THE STATES OF A GENE LOCUS IN MAIZE
Barbara McClintock

The locus of a gene whose action
is governed by a known system of
controlling elements may undergo a
change that alters the pattern of ex-
pression of the gene in mature tis-
sues. The change arises as a single
event and thus resembles a mutation.
Because the effect is on the gene-
control mechanism, the alteration is
not termed a mutation but rather a
change in “state” of the locus. The
states reflect the presence of a mech-
anism that is able to modify gene ex-
pression during development. The
nature of the states and their signifi-
cance will be discussed in this report.

The distinctiveness of the various
states of a gene locus may be illus-
trated by selected examples. For this
purpose, some of the states of a."°
and a,™? will be considered. These
symbols were assigned to designate
the effects on the Ai (Anthocyanin)
locus in maize of two independent in-
ceptions of control of the action of its
gene by the Spm (Suppressor-muta-
tor) system. This action is required
for production of anthocyanin pig-
ment in both plant and kernel. The
states of a.%? differ from those of
a, in their modes of regulating gene
action. Some aspects of the difference
were mentioned in previous Year
Books, but its breadth was not em-
phasized because knowledge of some
of the states of a."? was insufficient
for characterization. Recent investi-
gations have made it possible to com-
pare a number of states of a:”° with
those of a,* previously studied. The
comparisons demonstrate the manner
in which a single control system pro-

vides diversity of regulation of gene
expression. Although distinctions
among states are made apparent in
the phenotypes of both plant and ker-
nel, only those manifested in the ker-
nel will be considered here, as they
can be readily illustrated.

The States of a.™*

The states of a,"" were derived
through modifications of the original
state that arose when the first in-
stance of inception of control of Ay
gene action by the Spm system oc-
curred in my stocks. Each state was
recognized initially by the appearance
of an altered type or distribution of
anthocyanin pigment in the progeny
of a plant carrying the original state,
and most often in a single individual]
of the progeny. Investigation of the
cause of the changed phenotypes was
begun with such individuals and con-
tinued. with their progeny. The
studies indicated that each change
was initiated at the A, locus by the
element of the control system residing
there, and that it represented one
type of this element’s response to the
Spm element, located elsewhere in the
chromosome complement. In the
presence of a fully active Spm ele-
ment, the states of a,”* are distin-
guished from one another by differ-
ences in the time of “turning on” of
A, gene action during development of
a tissue, by the frequency of occur-
rence of such action in the cells of a
tissue, and by the pigment intensity
in those cells in which it is produced.
They are also distinguished by the
intensity of pigment produced in the
GENETICS RESEARCH UNIT

absence of an active Spm element.
One state produces no pigment
when Spm is inactive. Each of the
other states does produce pigment,
which is uniformly distributed over
the aleurone layer of the kernel. The
pigment is intense with one state but
pale to faint with the others. There is
no direct relation between the expres-
sion given in the presence of a fully
active Spm element and that given in
its absence. It should also be men-
tioned that when two states are com-
bined as alleles in a plant or kernel,
gene action is regulated independ-
ently by each, and when an active
Spm element is present, the pattern
of anthocyanin distribution and in-
tensity produced by one allele over-
laps that produced by the other. Ilus-
trations of the phenotypes of kernels
produced by the different states of
a,"™1, both in the presence and in the
absence of a fully active Spm ele-
ment, and the overlapping patterns
produced by combinations of states
as alleles, appear in the Brookhaven
Symposia in Biology, Number 18.
Analysis of the states of a,”1 was
not complicated. Simple rules could
be formulated that allowed prediction
of the phenotype each state would
produce in response to changes in ac-
tivity of the components of Spm.
These components—component-1, the
suppressor, and component-2, the
mutator—were considered in Year
Book 64. The distinctions among
states relate to the gene-associated
element of the system, which resides
at the A, locus. How this element
operates at the level of the gene and
within the nucleus to accomplish such
diverse modes of regulation of gene
action is not yet known. It is im-
portant to recognize, nevertheless,
that the gene-associated element of a
two-element control system does pro-
vide a means for directing a broad
range of types and patterns of gene
expression during development.

665
The States of a”?

Some of the states of a,"? are so
multifaceted in expression that an
understanding of them requires
forms of analysis not demanded by
any of the states of a,". The infor-
mation obtained from the analyses
has provided additional evidence of
the way in which gene action may be
regulated differentially by a single
system of controlling elements. The.
states of a,"? may be divided into
two major classes: those having an
Spm element at or close to the A,
gene locus, and those that show no
evidence of the presence of Spm at
the locus but respond to that element
in a distinctive manner when it is
located elsewhere in the chromosome
complement. The original state, from
which most of the other states were
derived, belongs in the first class. An
Spm element is present, either within
or just distal to the locus of the struc-
tural gene(s).

The original state of a,™?. The
phenotype produced by the original
state of a,”-? depends altogether on
the phase of activity of each of the
components of Spm. If component-1
is inactive, no pigment is produced
in the plant or in the aleurone layer
of the kernel. When this component is
active, the gene is activated and pig-
ment is produced. The type and dis-
tribution of the pigment depend on
the activity of component-2. If this
component is inactive, the aleurone
layer is lightly pigmented. If com-
ponent-2 acts only late in the develop-
ment of a kernel, then smail, deeply
pigmented spots appear in a lightly
pigmented background. If it is active
at all stages of development, both
Jarge and small deeply pigmented
spots appear in the lighter back-
ground. The kernels shown in Plate
1(A)illustrate responses of the initial
state of a.-? to these different phases
of activity of the components of Spm.
666

The phenotype produced by this state
does not depend solely on the action
of the Spm element that is situated
at or close to the locus. If either com-
ponent of that element is inactive, a
fully active Spm element located else-
where in the chromosome complement
will induce the same types of re-
sponse.

The responses of the original state
of a”? to phases of activity of com-
ponent-1 of Spm, the suppressor or
inhibitor component, are the reverse
of those given by most of the states
of a,”?. Responses to component-2,
the mutator component, are alike.
The term “mutator” is applied to this
component because it induces re-
sponses that modify the organization
of the locus and thus the expression
of the gene. One such modification is
responsible for the origin of the
different states that are under dis-
cussion in this report. Some states
are produced that respond to com-
ponent-1 of Spm but not to com-
ponent-2. Others show no evidence of
a response to component-1 but do
respond to component-2. Still others,
which respond to neither component,
are termed “stable states,” and each
of them gives rise to a distinctive
phenotype in both plant and kernel.

Only a few of the stable states de-
rived from the original state of a:”°
resemble that of the A: locus before
control of its gene action was taken
over by the Spm system. They pro-
duce deep anthocyanin pigmentation
in both plant and kernel, and in the
kernel the pigment is uniformly dis-
tributed over the aleurone layer. Most
of the stable states give rise to a dis-
tinctive pattern of pigment distribu-
tion and intensity in plant and kernel.
In the kernels, clusters of cells with
more intense pigment than the sur-
rounding cells are distributed over
the aleurone layer. The intensity of
the background pigmentation ranges
from faint with some states to deep

CARNEGIE INSTITUTION

with others. Three kernels with inter-
mediate but differing levels of back-
ground pigmentation are shown in
Plate 1(B), (C), and (D). Each of
the stable states is inherited in the
same manner as any stable mutant
allele of the gene. Intralocus cross-
over studies have shown that the
phenotype each produces relates to a
component residing at a particular
site within the locus.

New states, either responding or
not responding to Spm, arise only
from those states that react to com-
ponent-2 of Spm. The event respon-
sible for an altered state must occur
in a cell of the germ line and the
modified state must be included in a
gamete. A zygote produced by the
functioning of this gamete will give
rise to a plant having the new state.
The behavior of the state may then
be analyzed in this plant and its
progeny. With some states of both
a? and ai”?, responses to com-
ponent-2 of Spm occur only very late
in development of a tissue. None may
occur in the germ-line cells. Thus
these states remain unaltered through
successive plant generations even in
the presence of an Spm element with
an early-acting component-2. All
states are inherited unaltered if com-
ponent-2 is inactive or if its action is
effective only very late in develop-
ment.

The derived states of a?. The
phenotypes produced by four modi-
fied states of a,"? appear in Plate
1(E). A fully active Spm element,
placed at or close to the A; gene locus,
is present in each kernel. No visible
evidence of a response to com-
ponent-1 of Spm is shown by any of
these states. The dark background
pigmentation in kernels like the one
on the left, and the colorless back-
grounds of kernels carrying the other
three states, are the same regardless
of the phase of activity of compo-
nent-1 in the kernel. In the kernel on
GENETICS RESEARCH UNIT

the left the small deeply pigmented
spots represent one response of the
state present in the kernel to com-
ponent-2. The size of such spots is al-
ways small. Another response to com-
ponent-2 occurs early enough for the
modified locus to be included in a
number of gametes produced by a
plant carrying this state. The modi-
fications give rise to stable states
that no longer respond to compo-
nent-2. In the kernel, the aleurone
layer is uniformly pigmented, and

the intensity of color is the same as ,

the background pigmentation pro-
duced by the parent state (the dark
kernel in the photograph).

The second kernel from the left
illustrates one of the.responses of the
state present in that kernel to com-
ponent-2. The response produces pig-
mented areas, within which the pig-
ment distribution is similar to that in
the kernels shown in (B), (C), and
(D). A second type of response is
not visibly registered in the kernel
but is made apparent by the stable
derivatives this state produces. Some
of them give rise to phenotypes re-
sembling those in (B), (C), and (D).
Others, however, produce no pigment
in the aleurone layer of the kernel.

The state of a,”? present in the
second kernel from the right is dis-
tinguished by the range of phenotypic
expressions brought about by its re-
sponses to component-2. Areas in the
aleurone layer exhibit phenotypes
similar to those produced by several
other states. One such area is visible
in this kernel; it has many deeply
pigmented spots in a pale-pigmented
background. Other kernels carrying
this state may have sharply defined
areas containing a number of deeply
pigmented spots in a colorless rather
than a pale background. All the spots
may be small, or some may be large
and some small.

The kernel on the right in Plate
1(E) has only very pale spots in a

667

colorless background. Some of the
stable derivatives of this state give
rise to kernels whose aleurone layer is
uniformly pale-pigmented, but most
of them produce colorless kernels.

Five additional states, each derived
from the initial state, have been ex-
amined. None of them gives evidence
of the presence of an Spm element at
the locus of the A, gene. Each re-
sponds in a distinctive manner, how-
ever, to an active Spm _ element
located elsewhere. The phenotypes
produced by two of these states in
the presence of a fully active Spm
are shown in Plate 1(F). The two
left-hand kernels have one state.
Their deeply pigmented spots are
distinguished from those in most of
the other kernels illustrated by a near
absence of diffusion rims. Usually,
when deeply pigmented spots or areas
are formed, a product that diffuses
from the pigment-producing cells into
the surrounding cells, and often
through several rows of cells, allows
pigment to be formed there even
though the A, gene is not functioning
in these cells.

The common expression given by
the other state shown in Plate 1(F)
is seen in the second kernel from the
right. Small deeply pigmented spots
appear in a colorless background.
This state gives rise frequently to a
new state, characterized by a marked
increase in the number of pigmented
spots, as in the adjacent kernel on
the right.

The remaining three states have
been examined in greater detail than
most of the others, as they provide
information about the operation of
the Spm system that could not be
deduced readily from the behavior of
other states. Two of them, although
independently isolated, are so much
alike that they may be considered
jointly. These states (7977B and
7995) furnished the initial evidence
of the “presetting” and “erasure”
668

mechanism that was outlined and
illustrated in Year Book 63 (pp. 592-
602) and further commented on in
Year Book 64 (pp. 527-536). This
aspect of their behavior will not be
restated here. Another aspect, not
discussed earlier, will be considered
in conjunction with the behavior of
the remaining state, 8004. These
three states differ greatly from the
original state of a”? in their re-
sponses to the changes in activity of
the Spm element that can occur in
individual cells during development
of the endosperm of the kernel.

It is recognized that the compo-
nents of Spm undergo changes in
phase of activity. Control of the time
and frequency of their occurrence
resides in the Spm element itself,
each change regulating in a distinc-
tive manner the period when the next
will take place. The changes in phase
of activity of component-1, from ac-
tive to inactive and back to active, are
unambiguously registered by each of
the states of a.%'. The same unam-
biguity applies to the states of some
of the other gene loci that have come
under the control of the Spm system,
and it applies to the original state of
a;"-?. The phenotype that will appear
after each phase change can be pre-
dicted and the predictions validated
by tests that accurately determine the
phase. Illustrations are given in the
Brookhaven Symposium paper men-
tioned earlier. The states of a”?
designated 7997B, 7995, and 8004,
however, respond in a distinctly dif-
ferent manner to the alternating
phases of activity of Spm. The dif-
ference is especially well registered
when the changes in phase occur dur-
ing development of the kernel. State
8004 will be discussed first.

State 8004 of a."-?. Examination of
this state was conducted with plants
having different constitutions with
respect to the Spm element. These

CARNEGIE INSTITUTION

were: (1) no active Spm element;
(2) one Spm element with both
components fully active; (8) one Spm
element with an active component-l
and a late-acting component-2; (4)
two Spm elements, both as in (2);
(5) two Spm elements, one as in (2)
and one as in (3). In addition, some
plants had an Spm element whose
component-1 was inactive and re-
mained inactive in most plants and in
their progeny, returning to the active
phase only very rarely; the returns
could be observed in small regions
within individual kernels. Numerous
tests were conducted with plants hav-
ing these different constitutions. For
the purposes of this discussion, the
evidence obtained from only a few
kinds of test need be mentioned.
When state 8004 is propagated in
the absence of an active Spm element,
the aleurone layer of the kernel is
colorless. If, however, an active Spm
element is present, pigment is pro-
duced. Its type and distribution then
depend on the nature of the Spm and
the number of elements present. If
component-2 of the Spm element (s)
is inactive or becomes active only late
in development, the aleurone layer is
uniformly light-pigmented. If this
component is active initially, different
phenotypic expressions of the gene
appear. When two or more such ele-
ments are present, pigment in the
aleurone layer may range from nearly
colorless in some kernels to dark pale

in others, with, in most kernels, one

or several very small deeply pig-
mented spots. If, however, a kernel
starts development with a single ac-
tive Spm element that undergoes
change in phase of activity in some
cells, early in development, then pig-
ment intensities in the aleurone layer
of the mature kernel are strikingly
modified. Examples are seen in the
kernels of Plate 2(A)-(D). These
kernels have both large and small
GENETICS RESEARCH UNIT

areas outlined by rims of deep pig-
ment. The background pigmentation
in all but one kernel is faint: such
kernels were purposely selected for
clear illustration of the deeply pig-
mented borders. The kernel with a
darker background (B) is included
to suggest the range in intensity of
background pigmentation among
these kernels.

The rimmed areas consist of de-
scendants of individual cells in which
a change in phase of Spm activity has
occurred. The deep pigment outlining
an area is produced by the outermost
cells of the area. They receive a
diffusible substance from the cells
surrounding the area, in which Spm
is fully active. This substance allows
the border cells to make a pigment
that is more intense than that in the
cells either inside or outside the area.
Thus the deep pigment of the rims is
the result of a complementation reac-
tion. The rims indicate that the prod-
uct of action of the A, gene (or
genes) differs in the cells within and
without the areas. The larger
rimmed areas often contain , small
areas that also are rimmed with deep
pigment. Again, the rims are the
product of a complementation reac-
tion. Here, however, the diffusible
substance comes from cells within
the small area and enters the cells
surrounding it, where the deep pig-
ment is produced, It could be demon-
strated that the small rimmed areas
within the larger ones are composed
of descendants of cells in which Spm
has returned to an active phase. Not
all returns are made visible in this
way. Some are accompanied by a
change at the a,” locus that alters
its capacity to produce a substance
that can complement.

In some kernels the pigment with-
in a rimmed area is much more in-
tense than that outside the area. It
is not uniform in intensity, but

669

mottled, as illustrated in Plate 2(C),
(D). Some of the mottling is due to
very small areas in which a diffus-
ible complementing substance is pro-
duced. If a number of rimmed areas
are present in a kernel, the intensity
of the background pigmentation
within the areas is often the same.
In some kernels, however, it may
differ in one or more of the areas;
see Plate 2(B) and legend. It is sus-
pected that the differences among

- kernels in the intensity of pigment

within rimmed areas, exclusive of the
rims, reflect initial differences in or-
ganization of the a,”? locus in the
kernels, or differences that may arise
during kernel development.

That such differences do arise was
learned in tests conducted to deter-
mine the cause of interruptions of a
rim along a segment of an otherwise
rimmed area. Such interruptions, ex-
pressed as absence of deep pigment,
are often noted. Some of them are
not continuous but are arranged in
sequence along a continuous segment.
It was found that the interruptions
are due in some instances to loss of
ability of the adjacent outer cells to
produce a complementing diffusible
substance that will allow the border
cells of the rimmed area to form deep
pigment. The tests were made by
introducing wa™® along with state
8004 of a,”? into the primary endo-
sperm nucleus.

The gene at the Wx (Waxy) locus
functions to convert amylopectin into
amylose in the starch granules of
cells in the endosperm. Changes in
action of this gene during develop-
ment may be detected in the mature
kernel visually and also quite pre-
cisely by staining the starch with
an iodine—potassium iodide solution.
Gene action at the wx locus is
under the control of the Spm system,
and its responses to Spm result in
the production of amylose in the
670

starch granules of the cells. Because
all the cells of the endosperm below
the aleurone layer—the outermost
layer—have starch granules, each
change in action of the gene during
endosperm development is registered
in the mature endosperm. The de-
scendants of a cell in which a change
has occurred form a well-defined sec-
tor in which the altered action of the
gene is expressed in every cell. Most
of the large sectors terminate in the
aleurone layer. Small sectors, pro-
duced by changes occurring late in
kernel development, also may termi-
nate in the aleurone layer. Thus,
when state 8004 of aym? and wa™*
are both present in a kernel, the cell
lineage of the aleurone layer over-
lying such a sector, either large or
small, is also sharply defined. Parts
of some of these sectors are adjacent
to parts of a rimmed area, and some-
times these adjacent rim-area cells
do not form deep pigment. The inter-
ruption of the rim is precisely defined
by the common region of contact of
cells of the sector with those of the
rimmed area. It is evident that the
sector is formed from descendants of
a cell in which changes have occurred
coincidentally at the loci of a,™? and
wa®, The change at a,”? alters its
ability to make a diffusible substance
that can be utilized by the adjacent
cells in the rimmed area to form in-
tense pigment.

In Year Book 57 the first evidence
of phenotypic change produced by
Alternating cycles of activity of Spm
was outlined. The studies were con-
ducted with a selected state of a.”".
(A, is another locus in maize whose
gene is involved in the biosynthetic
pathway leading to anthocyanin for-
mation.) It was learned that the num-
ber and size of pigmented areas in a
kernel, produced as the consequence
of a change of Spm from an active to
an inactive phase, depend on the

CARNEGIE INSTITUTION

number of Sym elements present in
the kernel. When one is present,
many areas exhibit this change, and
many of them are large. With two
Spm elements, the areas are all small.
With three elements, no areas or only
some very small ones are formed. The
same relationship governs the produc-
tion of the rimmed areas with state
8004 of a,"-?. Few or no rimmed areas
are formed if two or more active Spm
elements are present in the kernel.
Both large and small areas are pro-
duced if only one Spm element is
present.

An additional aspect of the be-
havior of state 8004 should be men-
tioned. If a plant carrying this state
and also one active Spm element is
utilized as pollen parent in a cross to
a plant that lacks an active Spm ele-
ment, the kernels receiving the Spm
element from the pollen parent may
show large rimmed areas, and in
some kernels the pigment within the
areas is dark. Kernels that do not
receive the Spm element from the
pollen parent are colorless. Removal
of Spm from the nucleus by meiotic
segregation does not induce a setting
of the locus that allows pigment of
various intensities to be produced
subsequently among the kernels, as
happens with states 7997B and 7995.

States 7997B and 7995 of a,™*. An
earlier report (Year Book 63, pp.
592-602) showed that kernels having
one of these states and also one or
more fully active Spm elements de-
velop a number of small, deeply pig-
mented spots in a more lightly pig-
mented background. Should only one
Spm element be present and should
this element undergo change in phase
of activity during kernel develop-
ment, the response of either state to
the change results in a darkly pig-
mented area in the aleurone layer.
Many of these dark areas contain
small colorless areas. Examples are
GENETICS RESEARCH UNIT

seen in the two kernels in Plate 2(E),
each of which carries state 7995.
Kernels with a faintly pigmented
background were chosen for the illus-
tration, so that the dark areas might
be noted readily. In the parts of the
kernels where the Spm element was
active, the characteristic pattern of
small, deeply pigmented spots ap-
pears. Such spots are absent in the
darkly pigmented areas. Except for
the small colorless spots, the pigment
within the dark areas is uniform in
intensity. The edges of the areas are
not defined by deeply pigmented rims,
as are comparable areas in kernels
carrying state 8004. Rims will appear
if the two states 8004 and 7995 are
present as alleles in a kernel. An ex-
ample is shown in Plate 2(F). Here
the uniformly dark areas, produced
by state 7995, are bordered by deeply
pigmented rims, produced by state
8004. The small colorless spot within
the area on the left is also bordered
by a pigmented rim resulting from
the action of state 8004.

The pigment produced by state
7995 when a change in phase of Spm
activity occurs during kernel develop-
ment differs markedly both in inten-
sity and in pattern of distribution
from that produced by this state
when Spm is removed by a somati-
cally occurring transposition of Spm
or by means of meiotic segregation.
In a kernel that develops from the
functioning of a gamete that has lost
the Spm element by such means, the
aleurone layer has a distinctively dif-
ferent pattern of pigment distribu-
tion, as illustrated in Year Book 63
(pp. 592-602 and Plate 2).

This review of the states of a gene
locus is intended to illustrate the
extraordinary diversity in capacity of
a single system of controlling ele-
ments to regulate the action of a gene
during development. It also demon-
strates that the distribution and de-

671

gree of expression of the end product
of action of a series of genes is
mediated through control of the ac-
tion of one of these genes. Only the
fact that anthocyanin pigment—the
end product of such a series—is not
vital to the plant makes it possible
to learn about the many kinds of reg-
ulation such a system can provide. It
is well known that the various races
and strains of maize are distin-
guished from one another by a re-
markable diversity with respect to
distribution of anthocyanin to parts
of the plant and its intensity in any
one part. The patterns are so varied
that they defy a meaningful classifi-
cation. The same is true of the differ-
ent distributions of pigment produced
by the states of a,” and, especially,
the states of a,”?. Studies conducted
by maize geneticists have shown that
in some instances different alleles of
a gene locus involved in anthocyanin
production are responsible for the ap-
pearance of different patterns of pig-
ment distribution. The kind of regu-
lation exercised by some of these
alleles resembles that afforded by
some of the states described here. In
the studies of the alleles, with two
exceptions, it has not been possible
to determine the presence at the gene
locus of a control-mechanism compo-
nent that could be responsible for the
differences in action of the alleles.
That is understandable, for a means
of detecting such a component usually
is not available. It can be suspected,
however, that many of these alleles
represent different “states” of the
loci, in the sense of the term defined
in this report. Without a means of
distinguishing between a mutant of
the structural gene itself and a
mutant that is produced by a regula-
tory component at the locus, the term
“allele” must be retained even though
its significance in any one instance
will remain ambiguous.
672

Since 1962, the Brookhaven Na-
tional Laboratory has provided gar-
den space and cultivation facilities
for growing my maize plants. I
should like to express my apprecia-

CARNEGIE INSTITUTION

tion of this courtesy, and of the gen-
erous attitude and cooperative atten-
tion of those persons at Brookhaven
who are responsible for the care and
maintenance of plant materials.

BIBLIOGRAPHY

Butler, B., see Skalka, A.
Echols, H., see Skalka, A.
Goldberg, E. B., The amount of DNA be-

tween genetic markers in phage T4. Proce.
Natl. Acad. Sci. U.S., 56, 1457-14638, 1966.

Mosig, G., Distances separating genetic
markers in T4 DNA. Proc. Natl. Acad.
Sci. U.S., 56, 1177-11838, 1966.

Skalka, A., Multiple units of transcription
in phage lambda. Cold Spring Harbor
Symp. Quant. Biol., 31, 877-879, 1966.

Skalka, A., B. Butler, and H. Echols, Genetic

control of transcription during develop-
ment of phage a. Proc. Natl. Acad. Sci.
U.S., 58, 576-583, 1967.

Smith, M. G., A replicating form of \ phage
DNA, in IX International Congress for
Microbiology, Symposia, Moscow, pp. 483-
492, 1966.

Smith, M. G., Isolation of high-molecular-
weight DNA from normal and phage-
infected E. coli, in Metheds in Enzymol-
ogy, vol. 12, part A, L. Grossman and
K. Moldave, eds., Academic Press, pp.
545-550, 1967.

PERSONNEL
Year Ended June 30, 1967

Phyllis D. Bear, Carnegie Institution
Fellow

Elizabeth M. Bocskay, Chief Clerk

Jennie S. Buchanan, Curator of Drosoph-
ila Stocks

Elizabeth Burgi, Associate in Micro-
biclogy

Ruth Ehring, Carnegie Institution Fel-
low

Agnes C. Fisher, Secretary to Director ;
Editor

Alfred D. Hershey, Director

Laura J. Ingraham, Research Assistant

Barbara McClintock, Cytogeneticist

Shraga Makover, Carnegie Institution

' Fellow

Anna Marie Skalka, Carnegie Institu-
tion Fellow

Carole E. Thomason, Technical Assist-
ant

Rudolf Werner, Associate in Research

Temporary and Part-Time
John B. Earl, Technical Assistant