2 87 95 98 99 100 100 100 100 100 100 100
Final 85 95 95 97 98 99 100 100 100 100 100
:Net 4 2 5 2 2 l 0 0 0 0 0
.EPI 96 99 99 99 99 99 99 100 100 100 100
Final 96 98 98 99 99 99 99 100 100 100 100
:Net 0 l 1 0 0 0 0 0 0 0 0

 

70

TABLE XIII ( continued)

Antiserum 88A, Antigens CHFAlO, TRY, PCF

Tube
2 5 4 5 6 '7 8 9 10“Control

miFAlo* 70 79 84 87 87 88 89 i 89 87 89 90

Reading

Final 69 '77 81 86 86 88 89 89 8'7 89 90
Net 1 2 5 1 1 0 0 0 O 0 0
TRY 89 86 90 88 9O 9O 90 84 90 9O 90
Final 89 86 9O 86 89 89 9O 84 90 90 90
Net 0 O 0 2 1 1 O O 0 0 O
PCF 85 90 9O 86 85 89 90 8'7 89 8'7 90
Final 82 89 9O 86 85 89 90 87 89 8'7 90
Net 1 1 O 0 0 O O O O O 0
Antiserum 96A, Antigens CHFAB, TRY, PCF

Reading r Tube ~—
1 2 5 4 5 6 '7 8 9 10 Control
GIFAB’“ '74 82 84 8'7 88 89 89 8'7 88 89 90
Final '74 '79 81 84 8'7 8'7 89 8'7 88 89 90
Net 0 5 5 5 1 2 O O O O O
TRY 90 84 86 88 89 89 84 89 89 86 90
Final 89 83 85 87 88 89 84 89 89 86 90
Net 1 1 1 l 1 O 0 O 0 0 0
PCF 85 86 89 90 89 89 89 89 89 88 9O .

Final 85 86 89 89 88 89 89 89 89 88 90
Net 0 O 0 l l 0 O O O O O

 

I

TABLE XIII ( continued)

71

 

Antiserum 96B, Antigens CHFAB, BIM, 5ACF, MDG

 

 

 

 

 

 

 

 

* .-.-. homologous reaction.

Reading Tube
1 2 5 4 5 6 '7 8 9 10 Control

QIIE‘AB"I 74 81 84 87 88 88 86 9O 88 ' 89 90
Final ’72 '79 81 84 85 86 85 89 88 89 90
Net 2 2 5 5 5 2 l 1 O 0 O
BAM 65 77 78 82 88 85 89 87 85 89 90
Final 66 77 77 81 87 83 88 86 83 89 90
Net 0 O 1 l l 2 1 l 2 0 0
5ACF 9O 89 91 89 86 9O 90 89 8'7 8'7 5 90
Final 90 89 9O 88 86 9O 90 89 8'7 8'7 90
Net 0 0 1 1 0 O O 0 O O 0
MIX} 85 86 86 88 89 89 88 86 89 90 90
Final 85 86 86 86 89 89 1 88 86 89 9O 90
Net 0 O 0 2 O O 0 1 0 O O O

72
ships. Antigen TRY was made from a sphecid wasp, and PCF
was made from a vespid wasp. There has been some discussion
about the origin of the honey bees. One school holds that
the bees arose from the vespids, the other holds that they
were derived from the sphecids. Against antiapis 88A, anti-
gen CHFAIO gave a turbidity of 8 units, TRY 4 units and PCF
2 units. This gave a relationship ratio of 100 percent, 50
percent and 25 percent. On this basis the experiment appeared
to support. the proponents of sphecid ancestry.

A similar scheme resulted from the reactions involv-
ing antiapis serum 96A. Against CHFAB, TRY and PCF (Table
XIII) , the total turbidities were 12, 5 and 2. Here the re-
lationship is of the order 100 percent, 58 percent, 16 per-
cent, agreeing very well with the results above, and indicat-
ing the closer relationship of the sphecids and bees. It ap-
peared that this type of analysis would give much information
on related problems of "disputed parentage.”

A scheme illustrating relationships among some of the
bees resulted from the interaction of antiapis 96B and anti-
gen CHFAB, BAM, 3ACF and MDG (Table XIII). As expected, the
homologous reaction was strongest, with 18 units. The
strongest of heterologous reactions was with the bumble bee
antigen BAM at 8 units. This was followed by 5ACF and MDG,
each with 2 units of turbidity. The results indicated the
close relationship of the Apidae and Bombidae and the more

73
distant relation of the Megachilidae and Andrenidae. The
percent relationships were 100:45:11:11.

Tests with antibombus serum 80 against BAM, CHFAB,
and lACF gave fairly conventional results. BAM gave a tur-
bidity reading of 11 units as compared to 5 units with each
of the others (Table XIV). These reactions showed Apig and
Andrena about equally related to the bumble bees. They were
not necessarily closely related to each other, but perhaps
may be like two points at the ends of a straight line, being
equidistant from the center. Another reaction involving
antibombus serum and homologous antigen gave a total turbid-
ity of 15 units.

Serum 25B, an antitiphiid serum was reacted with sev-
eral homologous antigens giving fairly good turbidity read-
ings of 8 and 6 units. When the serum was diluted the tur-
bidity dropped to 5 units.

Antiandrena serum 52B showed an intragenus specifici-
ty. The homologous reaction with 4ACF was strongest at 8
units. With BACF, 5 units were formed and 2ACF showed 2
units. Against the related 6ACF and lACF no reaction of any
kind took place (Table XIV). These tests showed a type of
antibody production which reacted only with very closely re-
lated species, even eliminating some in the same genus. Ev-
idently the rabbit reacted only against those antigens which
were peculiar to species group 4ACF, 8ACF and 2ACF.

TABLE XIV

74

HETEROLOGOUS REACTIONS O‘IHER THAN WI'IH ANTIAPIS SERUM

 

 

 

Antiserum 80, Antigens BAM, CHFAB, lACF

 

 

 

 

Tube

368d“? 2 5 4 5 6 7 8 9 10 Control
BAM“ V 76 82 86 87 88 90 90 9o :90. 905 * 90'
Final 75 81 84 85 87 89 89 89 89 90 90
Net 1 1 2 2 1 1 1 l l 0 0
(ans 62 74 85 88 90 89 89 87 90 90 '90
Final 62 75 84 87 90 89 89 87 90 90 90
Net 0 1 1 1 0 0 0 0 o 0 o
lACF 90 86 89 84 88 90 86 ‘87 90 89 90
Final 90 85 88 84 88 89 86 87 90 89 90
Net 0 1 1 0 0 1 0 o 0 0

 

O

TABLE XIV( continued)

75

 

Antiserum 52B,

Antigens 4ncr, ancr, 21017, BACF, lACF

 

 

 

 

 

 

Reading Tube

1 2 54*4 5 6 7 8 9 10 Control
4.403" 89 9o 90 87 88 89 89 89 90 90 90
Final 89 90 89 85 86 88 88 88 90 9o 90
Net 0 0 1 2 2 1 1 l 0 o 0
BACF 90 89 85 90 88 88 87 89 87 85 90
Final 90 89 84 89 86 87 87 89 87 85 90
Net A 0 0 1 l 2 1 0 0 o 0 0
2ACF 89 9o 90 90 85 88 87 87 86 90 90
Final 89 89 89 90 85 88 87 87 86 90 90
Net 0 1 1 o 0 0 0 o 0 0 0 k
6ACF W 89 89 83 89 87 89 89 90 86 85 90
Final 89 89 85 89 87 89 89 90 86 85 90
Net 0 0 0 0 0 0 0 0 0 0 . 0
lACF 89 9o 88 90 88 89 90 87 89 89 90
Final 89 90 88 90 88 89 90 87 89 89 90
Net 0 0 0 0 0 0 0 0 0 0 0

 

'76
TABLE XIV (continued)

1
——‘-i

Antiserum 410, Antigens FPI, CHFAlO, BAM, '7ACF, MPP
Tube

 

 

 

 

 

 

 

 

 

 

Reading

2 5 4 5 6 7 8 9 10 Control
FPI" 63 "76 83 82 85 89 88 87 85 9o 90
Final 62 72 80 80 85 88 88 87 85 90 90
Net 2 4 5 2 0 l 0 0 0 0 0 A
(HFAlO 75 82 86 87 89 90 88 89 89 90 90
Final 75 81 85 86 86 88 86 88 89 90 90
Net 0 1 1 l 5 2 2 1 0 0 0
BAM 68 80 81 87 89 89 89 90 89 89 90
Final 68 80 81 87 89 89 89 90 89 89 90
Net 0 o 0 0 0 0 0 0 0 0 0
7ACF 87 90 87 89 89 89 88 89 90 90 90
Final 87 90 86 88 90 89 88 89 90 90 90
Net 0 o l l o 0 0 0 _0 0 0
MP? 82 89 86 90 86 87 88 90 90 90 90
Final 82 88 85 90 87 87 88 90 90 90 90
Net 0 1 1 0 0 o 0 0 0 0 :2:

 

* -.-. homologous reaction.

'77

The reactionswith antiformicid serum 410 were somewhat
confusing. Antigens used were FPI, CHFA'LO, BM, VACF and
MPP(Table XIV). As exPected the strongest reaction, 15
units, was With FPI, the homologous antigen. Turbidity
against @3110 was 11 units, VACF and MPP showed 2 units, '
and there was no reaction with BAN. This would.indicate
that the honey bee is closely related to the ant, with the
other bees:more distantly related.

Each series of reactions with the turbidity method gave
results which illustrated some principle of serologic test-
ing. In not one case was a heterologous reaction stronger
than the hemologous. Several tests illustrated the applica-
tion of the method to taxonomic problems. There were a few
marked lapses in technique. These were the failure to re-
peat tests using the same antigen and antisera, and the emis-
sion of reciprocal tests involving all of the heterologous
antigens and antisera. The reason for this was the lack of
sufficient volumes of reactants. This again pointed to the
need of collecting large numbers of each species and.the in-
jection of several rabbits with each antigen. However, the
reactions as they stood were believed sufficient to illus-
trate the types of results which could be expected and their

Possible application to taxonomic problens.

'78
Reactions in Agar

These techniques utilize an agar medium in which an an-
tigen-antibody complex forms a visible precipitate which is
lasting and not likely to be lost by accident. All three of
the methods are capable of furnishing the same type of in- V , 5““
formation such as the number of antigens present in an anti-
gen complex, the identity of similar antigens in different
specimens, and the number of antigens peculiar to any one
specimen.

The three methods were investigated at about the same
time and progress with each followed the sane pattern. When
consistent results were obtained with the Ouchterlony tech-
nique, trial with the others was stopped.

The Oudin technique. As discussed above, this was a
modification of the ring test in which the antiserum was
mixed with an agar solution, allowed to harden in a precip-
itin tube, and then overlaid with an antigen. Both homolo-
gous and heterologous antigens were used. The tubes were
checked every twenty-four hours until all signs of reaction
had stopped.

Tubes of an antiserum agar mixture consisting of one-
third antiserum 52B were overlaid with antigens CHFA5, 4ACF
and m2, and observed daily over a two week period. After
twenty-four hours a faint ring was noticed at the interface

of each system. By the second day there were two rings with

'79
antigen CHFAS, and still one faint ring with the others. On
the fourth day two strong rings were apparent with CHFAB, one
very sharp and the other irregular. The other two still had
the faint ring. This condition persisted for tm weeks,
after which the test was discontinued.

Nine tubes of agar and antiserum 4IC were prepared.
These were overlaid with antigens as follows: CHFAB, BND,
MIL, PCF, TCF, FPI, SPH and NaCl solution control. These
tubes were observed over a two week period and no reaction
was seen in any of the tubes.

Antigen 87B was mixed with agar and used to fill four
tubes. Three of these were overlaid with antigen QIFAS and
the other with saline solution. After twenty-four hours all
antigen tubes showed a faint ring at the interface. These
rings persisted and gradually increased in size. By the
fourth day the ring had doubled in size. On the sixth day it
appeared that these rings were beginning to separate, forming
two rings. However, the reading of the ninth day showed one
heavy ring in each tube. This remained as such until the
eighteenth day when the tubes were discarded. The control
tube remained negative throughout the test.

Antiserum 88B was mixed with agar and overlaid with
antigens (EFA5, 4ACF, and BN2. Readings after twenty-four
hours showed a faint ring at the interface of each tube. On

the second day there was a distinct ring at CHFAS, and the

80
other two were still faint. The control remained negative
for the remainder'of the test. No further reaction was seen
after two weeks and the tubes were discarded.

This same antiserumsagar cembination was also tested
with antigens CHFAB, BNZ, MIL, 4ACF, PCF, TCF, FPI, SPB and
saline solution control. In this case no rings were fonned
in the tube during the two week observation period.

ane of the above reactions were considered to be sig-
nificant. Just why the method failed may be due to several
factors. The tubes were left at room temperature (75° Fahren-
heit) and this may not have been favorable. The agar concen-
tration, 1.5 percent, may have been too high or the antigen
concentration may have been too low; No further tests were
run with this method, as one of the Petri plate techniques
was giving satisfactory results.

The Elek Petri plate method. In this technique a strip
of filter paper soaked in antiserum was laid on an agar sur-
face in a.Petri plate. Other filter strips which had been
soaked in antigen suspensions were placed at right angles to
the antiserum strip. The papers were then covered with a
second layer of agar (Fig. 7). A number of plates were pre-
pared using antisera 80, 52B, 410, 56B, 720, 78B, 780, and
88B, with a variety of antigens (Table XV) . Results of all
of these tests were very discouraging. NCne of these systems

formed a precipitate. The procedure was modified by using

81
TABLE IV
ANTISERUM-ANTIGEN SYSTEMS USED WITH ELFK MEIHOD

 

-: A r
T V—r-T—

 

 

Antiserum Antigens
‘ 80 BN2, came
86 , BN2, GBFA5
52B GHFA5, 4ACF, 8112
52B cams, 4ACF, FPI
410 FPI, m2, cam
410 FPI, PCF, FPI, TCF, FPI, 8N2
56B cams, CHFAB 1
56B CHFA'7, CHFAB
720 earns, 4ACF
720 came, 812, came, 4ACF, came, mm.
72C CHFA3, me
780 8N2, 085
780 BN2, cams, 4ACF
780 came, BNZ, 4ACF

888 8N2, cams, 4ACF

r 9
r

 

 

 

82
trenches filled with antisermmLagar and others filled with
antigen-agar. Varying concentrations of agar were tried.
Normal serum was mixed with the agar in an attempt to pro-
duce a.more natural substrate for precipitation. None of
these procedures proved satisfactory. gem

Results obtained with the other Petri plate method gave ‘
a reason for the failure of the Elek test. When the filter
paper was covered with the second layer of agar, there was no
way to add.more of the reactants to the filter paper strips.
The addition of sufficient volumes of reactants proved to be
a very necessary procedure.

The Ouchterlony technique. This method was very similar
to the Elek procedure. A.number of Petri plates were filled
with a layer of agar. After the agar hardened, absorbent
filter paper discs were placed on the surface. One disc was
placed in the center of the plate and others were placed
above, below*and on either side at about a distance of one
centimeter. An antisermm was placed on the center disc and
various antigen suspensions were placed on the peripheral
discs. Plates were prepared as follows: 26B 3: 5ACF; MIL,
GHFA5 and CHFA6; 32B 2: 2ACF, SACF; 4ACF and cans; 780 x- BND;
3N2, 4ACF and CHFAfi; and 883 x CHFAS, GHFA5,1MXL, and 4ACF
(Fig. 10 and Table XVI).

The plates were placed in an air incubator at 37° Centi-

grade and observed daily for fourteen days. Plates with

85

 

Fig. 10. Setup of Ouchterlony method using paper discs.

 

84
TABLE XVI

ANTISERUM-ANTIGE‘I SYSTEMS'USED WITH OUCHTMONY METHOD

 

 

 

Antiserum Antigens
258, 528, 568, 968 came
268 3ACF, mus, Mn, $815
528 2ACF, MP?
528 2ACF, BACF
528 2ACF, 5ACF, 4ACF, (HFA6
528 2ACF, 3ACF, MIL, cams
410 FPI, CHFAlO
56A camo, PPCF, spa
56A _ camo, BACF, 8v, MDG
56A MDG, m, MP?
568 m, SACF, 8ND, CHFA?
568 cane, cane, cane, cam
568 came, END, m, “or, rcr, TCF, FPI, SPH
568 FPI, BQ, CHFAV, POE '
720 PPCF, CHFAlO, TRY
78A 8v, BBI
78B BAM, BV, BBI
'780 4ACF, BN2, BND, (HFA6
888 MXL, came, 4ACF, CHFAS
94A PPCF, CHFAlO, BQ
968 CHFAV, cams
96B ems, came, cams, CHFA?

963 GHFAB, BACF, m, SPH

A
-

 

 

 

 

85
antisera 26B, 52B, and 780 remained negative throughout the
entire two week period. The plate containing antiserum 88B
gave the first demonstration that the test might work. Here,
after three days, one faint line of precipitate appeared be-
tween the antiserum and antigen CHFAS. The next day the line
was distinct and another line was present at CHFAS. On the
fifth day both lines were sharp and clear. No further change
was noted until the ninth day when a second line appeared at
GHFA6 (Fig. 11) . There were no other reactions at three
weeks and the test was terminated.

About this time it was thought that the reason for fail-
ure of these tests lay in the concentration of reactants.
Accordingly the technique was modified somewhat by substitut-
ing 18 x 10 millimeter micro slide rings of glass for the
filter paper discs. The method was performed as follows:
the bottom of asterile unmarred Petri dish was covered with
a layer of 1 percent agar and the agar allowed to harden;
the sterile glass rings were placed in position (Fig. 12);
another layer of agar was poured in the plate ,around the
rings; suitable antiserum and antigens were placed in the
wells formed by the rings; the plates were put in the air
incubator at 37° Centigrade and checked daily for two weeks
or until all reaction had stopped. Additional suspensions

were added as needed.

 

_

 

88B 4110.?

\

 

C D

Fig. 11. Antiserum 883 3:: MIL, CHFAG, and CHFAS. A. 8 days
incubation. B. 4 days incubation. C. 5 days incuba.
tion. D. 9 days incubation.

87

MIL

323

 

Fig. 12. Possible positions of reactants in Petri plate.

88

The first tests of this method were with antisera 56B
and 96B against their homologous antigens. The antigens
used were CHFA5, CHFA6 (2 wells) , and CHFAV. In the plate
with antiserum 56B, a faint line was seen at all antigens on
the third day (Fig. 15). By the fourth day these lines of
precipitate were distinct and had Joined, forming a square.
At this time a second faint line had appeared with all anti-
gens. On the next day these lines were more distinct. A
check on the sixth day showed a third faint line at one of
the CHFA6 wells. Twenty-four hours later the second lines
had joined so that there were now two squares of precipitate,
and the third line at CHFA6 was very distinct. The plate
was grossly contaminated on the next day and was discarded.

As far as it went, this plate showed the typical reac-
tion expected from similar antigen complexes. It appeared
that rabbit 56 had produced antibodies against at least
three single antigens in the injected antigen-complex. At
least two of these were shared by all four solutions and all
three were present in antigen CHFA6. Since these antigens
were common to all four of the reacting extracts, the lines
of precipitate formed between the antisera and antigens
joined at the point of contact, thus forming a square. This
happened with both antigens and illustrated one of the assump-

tions of the original method, i.e., that similar antigens

’- 5
t

 

‘l’
‘|’

89

 

Fig. 13 Ouchterlony method antiapis serum 56B
. x homol
antigens. A. After 1”: days. B. After 4 days. Otiggus
days. D. Seven days.

90
can be identified by the fact that their precipitates will
coalesce.

Another plate set up at the same time used antiserum
968 against antigens GHFAS, CHFA6, CHFA6 and CHFAV. One
faint line of precipitate appeared at each antigen on the
fourth day. These lines were distinct by the fifth day and
had formed a square by the next reading. Also at this time,
a second group of precipitates was visible. The next day
several of the lines crossed other lines (Fig. 14) and by
the eighth day two squares were visible, with some lines of
one crossing the other. A third set of lines was distinct
on the ninth day and a fourth set was faintly visible. The
next observation showed a third square and also that the
fourth lines were distinct. A fifth line appeared the next
day at (HFAG. No further precipitates were observed and the
fourth lines did not join. After another week of observation
the plates were discarded.

The results of this plate verified the results of plate
56B and also showed that dissimilar antigens cross each
other rather than joining. These results appear to show that
rabbit 96 had produced at least five kinds of antibodies
against the injected antigens. Four of the antigens were
found in all four extracts and antigen CHFA6 had all five of
them.

Alill . .0... put].

 

1.! I it." ‘1".-
> .
_

 

 

Fig. 14.

 

Ouchterlony technique.
GHFA6, CHFAV, CHFA6. A. 4 days. B. 8 days. 0. 10
days. D. 15 days.

 

 

 

 

 

 

Antiserum 963 x CHFA5,

92

A plate prepared with antiserum 523 against 2ACF, SAGE,
MXL and GHFAB showed some interesting reactions (Fig. 15).
No precipitate was seen until the fifth day, when one line
was visible at 2ACF. On the next day two lines had appeared
at SACF. A second was apparent at 2ACF on the eighth day.

A single line of precipitate formed at MXL on the ninth day.
This line joined the first line of 2ACF and EACF, on the ‘
tenth day. No further precipitate appeared and the plate ?
‘was discarded on the sixteenth day. Antigens 2ACF and 3ABF
were extracts of Andrena spp. and both had two antigens
which reacted with the antiandrena serum. ZMXL, an extract of
a megachilid bee, had a single antigen in common with the
Andrena spp. Apparently the honey-bee suspension did not
have a common.antigen. These same Andrena antigens had been
tested against szB'using the filter disc method and did not
form.a precipitate. The slight change in technique, allow»
ing the concentration of the solution to be maintained, was
enough to make the method a success.

Antigen 563 was tested wdth GHFA7,'MXL, 6ACF, and END,
'with interesting results (Fig. 16). No reaction appeared un-
til the eighth day when two faint lines were seen at CHFA?
and 6ACF. These lines were distinct on the ninth day. Bac-
terial growth appeared near CHFA? on the ninth day. On the
tenth day another line was visible at 6ACF. Gross contam-

ination appeared on the eleventh day and the plate was dis-

b

('3

Fig. 15.

93

CHFAB

 

Ouchterlony method, antiandrena 5213 x 2ACF, MXL,
SACF, and CHFAB. A. 5 days. B. 6 days. C. 8 days.
D. 9 days.

C

 

Fig. 16. Ouchterlony method, antiapis 56B x END, 6ACF, MXL
and CHFA'7. A. 6 days. B. 8 days. 0. 9 days. D. 10
days.

all; .I .1]

III; I! i It! Llll

95

carded. Here was a case more a heterologous antigen com-
plex formed more lines of precipitate than the homologous
antigen. This was not considered significant because of the
contamination. However this was another example of at least
three antibody systans in rabbit 56. The similarity or dif-
ferences in these two systans could not be determined, since
they were not in adjacent walls.

In a plate with antiserum 563 against FPI, BQ, CHFA? and
PCF, two lines of precipitate formed on the sixth day. Both
were at CHFAV. No further reaction was seen with any of the
antigens and the test was discontinued on the fourteenth day.

Antiserum 56B was also tested with (HFAB, BND, MIL,
4ACF, TCF, FPI, and SPH using glass cover slips over the
wells to hold down moisture loss. While this loss was min-
imized, the added hancfling favored contamination and the
plates were discarded.

Antiserum 96B was reacted with CHFAB, 5ACF, MXL and
SP3. No precipitate appeared until the eighth day when one
line formed at CHFAB. Another line formed at CHFAB on the
tenth day. No further precipitate formed and the test was
discontinued on the fourteenth day.

In a test illustrating the direct comparison of two
antigen-complexes against a single antiserum, 96B was re—
acted against GHFA'? and CHFAB. On the fourth day one faint
line was forming at each antigen. By the fifth day, these

,o ..v have... iv I I, I! !.

 

96
lines were distinct and had joined, forming a chevron-like
figure. A second line had formed with each antigen, on the
sixth day. A second "chevron" resulted on the followdng day
and also a third line had formed with each. These last
lines joined on the eighth day (Fig. 17). No further pre-
cipitate was formed and the plate was discarded on the four-
teenth day.

Results of the tests using one antigen complex and sev-
eral antisera were studied. Antigen GHFAB was tested against
antisera 25B, 32B, 56B and 96B (Fig. 18). The fourth in-
spection revealed two lines of precipitate at 96B. Twenty-
four hours later a third line had formed at 963 and a single
line at 32B and 56B. Readings on the sixth day showed three
lines at 963, three at 563 and one at 32B. A.fourth line
appeared at 96B on the next day, and one line in each systan
began to join the others. The eighth reading showed a
fourth line forming at 56B and.the junction of the combining
lines was complete. The next reading revealed a fifth line
at 563 and 96B and a single line at 25B. On the tenth day
this line had joined the others, forming a square. No other
precipitates were noted.

Near the end of the testing there remained small
amounts of various antisera and antigens. These were used hi

a series of plates as follows:

9‘7

 

Fig. 17. ' Ouchterlony method, antiapis 96B J: CHEM, CHFAB.
A. 4 days. B. 6 days. 0. 8 days. D. 10 days.

98

 

 

 

 

 

C

 

\M/

Fig. 18. Ouchterlony method, antigen CHFAB x antisera 25B,
52B, 56B and 96B. A. 4 days. B. 5 days. 0. '7 days.
D. 11 days.

99
32B 3: 2ACF, MPP
82B x 2ACF, BACF
410 x FPI, GHIO
56A.x GHIO, PPCF, SEH
56A x OHIO, 8ACF, BV, MDG
56A x MDG, m, MPP
720 x PPCF, 0810, TH!
78A x BV, BB
78B x BAM, BV, BBI
94A x PPCF, OHIO, BQ

The plates were checked daily for fourteen days and
then discarded, as no precipitate was visible in any of
them. This was anticipated with the small volumes used.
These negative results again pointed out the necessity of
keeping a sufficient volume of antisera and.antigen on hand.

The‘Petri plate method as a whole demonstrated the
types of information which could be obtained and gave proof
of great utility in relationship studies. ‘Various plates
showed that the total number of single antibodies in any
antiserum, the number of antigens held in common by differ-
ent species, and the number of antigens peculiar to a spe-
cies, could be determined.

Of the three agar techniques it was readily apparent'
that the glass ring modification of the Ouchterlony method
‘was the most practical. It was considered that any exten-
sive study would have to include this test as one of the
primary techniques.

Considering the precipitation techniques in general, it
‘was apparent that such testing could make valuable contribu-

tions to insect relations and entomological systematics. It

100
was felt that the ring test, turbidity test and Ouchterlony

test, used in conjunction with each other, could give much
information not only about the similarities of insects but
also about their essential differences. Eventually it

should be possible to prepare an antigenic formula for in-

sect species, similar to that used for bacteria.

DISCUSSION

.As stated above, the object of this investigation was
the appraisal of the methods of precipitin testing and their
application to entomological systematics, with special refer-
ence to the Hymenoptera.

Because of the difficulty encountered in raisinngymen-
optera, they had to be captured in the field. Seme:method
of killing the insects, which would not alter the body pro-
teins, had to be devised. Chloroform was decided on as the
agent of choice. The insects were stored in a freezing
chamber. When sufficient numbers of specimens had been
gathered, they were sorted and identified to species. It
was necessary to keep the specimens in the frozen state un-
til used in antigen preparation. In order that the insect
remain in a suitable condition, as little time as possible
had to be used in the identification process. The help of
Dr. R. L. Fischer, a specialist in Hymenoptera, who was fa-
miliar with most of the species used, was secured and the
insects were readily identified.

It was felt that a taxonomist would have to be consult-
ed in any future work in order to discover just what prob-
lems should be undertaken and also to facilitate the col-

lecting and identification of the species involved.

102

One of the recurring problans encountered during this
stuck; was that of insufficient volume of reactants. There
was no way of foretelling the approximate number of speci-
mens which would be necessary to conduct an entire series of
reactions. After many tests had been performed it was ap-
parent that the number of specimens of each species involved
should be such that the total bulk of. insects would equal
that of approximately five hundred honey-bees. The collab-
oration of a specialist is especially important in capturing
this number. Most insects have a seasonal abundance and also
a favored host plant. The specialist knows when the insects
are out and where to find than. When the species is present
they should be collected as soon as possible. If sufficient
numbers are not available one season, they should be stored
in a deep freeze until the proper number is secured.

mracts of each species involved should be used to in-
ject from three to five rabbits, thus insuring that a suffi-
cient supply of antiserum would be on hand and also provid-
ing against laboratory accidents or loss of animals.

In the actual extraction process, the use of buffered
saline solution is important. The insects are placed in a
mortar with a tissue-grinding medium and ground with a
pestle. The saline solution is added as needed, as an ex-
traction medium. The protein content of each extract

should be established by nitrogen determination.

103

When the rabbits are being injected some method of re-
straint is needed. mperience has shown that a rabbit box
is one of the best methods. If other persons are available
they may hold the animal. The injection schedule of choice
would seem to be that of Leone (1947) . Four doubling doses
of antigen are injected intravenously into the marginal ear
vein on alternate days. A week later a mall amount of
blood is drawn and the serum tested for antibody potency.

If the serum has sufficient titer the animal is exsanguinated.
If the potency is low, a second and third series of injec-
tions may be given. Other techniques may be used to supple-
ment this one.

Blood for trial testing is most easily obtained by
slitting the marginal ear vein and allowing the blood to
drip into a tube. If sterile blood is necessary it may be
obtained from the heart or central artery of the ear. When
blood is taken from the heart, an anesthetized rabbit is
much easier to work with than one fully conscious. Heart
puncture for exsanguination seems to be superior to cannula-
tion in ease of operation and time involved. The probabili-
ties of obtaining large amounts of blood appear to be about
equal with both methods. Food should be withheld from the
rabbit for twenty-four to forty-eight hours before bleeding
to keep down the possibilities of false reactions due to

104
fatty serum. This precaution is especially important in the
ring test.

hiring the investigation, all sera used to illustrate
precipitin technique were first used in the ring test to de-
termine potency. The type of information given by the inter-
facial technique using heterologous reactants is shown by
the interaction of antiserum 96B and antigens CHFAB, BN2,
MIL, 3ACF, TCF, FPI and BQ. The homologous reactions were
positive up to the ninth tube. The heterologous antigens
were not as strong with BN2 positive to the fifth tube, M11.
positive to the fourth tube and none of the rest positive at
all. As has always been the case, the homologous antigens
reacted in more tubes than any of the heterologous. The
bumble bees and leaf cutter bees appeared to be more closely
related to the honey-bees than were any of the other Hymenop-
tera.

It should be realized that the above reaction, and in
fact all reactions reported during this entire study are not
presented as final. Before significant results could be
presented, the tests would have to be repeated and reciprocal
testing performed.

The turbidity readings showed great promise in solving
problans of insect relationship. The tests conducted gave

some idea of the information which may be gained from this
technique. Some added data as to the origin of the bees

105
were obtained in the tests run with antiapis sera.88A.and
96B and antigen from.Apidae, Sphecidae and Vespidae. Both
88A.x GHFAlO, TRY and.PCF, and 96B x GHFAB, TRY and PCF gave
results which indicate that the Apidae are more closely re-
lated to the sphecids than to the vespids. Here again the 11
results are not presented as conclusive, but they do indi-
cate a possible use of the method. Other series of tests,
involving different species of sphecids and vespids, would
have to be conducted before definite statements concerning
these relationships could be made.

Relative placement of different families of Hymenoptera
were shown by the reaction of antiserum.750 with antigens
GHFAB, BN2 and FPI. The homologous reading was 25 units.
The reading of the closely related bumble-bee BN2 was 16
units and the distantly related ant showed 2 units.

A reaction illustrating relationships of some of the
bees was shown by antisermm 96B and antigens CHFAB, BAM,
5ACF, and MDG. The closer relationship of the honey-bees
and bumble-bees was again indicated and the relatively more
distant relationship of the Megachilidae and Andrenidae is
also illustrated.

Several specific antisera were produced. Some of these
showed a precipitate only when reacted with homologous anti-
gens. One antiserum showed an intrageneric specificity, re-

acting with some species but not math others in the same
genus.

106

The foregoing turbidity readings illustrate some of the
results which were obtained. They also indicate some of the
types of tests which could be performed in the future to ob-
tain more information about insect relationship. Such prob-
lems as placement of species in genera, and genera in fam-
ilies could be studied. Relative placement of the families
of bees and otheerymenoptera could be investigated. A.par-
ticularly ambitious project would be a serological classifi-
cation of all orders of insects.

Of the three precipitin methods using an agar base, the
Ouchterlony technique, as modified by using glass rings,
seemed to have the best applicability in insect studies.
There have been no reports of this method being used with in-
sects and indeed fairly few reports with other'organisms.
Its utility appears to be very great.

The plates using antisera 56B and 96B against homolo-
gous antigens aptly demonstrate the principles behind the
method. Application of these principles to heterologous re-
action gave some interesting results. Antiserum 32B showed
the types of reaction with closely related species and.wdth
species:more distant.

The possibility of direct comparison of two species was
shown by antiserum 968 with two homologous antigen complexes.
Three lines of precipitate were formed with each antigen and

these joined to form a "sergeant's chevron." Thus the es-

107
sential similarity of the two antigen complexes was demon-
strated.

The reaction involving one antigen and a series of an-
tisera showed how identical antigens in different species
extract can cause the formation of similar antibodies which
will form a precipitate with a homologous antigen. A.number
of lines of precipitate were formed with the homologous an-
tisera and only one with the heterologous. That these were
caused by the interaction of the same system was shown in
the joining of the lines to form a square.

Considering what has been said, it would appear that in
order to make a.significant contribution to insect relation-
ships a complete study would.involve the collaboration of a
competent serologiet and a competent systematist. It does
not appear likely, in the field of entomology'at least, that
these two specialists would be found in the same man. The
function of the systematist would be to delineate the prob-
lems to be investigated and to direct the actual capture and
identification of the specimens. The serologist would per-
form or direct the performance of the entire extraction and
testing, using the ring test, turbidity test and Ouchterlony
test as primary sources of information.

A.suggested procedure for conducting an investigation

of insect relationships would proceed as follows:

3.

4.

5.

6.

7.

8.

9.

10.

11.

108

The problem to be investigated would be determined.
Insects would be collected and killed with chloro—
form, or by freezing.

Insects collected would be field-stored in an ice
chest.
All specimens would be laboratory-stored in a deep
freeze.

A specialist in the group muld sort and identify
all specimens.

Thoraces of all specimens of one species would be
ground in a mortar using a tissue grinding medium
and employing buffered physiological saline solu-
tion as the extractant.

The extracts would be stored in the refrigerator
for forty-eight hours with occasional shaking of
the container to facilitate suspension of the
antigen.

Extracts would then be centrimged and filtered to
renove extraneous matter.

Recentrifugation at high speed and low temperature
to rmove bacteria would follow.

The sterile extract would be held in the cold until
used.

The protein content would be standardized by deter-

mining the nitrogen content and making suitable dilu-
tions.

.‘,v .5 t .3 I}.
.. . ,1 ,lr: . .I

in! ti

12.

14.

15.

16.

1'7.

18.

109
Several rabbits would be injected with each ex-
tract.
Rabbits would be trial bled and the ring test used
to estimate potency of the antiserum.
Complete exsanguination of rabbits with high titer
antiserum would follow.
The antiserum would be separated from the blood
and stored.
Relationship studies using reciprocal turbidity
test would follow.
Petri plate studies would be made to determine an-
tigen identity.
Final analysis of data to establish relationship of

insects studied could now be made.

. c5.“-

Wu‘wg‘. rump-pa.

2.

3.

4.

5.

6.

SW AND CONCLUSIONS

An appraisal of the methods of precipitin testing
as applied to entomological systematics with refer-
ence to the Hymenoptera has been made.

Insects used in the study were collected in the
field or at the Entomology Department Apiary. The
importance of collecting sufficient numbers for
complete testing is enphasized.

Antigen complexes consisting of suspended proteins
were made by extraction of the thoraces with a tis-
sue grinding medium in buffered saline solution.
Nitrogen content of each sample was determined so
that protein content of all samples could be stand-
ardized.

Rabbits were used as antibody producers. A variety
of injecting schedules was used. An intravenous in-
jection schedule was considered most efficient.
Varied methods of obtaining blood from the rabbits
were-considered. The slitting of the marginal ear
vein was useful in obtaining small amounts of
blood. Heart puncture of the anesthetized animal

was best for large volumes of blood.

Ills} lit, III..||| |, I
gilt-I'll. ll, nll'rl'll. I till, all." (lit). gall. .I1

q!‘ I'll" VIII.~

Id..r...lllull. il!’,w

'7.

8.

9.

10.

11.

12.

13.

111
It was essential that animals be starved twenty-
four to forty-eight hours before bleeding in order
to reduce fat concentration in the blood.
The application of the various testing methods was
demonstrated and their use discussed.
The ring test, turbidity test and Ouchterlony agar

method were considered best and their use recommend-

 

ed. Reciprocal testing was also recommended.
Collaboration of a serologist and systematist was
stressed.

The possibilities of precipitin testing and the
types of studies in which such testing would be val-
uable were discussed.

A procedure for a complete investigation was listed
in some detail.

It was the opinion of the author that precipitin
testing will prove to be a valuable supplement to
the conventional methods of determining insect re-

lationships.

 

 

LITERATURE CITED

Aoki, K.

1915 Baobachtungen fiber die Prfizipitationsreaktion bei
Seidenraupen. Mitteil. aus d. med. Fakulth d.
kaiserl. Univ. zu Tokyo, Vol. 2%,, p. 81. (Cited
by Erhardt, 1931.)

“301.1, Me
1902 Ueber den Mechanismus der Albuminurie durch Bier-
weise. Munchner med. loch” _4_._9_: 398-401. (Cited
by Erhardt, 1931.)

Boyd, 1!. C.
1947 Fundamentals of Immunology. Ed. 2, Interscience
Publishers, New York, 503 pp.

 

‘91:... 3 «Te—"*—

Boyden, A. A.
1926 The precipitin reaction in the study of animal rela-
tionships. Biol. Bull. _59_ (2): 73-107.

1934 Precégitins and phylogeny in animals, Am. Nat. pg:
516- 6.

1942 Bystanatic Serology: A critical appreciation.
Physiol. 2001.. 12: 109-145.

1953 Fifty years of Systematic serology. Bye. 2001. g

Boyden, A. A., and G. K. Noble
1933 The relationships of some common Anphibia as deter-
mined by serological stuchr. Amer. Mus. Novit. #600,
PP. 1-240

Bordet, J. ’
1899 Le mecanisme de 1' agglutination. Ann. de l'Inst.
Pasteur. 13: 225-250.

Brook, M. M., and H. C. Proske
1946 Precipitin test for determining natural insect
predators of immature mosquitoes. Nat'l Malarial
Soc. J. Q: 45-46.

Brom, F. M., and H. M. Heffron
1928 Serum diagnosis and Rhopalocera. J. N. Y. Ent. Soc.
§_§_: 165-168.

.. It)... 1 III. lililll‘liré !1¢\{l1lllluii .I‘. ii. .. i i , In I ‘1‘. l: I!

 

 

113

Bull, 0. G., and W. V. King
1925 The Identification of the Blood Meal of Mosquitoes
by means of the Precipitin test. Amer. Jour. Hyg.
5: 491-496.

Chester, K. S.
1957 A critique of plant serology. I, II, III, Quart.
Rev. Biol. J_:_2_: 19-46, 165-190, 294-521. ,

Cumley, R. I.
1959a Precipitin absorption with Drosophila antigens.
Amer. Nat. 25: 575-580.

1959b The relations among Drosophila species as determined
by the complanent-fixation reaction using either
insoluble fractions. J. hp. 2001. 89; 299-514.

1940 Comparison of serologic and taxonomic relationships

of Drosophila species. J. N. Y. Ent. Soc. 58: 265-
274.

Defalco , R. J.

1942 A serological study of some avian relationships.
Biol. Bull. 832: 205-218.

1952 Serological identification of mosquito blood meals.
Proc. N. J. Mosq. Exterm. Assoc. 53: 168-169.

Downe, A. E. R., and A. S. West
1955 Progress in the use of the precipitin test in ento-
mological studies. Can. Eat. 86(4): 181-184.

Eisenbrandt, L. L.
1956 Precipitin reactions of Helminth extracts. Proc.
Soc. Exp. Biol. and Med. 52: 522-525.

flex 8. DO
1948 The recognition of toxicogenic bacterial strains in
vitro. Brit. Med. Jour., March 15, 1948: 495-496.

1949 The serological analysis of mixed flocculating sys-
tems by means of diffusion gradients. Brit. J.

E11811. G. S.
1952 Factors influencing the performance of the precip-
itin test in the determination of blood meals of
insects. Canadian J. Zool. 50 (4): 215-218.

114

khardt, A.

1951 Die Verwandtschaftsbestinmlungen mittels der mmuni-
tfitsreaktionen in der Zoologie and ihr Wart fur phy-
logeneti sche Untersuchungen. Ergebnisse und Fort-
schritte der Zoologie. 1: 279-577.

Hana, A. G.
1922 A buffered physiological salt solution. Jour. Inf.
Diseases. 50: 95-98.

Fox, Allen S.
1949a Imunogenetic studies of Drosophila Meganogaster.
I. The development of techniques and the detectiOn
of strain differences. Jour. mnun. _§_2_ (1): 15-27.

1949b Imunogenetic studies of Drosophila Melanogaster.
II. Interaction between the R B and V loci in pro-
duction of antigens. Genetics 54; 647-664.

Gortner, R. A.
1958 Outlines of Biochanistry. Ed. 2. John Wiley and
Sons, Inc., New York, 1017 pp.

Hall, R. R., A. E. R. Downe, A. 8. West, and E. R. MacLellan
1955 Evaluation of insect predator-prey relationships by
precégitin test studies. Mosquito News. 1.5: (5):
199- 40

Kraus, R.

1897 Ueber spezifische Reaktionen in keimfreien Filtraten
aus Cholera, Typhus, und Pestbouillon culturen, er-
zeugt durch homologes Serum. Wien. klin. Wochenschr.
10: 756-758. (Cited by hhardt, 1951.)

Leone, Chas. A.
19478 A serological study of some Orthoptera. Ann. Eat.
Soc. finer. sq: 417-455.

1947b Bystanatic serology anong certain insect species.
Biol. Bull. 92. (1): 64-71.

1949 Comparative serology of some brachyuran Crustacea
and studies in hemocyanin correspondence. Biol.
Bull. 91(5): 275-286.

1950 Serological systematics of some paninuran and aste-
curan Crustaceae. Biol. Bull. §_8_ (2): 122-127.

115

Libby, R Lo
1958 The photronreflectometer--an instrument for the
measuranent of turbid systems. J. Immun. 14:71-75.

Martin, 8., and F. B. Cotner
1954 Serological studies of moth proteins with special
reference to their phylogenetic significance. Ann.
mt. m0. mere g1: 372-3830

Mez C. -
1926 Die Bedeutung der Sero- Diagnostik fflr die stammes—
eschichtliche Forschung. Bot. Archiv. l_§_ (1):
-23.

M82, 0., and H. Ziegenspreck
1926 Der KBnigsber er serodiagnostische Stammbaum. Bot.
”ChIVQ E (Z 3 482.4880

Nuttall, G. H. F.
1904 Blood immunity and blood relationship. Cambridge
Univ. Press, Cambridge, England, 444 pp.

Ouchterlony, Orjan
1949 An in-vitro test of the toxin-producing capacity of
Corynebacterium Diphtheriae. Lancet 256 (l): 546-

Oudin Jo
1948 L' Analyse immunochimique qualitative methods par '
. diffusion des Antigenes au sein de 1' immunsemm pre-
cipitant gelose. Ann. Inst. Pasteur f_7_§ (2):109-
129. .

Perlmann, P.
1955 Soluble antigens in sea urchin gametes and develop-

mental stages. mperimental Cell Research 5: 594-
599.

Pryor, C. R., and C. A. Leone
1952 Serological comparison of astacuran Crustacea. Biol.
Bull. 105 (5): 455-445.

Tchistovitch, T.

1899 Etudes sur 1' immunisation contre le sérum d'anguilles.
Ann. de l'Inst. Pasteur. 15: 406-425.

 

116

Salk, J. 3., and A. M. Laurent
1952 The use of adjuvants in studies on Influenza immuni-
zation. I. Measuranents in monkeys of the dimen-
sions of antigenicity of Virus-mineral oil emulsions.
J. mpt. Med. 22 (5): 429-447.

Telfer, m. H. '

1954 Immunological studies of insect metamorphosis. II.
The role of a sex-limited blood protein in egg
formation by the Cecro ia silkworm. Jour. Gen.
Physiol. 51 (4): 559-5g'8'.

Telfer, Vin. R., and Carroll M. Williams
1955 Immunological studies on insect metamorphosis. I.
Qual. and Quant. description of the blood antigens
of the Cecropia silkworm. Jour. Gen. Physiol. §_6_
(5): 589-415.

Uhlarhuth, P.
1901 Waitere Mittheilungen fiber meine Methods zum Nach-
weise von Menschenblut. Deutsche med. lochen. 21:
260-2610

West, A. 8.
1950 The precipitin test as an entomological tool. Can.
Wilhelmi, R. W.
1940 Serological reactions and species specificity of
some Helminths. Biol. Bull. 23 (1) 64-90.

1942 The application of the precipitin ”technique tothe diseases
concerning the origin of vertebrates. Biol. Bull.
82: 179-189. .

1944 Serological relationships between the Mollusca and
other invertebrates. Biol. Bull. 81 (1): 96-105.

Wolfe, H. R.
1955 The effect of injection methods on the species
specificity of serum precipitins. J. Immun. 39:
1-110

- 1956 The specificity of precipitins for serum. J. m-
mun. 51: 103.

19598 Standardization of the precipitin technique and its
application to studies of relationships in mammals,
birds and reptiles. Biol. Bull. 2.9. (1): 108-120.

 

11'7

(Wolfe, H. R., cont.)
1959b Serological relationships among Bovidae and Cer-
vidae. Zoologica. _2_4_: 509-521.

Wolfe, H. R., and J. G. Baier, Jr.
1958 Comparison of quantitative precipitin techniques as
influenced by injection procedures. Physiol. 5001.
11: 65-74.

01..

will!!! .In'