/. Embryol. exp. Morph. Vol. 37, pp. 173-185, 1976
Printed in Great Britain
173
Sex determination in germ line chimeras of
Drosophila melanogaster
By E. B. VAN DEUSEN 1
From the Department of Cell Biology, Biocenter,
University of Basel
SUMMARY
Of 55fliesdeveloping from blastoderms which had received male or female pole cell transplants, 15 (7 females and 8 males) were shown by progeny testing to be germ line chimeras.
Since donor and host pole cells were genetically marked with contrasting X- or Y-linked
alleles, the progeny testing scheme enabled the genotypic sex of the donor component undergoing gametogenesis to be identified as either the same as ('homosexual' chimeras) or opposite (' heterosexual' chimeras) that of the host.
All seven of the female chimeras were identified as 'homosexual' chimeras carrying only
chromosomally female XYX donor and XX host germ cells. Similarly, all eight males were
shown to be 'homosexual' chimeras with chromosomally male XYdonor and XFhost germ
cells. The chromosomal sex of the donor component undergoing gametogenesis was in every
case the same as the phenotypic sex of the host.
Since there is an equal probability of constructing either a' homosexual' or a' heterosexual'
chimera during pole cell transplantation, the ability of pole cells to differentiate functional
gametes in hosts of the opposite sex was tested 50 % of the time even if sex reversal of these
donor pole cells could not be demonstrated. Thus the absence of 'heterosexual' chimerism
strongly supports the interpretation that the phenotypic sex of a germ cell in Drosophila
is determined entirely by its own chromosome constitution, not by that of the gonadal
mesoderm.
INTRODUCTION
Analyses of phenotypic sex transformations induced by autosomal mutants of
Drosophila melanogaster such as tra, traD, ix, and dsx (cf. Lindsley & Grell,
1968) have already shown that the mechanism of sex determination in the soma is
controlled by autosomal genes as well as by the ratio of X chromosomes to
autosomes (Bridges, 1939). In addition, the fact that germ cell differentiation is
blocked at an early stage in XjX, trajtra (or XXY, trajtra)2 transformed males
(Sturtevant, 1945; Seidel, 1960,1963) has lead to the suggestion that the chromosome constitution of a germ cell cannot be at variance with its phenotypic
differentiation.
The present paper reports on the results of experiments designed to determine
1
Author's present address: Medical and Biological Research Division, Sandoz, Ltd.,
4002 Basel, Switzerland.
2
XX stands for attached- X chromosomes.
12
EMB 3 /
174
E. B. VAN DEUSEN
the extent to which the phenotypic differentiation of germ cells in sex
chimeras of D. melanogaster depends on their own chromosome constitution or
on that of the gonadal mesoderm. The most direct approach was to restrict
chromosomal sex chimerism to the germ line by transplanting appropriately
marked male or female primordial germ cells (pole cells) at random into wildtype male or female blastoderms and to ask whether functional sex reversal
occurs in donor germ cells of the 'heterosexual' chimeras, i.e. whether a genotypically male germ cell can differentiate a functional oocyte in a female host,
and whether a genotypically female germ cell can differentiate a functional
sperm in a male host.
Previous attempts by Stern & Sherwood (1962) to construct germ line mosaics
in Drosophila in which the soma would be normally XFmale and some of the
primordial germ cells XX Y were unable to detect functional XX-bearing sperm
by progeny testing. However, since they did not know whether any of the nearly
8500 males tested actually had been gonosomic mosaics, it was not possible to
conclude from their results that XXY germ cells are unable to differentiate
functional sperm. Negative results like these could be interpreted in support of
such a conclusion only if the frequency of germ line mosaicism could have been
determined.
The present study was able to detect chimerism and differentiation of fertile
gametes by donor germ cells in breeding experiments designed to distinguish
progeny derived from host and donor pole cells. This scheme also enabled the
genotypic sex of the donor component undergoing gametogenesis to be identified as either the same as ('homosexual' chimeras) or opposite ('heterosexual'
chimeras) that of the host.
In each of the 15 flies shown to be germ line chimeras on the basis of their
progeny phenotypes the gametes derived from donor pole cells had the same
genotypic sex as their host. Non-autonomous differentiation of the transplanted
pole cells into germ cells of the opposite sex was not observed, providing strong
evidence that the sexual phenotype of germ cells in D. melanogaster is
determined entirely by their chromosome constitution.
MATERIALS AND METHODS
The Drosophila blastoderm is particularly suitable for constructing sex
chimeras since at this stage germinal and somatic cell lines of the gonad are
distinctly separated from one another at the posterior pole. Pole cells can be
transplanted to the posterior pole of a genetically different host blastoderm to
produce chimeras which are mosaic for the germ line (Illmensee, 1973). When
both donor and host pole cells migrate to the interior of the embryo during
gastrulation, both genetic types should be surrounded exclusively by mesodermally-derived host tissue throughout gametogenesis. Each successfully
transplanted host embryo in these experiments originally contained donor pole
Sex determination in chimeras of Drosophila
?
V VV
175
CY
j-36a
/
f
y «' rb
\
X
y w
y >v /
iv"
rb
r36a
36a
\
X Q donor
/
f 36a
~1
f 1
A
XY
/
Bs
_ Attached
XY
XY 0IT
/
donor
-
X
Fig. 1. Source of donor pole cells.
cells of either the same or the opposite chromosomal sex and in each case the
sex chromosomes of the donor cells carried contrasting alleles in homozygous
or heterozygous form from crosses of ywf36alywf*6a females with YLXYS,
y wa rbflY, Bs malesJFig. 1 and Table 1; also cf. Lindsley & Grell, 1968, for
description of the XY compound chromosome).
Stocks and sex chromosome markers. (1) The cross shown in Fig. 1 produced
female and male pole cell donors. The donor female blastoderms were YLXYsy
y wa rbfl ywfUa. The attached-lT compound chromosome was used to
ensure sperm motility in combinations where chromosomally female XYX
pole cells in a male host might be sex reversed to male germ cells. Although the
gene ruby (rb) is also present in this chromosome, it was not used as a phenotypic marker in the following description. (2) The male donors were y w/ 3 6 a / Y,
Bs. (3) Oregon R wild-type blastoderms were used as recipients for either male
or female donor pole cells. (4) A y wf3ea testcross stock was used for progeny
testing male and female wild-type recipients of pole cell transplants.
176
E. B. VAN DEUSEN
Pole cell transplantation. Germ cell chimeras were obtained by transplanting
at random pole cells from a chromosomally male or female blastoderm donor
among those of a recipient blastoderm of the same or opposite sex. The simplified method of pole cell transplantation developed for these experiments is
accurate and easy to set up, use and maintain. It can also be used without
further modification for accurately transferring living cytoplasm between
specific regions of the pre-blastoderm embryo.
(a) Donor and recipient eggs were collected for 3 h at 22 °C, dechorionated
in 3 % hypochlorite, blotted for 30 sec on bibulous paper, then spread with a
brush in a dry Falcon plastic dish and flooded with Ringer's. The eggs remain
attached to the bottom of the dish where they can be staged quickly and selected
for pole cell transplantation.
(b) Male and female pole cell donors and recipients were selected just before
their pole cells became well separated from the blastoderm. Several minutes
later at the time of transplantation, these pole cells have grouped themselves
into a polar 'cap' (Fig. 2B) around which the perivitelline space can be seen
easily.
(c) A strip of double-stick 'Scotch' tape was pressed along the edge of a 24 x
45 mm coverslip (Fig. 2AX) facing the injection micropipette (Fig. 2 A2). Parallel
to this edge two 'tongues' of tape were cut on three sides (Fig. 2A3) and looped
back on themselves (Fig. 2B3, C3). On each loop of tape an embryo was mounted
with its long axis parallel to the surface of the coverslip so that it was suspended
(Fig. 2C 5 ) over a clear window in the tape (Fig. 2C4) through which the injection procedures could be followed in detail. One donor and one recipient blastoderm were mounted over separate windows on the same coverslip.
(d) Mounted embryos were placed in a silica gel desiccation chamber for
5-7 min immediately prior to transplantation. Proper dehydration is critical
for preventing leakage of cytoplasm during and after the injection without
impairing development.
(e) After desiccation the donor blastoderm was covered with medium viscosity
'Voltalef 10 S' fluorocarbon oil (Plastimer, Paris) and the recipient with high
viscosity 'Voltalef 20 S' oil (Fig. 2C7). The coverslip was then mounted on the
microscope stage and adjusted so that the long axes of the embryos and the
injection micropipette were parallel (Fig. 2B 5 , B6).
(/) All transfers were made with one siliconized glass micropipette (i.d. 8-10
jum) mounted on a Prior micromanipulator. The pipette was tipped either by
breaking or grinding the tip at 45°, attached to a water-filled syringe system
(Gordon Instruments, Bloomington, Indiana) and filled with 'Voltalef 10 S'
fluorocarbon oil. All transplantations were made at 500 x using Zeiss inverted
phase optics.
(g) Depending on the orientation of the embryo on its tape loop, pole cells
can be removed after penetrating either the anterior or posterior pole of the
donor. All of the pole cells were drawn slowly into the micropipette, taking care
Sex determination in chimeras o/Drosophila
Fig. 2. Scheme for pole cell transplantation.
177
178
E. B. VAN DEUSEN
to avoid contamination by blastoderm cells. After the micropipette was withdrawn from the donor a small quantity of oil was drawn into only the very tip to
protect the donor cells when the micropipette was raised from the oil droplet
and exposed to the air.
(/?) In order to inject these pole cells, the coverslip must be repositioned so
that the recipient is located directly below and parallel to the loaded micropipette. Then the micropipette was lowered into the second oil droplet and the
protective oil plug expelled immediately prior to penetrating the recipient from
the posterior pole. The best results were obtained when all of these donor pole
cells were injected directly from the posterior end into the perivitelline space
among the pole cells of the wild-type host.
(/) The coverslip was finally transferred to a moist chamber until the
recipient hatched. Before it crawled oui of its oil droplet, the larva was transferred to a vial in which yeast suspension had been spread over normal
medium.
(j) Female Oregon R recipients were testmated with y w/ 3 6 a \ Y males and
each male Oregon R recipient with two y wf36a / y wf3Ga females of the same
stock. If matings of the recipients failed to produce fertile eggs, the y w/ 3 6 a
test mates were replaced. All F 1 progeny were classified according to their sex,
body colour, eye colour and shape, and bristle morphology (Table 1).
RESULTS
Tests for sex reversal in transplanted germ cells (Table 1)
The experiments reported here follow the developmental behaviour of genotypically male and female pole cells (Fig. 1) which were transplanted among the
pole cells of Oregon R recipients of the same (' homosexual' chimeras) or the
opposite sex ('heterosexual' chimeras). Table 1, lines A to D, lists the four possible random transfers of donor pole cells to wild-type recipient blastoderms.
Fifty per cent of the successful transplantations will produce 'homosexual'
chimeras, while the remaining 50 % will be 'heterosexual' chimeras of the germ
line. Progeny testing of all these recipients can show whether the progeny are
exclusively derived from host pole cells or, alternatively, whether some of the
progeny stem from donor pole cells. Furthermore, the chromosomal constitution of the contributing donor germ cells can be derived from the phenotypes
of the donor progeny (classes 1 through 12 in Table 1).
Progeny testing of a female host. These testcrosses are shown in lines (A) and
(B) of Table 1 and will produce phenotypically wild-type progeny derived from
wild-type host pole cells. If the female recipient is chimeric, she has a 50 %
probability of having received either XYX or XY pole cells.
Line A (homosexual combination). Four classes of progeny are expected from
XYX donor pole cells if they differentiate ova in XX hosts which are testcrossed
to ywfUa I Y males: y w/ 3 6 a females (25%) and males (25%), y wafj Y
Sex determination in chimeras 0/Drosophila
179
Table 1. Progeny testing scheme to distinguish 'homosexual' chimeras from
'heterosexual' chimeras
Donor -*- Host X Testcross
v wr
6a
+
Progeny from
host pole cells
Progeny from donor pole cells
y w /36U
/•36 a
Ma
y w f
\y w . / /
j XYX
v w fM"
+
» z 36 "
v w fMa
y w f36"
y w f3ba
XX
XY
y w f36a
vw f
Bs
XY, Bar
7
XX
v w f3
v
y w f36"
.v w f36a /y w f
YY, lethal
8
y w f
iv . / •
males (25%) and ywaflywf™a
females (25%). The presence of ywafl
36a
a
y w/
females (class 1) and y w fj Y males (class 4) would provide evidence
for the capacity of XYX female pole cells to differentiate functional ova in a
female host. Although rb is present in this XY chromosome of the female donor
(Fig. 1), it was not used as a phenotypic marker and therefore does not appear in
lines A and D.
Line B {heterosexual combination). Three classes of progeny can be expected
from XY donor pole cells which are sex-reversed in XX female hosts and differentiate functional ova: y w/ 3 6 a females (33-3%) and males (33-3%) and
y w/ 3 6 a / Y, Bs Bar-eyed males (33-3 %). Since the y wf3ea females and males
(classes 5 and 6, respectively) cannot be distinguished from classes 2 and 3, only
the presence of y w/ 3 6 a / Y, Bs males (class 7) would unequivocally demonstrate that an XY pole cell can undergo sex reversal and differentiate into a
normal egg. Likewise the presence of classes 1 and 4, but never class 7, in all
testcrosses of chimeric female hosts would be strong evidence that such sex
reversals very rarely occur, if at all.
Progeny testing of a male host. These testcrosses are shown in lines (C) and
(D) of Table 1 and will produce wild-type females and y wf36a males derived
from wild-type pole cells of the host. If this male recipient is a sex chimera, he
180
E. B. VAN DEUSEN
also has a 50 % probability of having received either XYX female or XY male
pole cells.
Line C {homosexual combination). Only two classes of progeny are expected
from XY donor pole cells in the testcross: y wf36a females (50 %) and y w/ 3 6 a /
Y, Bs Bar-eyed males (50 %). Only the presence of the Bar-eyed males (class 10)
would demonstrate that XY male pole cells can differentiate functional sperm
in male hosts.
Line D {heterosexual combination). Two classes of progeny can be expected
from XYX donor pole cells which differentiate functional sperm in XY male,
hosts: y wafl y w/ 3 6 a females (50%) and y wf36a females (50%) but no
males. Since thty w/ 3 6 a females (class 12) could not be distinguished from class
9, only the presence of y wafl y w/ 36ft females (class 11) in the testcross would
clearly indicate that an XYX pole cell can undergo phenotypic sex reversal and
give rise to functional sperm.
Germ cell differentiation in chimeric gonads {Tables 2-4)
Six series of experiments (Table 2) totalling 275 successful pole cell transfers produced 55 (20%) fertile flies of which 15 (5-5%) were shown by the
progeny testing schemes to have chimeric gonads containing functional gametes
derived from donor pole cells. All 15 of these chimeras - 7 females and 8 males produced, in addition to progeny derived from host pole cells, progeny which
could have been derived only from donor pole cells which had the same chromosomal sex as their host.
The testcross of seven chimeric females to y wf36a / Y males produced
(Table 3), in addition to wild-type males and females derived from host pole
cells, progeny classes 1, 2, 3 and 4 predicted when donor XYX female pole cells
differentiate functional ova in XX female hosts: phenotypically y wf (class 1)
and y wf36a (class 2) females and y wf36a (class 3) and^ w a /(class 4) males. The
presence of classes 1 and 4 demonstrates that all of the recipient females giving
rise to donor progeny were 'homosexual' XYXjXX germ line chimeras. This
conclusion is also supported by the complete absence of the Bar-eyed male
(class 7) expected if XYVMLIQ pole cells differentiate functional ova in female hosts.
The testcross of eight chimeric males to y w/ 3 6 a females always produced a
1:1 ratio of the two donor progeny classes 9 and 10 (Table 4) predicted when
XY donor pole cells differentiate functional sperm in XY male hosts: y wf3Qa
females (class 9) and y w fUa \ Y, Bs males (class 10). The presence of class
10 males clearly indicates that all of the recipient males giving rise to donor
progeny were 'homosexual' XYjXY germ line chimeras. The regular absence
from the testcross progeny of phenotypically ywf females (class 11) is additional evidence that these male recipients were not 'heterosexual' XYXjXY
germ line chimeras and that XYX female pole cells cannot differentiate functional sperm.
Sex determination in chimeras of Drosophila
181
Table 2. Summary of pole cell transplantation experiments which produced
'homosexual' chimeras
Germ line
chimerism
(donor/host)
Chimeras as determined
Fertile ,
by progeny
Adults adults
testing
9 <J
A.
f
Series Injected Hatched Pupae
1
37
75
8
7
4
(5 %)
11
—
66
11
—
11
(16 %)
111
22
39
16
12
12
(30 %)
IV
30
5
11
5
5
(17%)
V
33
24
15
15
VI
32
29
19
19
11
(33 %)
12
(40 %)
Total
69
275
55
(20 %)
\
0 2
(2-5 %)
0 1
(1-5%)
0 1
(2-5 %)
4 0
(13 %)
2 0
(6%)
1 4
(16%)
7 8
(5-5 %)
XY/XY
XY/XY
XYjXY
XYXIXX
/N.
XYX/XX
XYX/XX
XY/XY
Table 3. Progeny testing of chimeric females
Host
female
no.
1
2
3
4
5
6
7
Progeny phenotypes from donor pole cells
Wild-type
progeny ,
N
33
9?
from
r
host
<
>
^
_>
„
V W f^a
y Wf36a y waf
pole
ywf
Bars S3
(class 1) (class 2) (class 3) (class 4) (class 7)
cells
119
178
91
53
62
172
231
5
4
47
12
0
0
0
38
23
141
11
22
28
13
20
21
77
25
18
6
17
18
14
23
0
8
7
6
0
0
0
0
0
0
0
DISCUSSION
Functional sex reversal can occur in fish, amphibia and birds. Hermaphroditism is relatively common in fish which may undergo transient sex reversal
depending on the stage in their life-cycle (Forbes, 1961) or alternate between
consecutive phases of hermaphroditism depending on the male-to-female ratio
(Fishelson, 1970). Rivulus marmoratus is a unique hermaphroditic fish which
182
E. B. VAN DEUSEN
Table 4. Progeny testing of chimeric males
Progeny from host
pole cells
Host ,
\
W.T.
y\
male
Progeny phenotypes from
donor pole cells
A
r
y wf3Ga
Bars
9
$
no.
9
a
(class 9)
1
2
3
4
5
20
92
66
10
37
35
20
101
12
22
40
9
41
51
12
75
23
39
57
1
8
24
23
17
6
7
8
ywf
9
(class 10) (class 11)
22
31
25
3
7
37
22
15
0
0
0
0
0
0
0
0
lays self-fertilized eggs throughout its reproductive life (Harrington, 1961). In
addition to natural hermaphroditism, functional sex reversal in fish can be induced by mammalian steroids or by Z-irradiation to produce individuals which
are hetero- or homogametic with respect to either of the sex-determining
chromosomes (Mittwoch, 1973).
Genetic analyses of sex-reversed ambystomid salamanders (Humphrey,
1945) and Xenopus (Mikamo & Witschi, 1963) have shown that the germ cells
of sex-reversed males (originally ZW females) and females (originally ZZ males)
retain their original chromosome' constitution and differentiate functional
gametes.
The left gonad of both male (XX) and female (XY) chick embryos is at first
potentially hermaphroditic, while the right gonad has only the potential for male
differentiation (Mittwoch, 1973). Thus if the single left ovary of a hen is removed
surgically, the remaining gonadal remnant can develop into a functional testis
containing X- and 7-bearing sperm.
Sex reversal in mammals, by contrast, has been shown to be limited to the
soma. The phenotypic sex of the germ line seems to be determined entirely by its
chromosome constitution (cf. Short, 1971; Mittwoch, 1973). Natural sex reversal in mammals transforms females into completely sterile males. For example,
germ cells with two X chromosomes reach the testicularized gonad of the intersex goat, intersex pig, sex-reversed mouse (XIX, Sxr / -), bovine freemartin and
the XXYmale, but fail to differentiate functional male gametes. In hermaphroditic pigs and cattle with an XX ovotestis, germ cells survive in the ovarian but
not in the testicular part of the gonad. Attempts to reverse the gonadal sex of
the mammalian embryo by the administration of steroids have been so far
unsuccessful (Burns, 1961).
There is also indirect evidence obtained from the sex ratio of chimeric tetraparental mice and their breeding records (Mystkowska & Tarkowski, 1968;
Sex determination in chimeras o/Drosophila
183
Mintz, 1968; McLaren, 1972, 1975) and direct evidence from chromosomal
studies (Mystkowska & Tarkowski, 1968, 1970; Milet, Mukherjee & Whitten,
1972; McLaren, 1975) that only the XY germ cell population is capable of
differentiating into functional gametes in XYjXX male chimeras. Only a very
small minority of these sex chimeras showed morphological evidence of hermaphroditism. Although the percentage of XXcells detected by karyotype analysis
was in some cases as high as 98 % in blood and 80 % in bone marrow of seven
XYjXX sex chimeras reported by McLaren (1975), all seven developed as
phenotypic males and their genetic component undergoing spermatogenesis as
shown by the genotype of the progeny corresponded in every case to the component identified by chromosome morphology as XY. Thus both XX and XY
primordial germ cells enter the gonadal ridges of a phenotypic male, but the XX
population cannot continue meiosis in a testicular environment. Although the
interpretation of these experiments in chimeric mice is limited mainly by the fact
that mosaicism is not confined to the germ line, they have shown convincingly
that no functional sex reversal of XX germ cells occurs in mice.
The results of the present study have shown that the phenotypic sex of the
germ line in Drosophila, as in mammals, is determined entirely by its chromosome constitution. This is demonstrated by the fact that all 15 wild-type hosts
identified by progeny testing as germ line chimeras (27 % of the surviving fertile
recipients) were 'homosexual' chimeras (XYXjXXor XYjXY). In other words,
the donor component undergoing gametogenesis always had the same genotypic
sex as that of its host. The donor component undergoing spermatogenesis in all
eight male chimeras corresponded to the XY genotype and the donor component undergoing oogenesis in all seven female chimeras corresponded to the
XYX genotype. The remaining 73 % of the surviving wild-type recipients
produced progeny classes expected from their own, but not from a donor
pole cell population.
The probability that this result was obtained just by chance is extremely low.
The probability that all successful pole cell transfers had been made only
between 15 pairs of blastoderms of the same genotypic sex ('homosexual' combinations) is (0-5)15 or 3-05 x 10~5 for both sexes or 7-8 x 10"3 for female and
3-9 x 10~3 for male combinations.
Since the probability of constructing either a 'homosexual' or 'heterosexual*
chimera during pole cell transplantation is 0-5, it is clear that 'heterosexual'
combinations had been tested in the present experiments even if they could not
be detected among 73 % of the fertile recipients which were progeny tested.
Therefore as many as 15 successful male/female combinations were also tested
for sex reversal of the donor pole cell component and the overall success of these
experiments in producing viable chimeras of either kind is likely close to 55 %
(i.e. 2x15/55 = 0-55).
The absence of 'heterosexual' chimerism also has been found by Illmensee
(1973 and personal communication), but in his experiments female (XX) pole
184
E. B. VAN DEUSEN
cells developing into sperm could not have been detected since their lack of a Y
chromosome would lead to sperm immotility.
The inability to detect 'heterosexual' chimeras after pole cell transplantation
also could be explained if none of the donor pole cells in such male/female
combinations entered the gonad but, instead, survived as cuprophilic cells in the
midgut of the host (Poulson & Waterhouse, 1960). Factors such as proliferative
disadvantage of donor cells or small donor 'patch size' which might select
against donor pole cells in 'heterosexual' chimeras do not account fully for the
failure to detect these chimeras since at least 27 % of the fertile recipients in the
present experiments were 'homosexual' germ line chimeras.
The absence of 'heterosexual' chimerism reported here suggests strongly that
donor pole cells which are successfully incorporated into a host gonad of the
opposite genotypic sex cannot reverse their phenotypic sex and differentiate
functional gametes. The conclusion emerging from these results is that the
phenotypic sex of a germ cell in Drosophila is determined entirely by its own
chromosomal constitution, not by that of the gonadal mesoderm.
I am particularly grateful to Professor Walter Gehring for the use of his laboratory facilities and his helpful guidance, to Dr Eric Wieschaus for providing the expertise which led to
the use of these genetic stocks, to Dr Larry Marsh for critically reading the manuscript, and
to the American Cancer Society (Postdoctoral Fellowship No. PF-896) and the Kanton
Basel-Stadt for financial support.
REFERENCES
BRIDGES, C. (1939). Cytological and genetic basis of sex. In Sex and Internal Secretions
(ed. E. Allen), 2nd ed., pp. 15-63. London. Bailliere.
BURNS, R. K. (1961). Role of hormones in the differentiation of sex. In Sex and Internal
Secretions (ed. W. C. Young), 3rd ed., pp. 76-158. London: Bailliere, Tindall and Cox.
FJSHELSON, L. (1970). Protogynous sex reversal in the fish Anthias squamipinnis (Teleostei,
Anthiidae) regulated by the presence or absence of a male fish. Nature, Lond. 227, 90-91.
FORBES, T. R. (1961). Endocrinology of reproduction in cold-blooded vertebrates. In Sex and
Internal Secretions (ed. W. C. Young), 3rd ed., pp. 1035-1087. London: Bailliere, Tindall
and Cox.
HARRINGTON, R. W., JR. (1961). Oviparous hermaphroditic fish with internal self-fertilization. Science, N. Y. 134, 1749-1750.
HUMPHREY, R. R. (1945). Sex determination in ambystomid salamanders: a study of the
progeny of females experimentally converted into males. Am. J. Anat. 16, 33-66.
ILLMENSEE, K. (1973). The potentialities of transplanted early gastrula nuclei of Drosophila
melanogaster. Production of their imago descendants by germ-line transplantation. Wilhelm Roux Arch. EntwMech. Org. 171, 331-343.
LINDSLEY, D. L. & GRELL, E. H. (1968). Genetic variation in Drosophila melanogaster.
Carnegie Inst. Wash. Publ. No. 627.
MCLAREN, A. (1972). Germ cell differentiation in artificial chimeras of mice. Proc. Int. Symp.
The Genetics of the Spermatozoon (ed. R. A. Beatty & S. G. Waelsch), pp. 313-324. Edinburgh and New York: R. A. Beatty and S. G. Waelsch.
MCLAREN, A. (1975). Sex chimerism and germ cell distribution in a series of chimeric mice.
/. Embryol. exp. Morph. 33, 205-216.
MIKAMO, K. & WITSCHI, E. (1963). Functional sex reversal in genetic females of Xenopus
induced by implanted testes. Genetics, Princeton 48, 1411-1421.
Sex determination in chimeras 0/Drosophila
185
MILET, R.
G., MUKHERJEE, B. B. & WHITTEN, W. K. (1972). Cellular distribution and possible
mechanism of sex-differentiation in XY/Xychimeric mice. Can. J. gen. Cytol. 14, 933-941.
MINTZ, B. (1968). Hermaphroditism, sex chromosomal mosaicism and germ cell selection in
allophenic mice. / . Animal Sci. 27 (Suppl. 1), 51-60.
MITTWOCH, U. (1973). Genetics of Sex Differentiation. New York: Academic Press.
MYSTKOWSKA, E. T. & TARKOWSKI, A. K. (1968). Observations on CBA-p/CBA-T6T6
mouse chimeras. /. Embryol. exp. Morph. 20, 33-52.
E. T. & TARKOWSKI, A. K. (1970). Behavior of germ cells and sexual differentiation in late embryonic and early postnatal mouse chimeras. /. Embryol. exp. Morph.
23, 395-405.
POULSON, D. F. & WATERHOUSE, D. F. (1960). Experimental studies on pole cells and midgut
differentiation in Diptera. Aust. J. biol. Sci. 13, 541-567.
SEIDEL, S. (1960). Aufhebung der Sterilitat von 'Transformer' Mannchen durch Erzeugung
von Keimbahn-Mosaiken. Naturwissenschaften 47, 609-610.
SEIDEL, S. (1963). Experimented Untersuchungen iiber die Grundlagen der Sterilitat von
Transformer (fra)-Mannchen bei Drosophila melanogaster. Zeitschrift fur Vererbungslehre
94, 215-241.
SHORT, R. V. (1971). Germ cell sex. Proc. Int. Symp. The Genetics of the Spermatozoon (ed.
R. A. Beatty & S. G. Waelsch), pp. 325-345. Edinburgh and New York: R. A. Beatty
and S. G. Waelsch.
^
STERN, C. & SHERWOOD, E. R. (1962). Can primordial germ cells of the genotype XXY
produce functional sperm? Drosophila Information Service 36, 118.
STURTEVANT, A. H. (1945). A gene in Drosophila melanogaster that transforms females into
males. Genetics, Princeton, 30, 297-299.
MYSTKOWSKA,
(Received 12 July 1976)