In Anura the primordial germ-cells are discernible in the dorsal crest endoderm of tail-bud stages of development and may be traced from this position throughout their migration into the undifferentiated gonadal rudiment. These facts have been established by the descriptive studies of a number of workers (see review by Johnston, 1951), the cells being recognizable by their large size, the retention of yolk platelets long after their disappearance in neighbouring cells, the sharply defined and often kidney-shaped nuclear membrane, and the poor staining affinity of the nuclear contents.

By means of the application of the Altmann-Volkonsky staining technique, Bounoure (1934) was able to demonstrate that germ-cells of the dorsal crest endoderm are the lineal descendants of certain cells found in the ventral region of the blastula. This discovery has been confirmed for Rana temporaria (the species investigated by Bounoure) by Blackler (1958), and extended to other Anuran species by Nieuwkoop (1956 a, b), Blackler (1958), and Di Berardino (1961).

In a recent experimental study by Blackler (1960 a, b) and Blackler & Fisch-berg (1961), regions containing the so-called primordial germ-cells were grafted from one Xenopus laevis embryo to another at the neurula stage. Graft cells were distinguished from host cells with the help of a nuclear marker. The object of this experiment was to ascertain if the stainable gonocytes of neurulae are indeed primordial germ-cells by demonstrating that they give rise to definitive gametes.

The results showed that true gonocytes occur in neurulae, since these grafts contributed at least some of the functional gametes, although whether these germ-cells are identical with those observed in histological studies remains a matter awaiting proof.

This present paper describes experiments which follow up the success in establishing a germ-cell transfer technique by exploring the possibilities of the technique for elucidating another long-standing problem of germ-cell studies and in making a contribution to the study of oogenesis.

In female animals the maturation of the germ-cells takes place in the ovarian environment and it is possible that ovarian influences manifest themselves in the morphogenetic events that take place in an egg after its ovulation and subsequent fertilization. Indeed, Raven (1958) has already suggested the possibility that the follicle cells impart information which is stored in the cortex of the egg. Normal females of the two subspecies employed in this present analysis produce eggs which are distinctive in their size, colour, and later development, so it was reasonable to assume that if the germ-cells of one sub-species were developing in the ovarian environment of the other subspecies, the nature of the eggs finally produced would be some measure of the influence of the host ovary on the growth of the graft oocyte. Moreover, the point at issue is particularly relevant to the germ-line theory, i.e. that theory claiming the mutual independence of the reproductive and somatic lines of cells.

Furthermore, there has long been controversy in germ-cell studies concerning the opposing views that the primordial germ-cells give rise to all the definitive gametes or that these cells give rise to only some gametes (the remainder deriving from some secondary, somatic, source). The transfers of germ-cells between the two subspecies were also based on the possibility that at least some experimental animals might be obtained whose gametes were entirely of graft origin, in which case the view of complete integrity of the germ-line would be upheld.

The two subspecies used were laboratory specimens of X. laevis laevis (Daudin) and X. laevis victorianus (Ahl).

X. laevis laevis (hereafter referred to as XU) is the usual laboratory form of X. laevis. Mature females lay eggs varying from 1·3 to 1·6 mm. in diameter, all the eggs of any one female being of the same size and the smaller eggs being usually laid by females shortly after gaining maturity (Plate, fig. A). The animal hemisphere pigment is rather variable, but the commonest colour is a rich chocolate brown (beautifully imitated in colour drawings published by Bles, 1905).

X. laevis victorianus (hereafter referred to as Xlv) is the Uganda form of X. laevis, smaller in adult size than Xll and clearly recognizable by other characters as well. Mature females lay eggs varying only from 1·00 to 1·09 mm. in diameter, and the commonest egg colour is a very pale brown (Plate, fig. B).

The subspecific status of the two forms, founded by Parker (1936), has been experimentally tested by Blackler, Fischberg, & Newth (in preparation) who find that Xll and Xlv are mutually interfertile, as well as the offspring of such hybrid crosses.

The development of eggs of the two subspecies has also been subjected to examination by Blackler, Fischberg, & Newth, and two characteristics of this development are relevant to the present paper and demand some attention. The first is that black melanophores appear on the rectal tube in Xlv at stages 49-50 of Nieuwkoop & Faber’s (1956) table of Xenopus development (Plate, fig. C) but do not manifest themselves until stage 57, i.e. just before the onset of metamorphosis, in Xll (Plate, fig. D). Further to this, it is clear that the genes affecting the time of appearance of these melanophores in Xlv are dominant over those of Xll, as revealed in the study of the reciprocal hybrids in which all develop like Xlv in this respect. The significance of these data is commented on below.

The technique of germ-cell grafting was almost the same as that described by Blackler & Fischberg (1961). The primordial germ-cells in the piece of endoderm grafted were later constituents of the experimental gonad, so it can be justifiably assumed that they had migrated out of their graft milieu and come into contact with the somatic cells of the host gonad. The slight variation in the technique lay in transplanting between neurulae of stages 21–23 and not stage 26, since in their discussion of graft failures Blackler & Fischberg (1961) drew attention to the possibility that the germ-cells of this latter stage were no longer present in the piece of tissue grafted from one embryo to another.

Grafts were made reciprocally with respect to both subspecies. The smaller egg size of Xlv causes the neurulae to be smaller than neurulae of Xll at stage 23 (2·2 mm. as opposed to 2·5 mm.). This difference in size is small and yet is sufficient to make the grafting of pieces of Xll neurulae into Xlv more difficult than the reciprocal operation. One may add that this difficulty lies not in the actual manipulation but in transferring a piece of Xll neurula tissue large enough and likely to incorporate the primordial germ-cells.

In order to distinguish between graft and host cells, the nuclear marker discovered by Fischberg (Elsdale, Fischberg, & Smith, 1958) has been used. In theory, the removal of the host germ-cell region should result in host sterility with respect to the host germ-cells, but in practice some of the host’s germ-cells can be left behind. The nuclear marker has been described so often in publications from this laboratory (Elsdale, Fischberg, & Smith 1958; Fischberg, Gurdon, & Elsdale, 1958; Gurdon, 1959; Elsdale, Gurdon, & Fischberg, 1960; Gurdon, 1960 a, b;,Fischberg & Wallace, 1960; Wallace, 1960; Blackler, 1960 a, b;Blackler & Fischberg, 1961) that a further repetition is hardly justified. The reader is referred to Blackler & Fischberg (1961) for an account of the use of the nuclear marker in the germ-cell transfer technique.

The results of the experiments performed have involved observations of nuclei (for the marker), egg diameters and pigmentation, and the times of appearance of the rectal tube melanophores.

Grafts

Forty-eight transfer operations were performed, of which 34 were successful in producing metamorphosed frogs. The successful healing-in of grafts was thus a little over 70 per cent, of all attempted transfers. This is an improvement on the healing (40 per cent.) obtained in the development of the technique as reported in Blackler & Fischberg (1961). Applications of the technique for other purposes, not commented on here, have resulted in over 90 per cent, healing. This improvement has been obtained by culturing the experimental embryos in 0·1 Niu & Twitty solution until they reach the feeding stage. After an operation the obliteration of the edges of the graft by ectodermal overgrowth is a poor criterion of successful healing and the graft may break down at any time between the neurula and swimming tadpole stages. Transfer of experimental embryos to water too soon after operation increases the chance of graft breakdown.

Not all the 34 experimental frogs obtained have been tested for graft success. Two were retarded in growth, 2 died shortly after metamorphosis, and 12 males derived from operations in which the nuclear marker was not used were not tested.

Transfer of 2-nucleolate germ-cells of Xlv to 1-nucleolate Xll recipients

All 9 experimental frogs of this combination developed into males and have been mated with 2-nucleolate Xll females. Xll females were chosen since (a) it is easier to mate Xll with Xll than Xll with Xlv because of adult size differences, and (b) the graft success, as derived from observation of nuclei in tadpoles, can be checked against the percentage graft success obtained from observations of rectal tube melanophores (which, as explained above, are dictated in their time of appearance by Xlv genes).

When a non-experimental 1-nucleolate (1 nu) frog1 is mated with a normal 2-nucleolate (2 nu) frog, approximately half the offspring are 1 nu and half 2 nu. Thus, in the case of 2 nu-in-l nu experimental frogs mated to 2 nu normal frogs, the percentage graft success may be calculated from the numerical excess of 2 nu over 1 nu offspring. Employing this principle, the graft success of the 9 males in this group was calculated from the nucleolarities of their offspring, as recorded in Table 1.

TABLE 1.

Transplantation of Xlv germ-cells into Xll at the neurula stage, using the nuclear marker

Transplantation of Xlv germ-cells into Xll at the neurula stage, using the nuclear marker
Transplantation of Xlv germ-cells into Xll at the neurula stage, using the nuclear marker

The nucleolarities of the tadpoles in Table 1 were determined in phase-contrast preparations of the tail-tips. Other tadpole samples of the same matings were kept and scored for the appearance of rectal tube melanophores as following the Xlv pattern (st. 50) or the Xll pattern (st. 57). Graft success, as calculated from these observations, is recorded in Table 2. It is apparent that very comparable values for graft successes are obtained by using the two different methods of assessment.

TABLE 2.

Transplantation of Xlv germ-cells into Xll at the neurula stage

Transplantation of Xlv germ-cells into Xll at the neurula stage
Transplantation of Xlv germ-cells into Xll at the neurula stage

Transfer of 1-nucleolate germ-cells of Xll neurulae to 2-nucleolate Xlv recipients

Six experimental frogs of this combination have been obtained and mated with 2-nucleolate Xlv frogs. Once again, the choice of the normal frog was governed by the ease of mating, although in this case the observation of melano-phore appearance cannot be used to determine graft success.

When a 1 w-in-2 nu experimental frog is mated with a normal 2 nu frog, any 1 nu tadpoles among the offspring must have been derived from graft cells. Moreover, a proportion of the 2 nu offspring (numerically equal to the 1 nu tadpoles) are also of graft origin although these cannot be distinguished from those offspring derived from host cells. The percentage graft success may, therefore, be calculated from the number of 1 nu tadpoles. Graft success in the 6 frogs of this group, using this principle, was calculated from the data set out in Table 3.

TABLE 3.

Transplantation of Xll germ-cells into Xlv at the neurula stage, using the nuclear marker

Transplantation of Xll germ-cells into Xlv at the neurula stage, using the nuclear marker
Transplantation of Xll germ-cells into Xlv at the neurula stage, using the nuclear marker

Three of the 6 frogs were females, and it became clear that observation of the type of egg laid could serve as a check on graft success as calculated from nucleolarity. In Table 3, ♀ ♀ 10 and .12, both negative for graft success as calculated from the nucleolarity of their offspring, laid eggs absolutely typical of Xlv—that is, with a mean diameter of 1 mm. and pale grey-brown pigment, ♀11, on the other hand, demonstrated the graft origin of her oocytes by laying typically Xll eggs, with a mean diameter of 1-32 mm. and coloured chocolate brown. The eggs of ♀11 are presented in Plate, fig. A. Every egg laid by this female was of Xll type and nothing could be observed, either in the eggs or their mode of development, that indicated qualities in them intermediate between Xlv and Xll that could be ascribed to the ovarian environment. For ♀11, observations of the nucleolarity of offspring and the type of eggs from which they developed indicated total graft success.

Transfers of germ-cells between Xlv and Xll in which the nuclear marker was not employed

(a) Xll germ-cells transferred to Xlv

Seven frogs were obtained from operations in this group, but all were male and have not been tested. Graft success could have been determined from observations of melanophores in offspring derived from matings with Xll frogs, but such crosses are obtained only with difficulty and it was unlikely that any information further to that obtained from transfers of this kind in which the nuclear marker had been used would be forthcoming.

(b) Xlv germ-cells transferred to XU

Eight frogs were obtained from operations in this group. Of these, 5 were male and have not been tested, although graft success could have been determined from melanophore observations. The object in testing the females was to obtain further data on the type of eggs laid.

The 3 female frogs in this group were mated to normal Xll males. The results of mating these females are recorded in Table 4 and it can be readily appreciated that all three operations were successful. The great majority of eggs laid were of Xlv type and hence these females are the reciprocals of ♀11 (vide supra) in respect of egg type. Again, no indications of qualities that could be ascribed to the ovarian environment were observable, and the melanophore observations on offspring support the egg data.

TABLE 4.

Transplantation of Xlv germ-cells into Xll at the neurula stage

Transplantation of Xlv germ-cells into Xll at the neurula stage
Transplantation of Xlv germ-cells into Xll at the neurula stage

Grafting technique

In their 1961 study, Blackler & Fischberg were unable to obtain female frogs showing graft success. However, in the present analysis females as well as males have been obtained which show graft success. Thus there is no proof that transfers of primordial germ-cells into female recipient embryos are less likely to be successful than transfers into males. The numbers of experimental frogs obtained and tested have been small, but, nevertheless, it does seem that more successful transfers result when neurulae not later than stage 23 are used.

Egg type

Female frogs of one subspecies showing graft success have laid eggs typical of the other subspecies. These eggs were the direct descendants of those primordial germ-cells originally present in the graft and which have migrated from it into the ‘foreign’ somatic environment of the host ovary. The results are limited in that the eggs are defined as of graft type on only two characters— dimension and colour, and involve only subspecific grafts. Nevertheless, it is clear that the size and colour of an egg is, in this graft combination, determined by the egg itself and not by the ovarian environment in which it develops. Although the egg acquires much of its substance from the host via the somatic cells of the ovary, it seems that the organization of this substance is specified by the egg itself. Once again, the independence of the germ-line is manifest.

In conclusion, one could go further and cite this control of egg type as being centred in the egg nucleus. Gurdon (1961) has been able to show, by means of nuclear transplantations between the two subspecies used in this study, that the kind of frog produced, and the type of eggs laid by female transplant-frogs, is determined by the subspecific nature of the nuclei transplanted.

Reproductive metaplasia

The primordial germ-cell grafting technique, employing the nuclear marker, has shown that primordial germ-cells exist in neurula stages and that these cells give rise directly to cells which, in turn, form the definitive gametes. However, the cases of complete graft success (100 per cent.) can be cited as evidence against the hypothesis of a secondary, somatic, origin of some definitive sex-cells (reproductive metaplasia). This hypothesis originated in a paper by Butcher (1929) on the germ-cells of the lamprey, in which the gradual transition of peritoneal cells into primordial germ-cells was described. The issue has remained a controversial topic in germ-cell studies, and evidence for the existence of such reproductive metaplasia, derived entirely from descriptive studies, permeates the literature (particularly the literature concerned with the germ-line of mammals).

Experimental approaches to the problem of reproductive metaplasia have been few. In birds, Benoit (1930) and Dulbecco (1946) by irradiation methods, and Dantschakova (1931) by cauterization, have shown that removal of early embryonic extra-gonadal germ-cells leads to the production of embryonic sterile gonads. However, the design of these experiments does not eliminate the possibility that such gonads might subsequently acquire a complement of sex-cells from a somatic source. The experiment of Domm (1929) does, by contrast, point to the incapacity of the post-embryonic bird ovary to produce gonocytes after destruction of the primordial germ-cells. Briefly outlined, Domm found that extirpation of the left ovary is followed by hypertrophy of the normally atrophying right ovary to form a testis. This testis is only fertile, however, if the extirpation of the left ovary is performed while the primordial germ-cells of the right ovary have not degenerated. This degeneration is complete 3 weeks after hatching of the chick and operations result only in the production of sterile right testes thereafter.

In mammals, Mintz & Russell (1957) have worked on the problem of sterility in mice caused by the presence of certain mutated genes. These workers find that the primordial germ-cells are segregated normally in the embryos but that they degenerate in the course of their migration to the gonadal sites. The analyses of Mintz & Russell strongly deny the existence of reproductive meta-plasia in the mouse, although it should be borne in mind that their mice were abnormal in ways other than sterility, and the abnormal genotype might affect the expression of metaplasia adversely.

To these experimental approaches in birds and mammals, the present contribution for the frog Xenopus may be added. In this study certain experimental frogs, absolutely normal in their phenotype but containing germ-cells of graft origin, produce normal eggs and sperm entirely of the graft type. The conclusion must be that the existence of reproductive metaplasia in vertebrates is not only unproven, but is, to say the least, highly improbable. The germ-line is not only distinct from the somatic line in adults (shown unequivocally for the guinea-pig in an experimental study by Castle & Phillips, 1911) but in embryos as well.

  1. Transplantations of primordial germ-cells have been carried out between two subspecies of X. laevis at the neurula stage according to the technique of Blackler & Fischberg (1961). Adult experimental frogs have been tested for graft success by matings with normal frogs, and graft success quantitatively assessed with the help of a nuclear marker, rectal tube melanophores, and egg-type.

  2. Some experimental frogs only produced gametes of graft origin. This finding is adduced as strong evidence against the theory which claims that some normal gametes originate from sources other than the primordial germ-cells.

  3. After successful grafting, female frogs of one subspecies laid eggs in size and colour identical with those of the other subspecies and no intermediate characters were observed. It is concluded that for the two subspecies used, the oocyte specifies these egg characters itself and independently of the ovarian environment in which it develops.

  1. Les transplantations des cellules germinales primordiales ont été pratiquées entre les deux sous-espèces de Xenopus laevis au stade neurula, selon la technique de Blackler & Fischberg (1961). Pour le contrôle du succès des greffes, les grenouilles expérimentales adultes ont été accouplées avec des grenouilles normales et le résultat quantitatif des greffes a été évalué par les signes nucléaires, les mélanophores du tube rectal et par le type de l’œuf.

  2. Certaines grenouilles expérimentales ont produit seulement des gamètes d’origine provenant de la greffe. Cette constation est une preuve évidente contre la théorie qui pretend que certains gamètes normaux ont comme origine d’autres sources que les cellules germinatives primordiales.

  3. Parmi les greffes réussies, les grenouilles femelles d’une des sous-especès ont pondu des œufs identiques, et de taille et de couleur, à l’autre sous-espèce. Des caractères intermédiaires n’ont pas été observés. Il a été conclu que dans les deux sous-espèces expérimentées, seules les oocytes sont responsables de ces caractères de l’œuf et celà indépendamment de l’environnement de l’ovaire dans lequel il se développe.

It is a pleasure to acknowledge the advice afforded by my colleague Prof. M. Fischberg and the technical assistance of Miss J. D. McConnell. The author wishes also to thank Mme P. Ahmad-Zadeh for some help with the manuscript and the British Empire Cancer Campaign for a grant to support work of which this study forms a part.

Benoit
,
J.
(
1930
).
Destruction des gonocytes primaires dans le blastoderme du poulet par les rayons ultra-violets, aux premiers stades du développement embryonnaire
.
Proc. sec. Inter. Congr. Sex Res
.,
162
70
.
Blackler
,
A.
(
1958
).
Contribution to the study of germ-cells in Anura
.
J. Embryol. exp. Morph
.
6
,
491
503
.
Blackler
,
A.
(
1960a
).
Transfer of germ-cells in Xenopus laevis
.
Nature, Lond
.
85
,
859
60
.
Blackler
,
A.
(
1960b
).
‘The transfer of primordial germ-cells in Xenopus laevis.’
In Germ-cells and Earliest Stages of Development, symposium of I.I.E.—Pallanz
.
1960
,
25
28
.
Blackler
,
A.
Fischberg
,
M.
(
1961
).
Transfer of primordial germ-cells in Xenopus laevis
.
J. Embryol. exp. Morph
.
9
,
634
41
.
Bles
,
E. J.
(
1905
).
The life history of Xenopus laevis, Daudin
.
Trans. Roy. Soc. Edinb
.
51
,
789
819
.
Bounoure
,
L.
(
1934
).
Recherches sur la lignée germinale chez la grenouille rousse aux premiers stades du développement
.
Ann. Sci. natur. zool
., 10e sér.,
17
,
67
248
.
Butcher
,
E. O.
(
1929
).
The origin of the germ-cells in the lake lamprey (Petromyzon marinus unicolor)
Biol. Bull
.
56
,
87
99
.
Castle
,
W. E.
, &
Phillips
,
J. C.
(
1911
).
On germinal transplantation in Vertebrates
.
Contr. Embryol. Carneg. Instn
.,
144
,
1
26
.
Dantschakova
,
V.
(
1931
).
Keimzelle und Gonade. Die entodermale Wanderzelle als Stammzelle in der Keimbahn
.
Z. Zellforsch
.
14
,
376
84
.
Diberardino
,
M. A.
(
1961
).
Investigations of the germ-plasm in relation to nuclear transplantation
.
J. Embryol. exp. Morph
.
9
,
507
13
.
Domm
,
L. V.
(
1929
).
Spermatogenesis following early ovariotomy in the brown leghorn fowl
.
Roux Arch. EntwMech. Organ
.,
119
,
171
87
.
Dulbecco
,
R.
(
1946
).
Sviluppo di gonadi in assenza di cellule sessuali nell’embrione di polio
.
R. C. Acc. Naz. Lincei
,
8
,
1211
13
.
Elsdale
,
T. R.
,
Fischberg
,
M.
, &
Smith
,
S.
(
1958
).
A mutation that reduces nucleolar number in Xenopus laevis
.
Exp. Cell Res
.
14
,
642
3
.
Gurdon
,
J. B.
, &
Fischberg
,
M.
(
1960
).
A description of the technique for nuclear transplantation in Xenopus laevis
,
J. Embryol. exp. Morph
.
8
,
437
44
.
Fischberg
,
M.
,
Gurdon
,
J. B.
, &
Elsdale
,
T. R.
(
1958
).
Nuclear transfer in Amphibia and the problem of the potentialities of the nuclei of differentiating tissues
.
Exp. Cell Res. Suppl
.
6
,
161
78
.
Fischberg
,
M.
, &
Wallace
,
H.
(
1960
).
A mutation which reduces nucleolar number in Xenopus laevis
.
In The Cell Nucleus
, ed.
B.
Holmes
,
London
:
Butterworth
.
Fischberg
,
M.
, &
Wallace
,
H.
Gurdon
,
J. B.
(
1959
).
Tetrapioid frogs
.
J. exp. Zool
.
141
,
519
43
.
Fischberg
,
M.
, &
Wallace
,
H.
Gurdon
,
J. B.
(
1960a
).
The effects of ultra-violet irradiation on uncleaved eggs of Xenopus laevis
.
Quart. J. micr. Sci
.,
101
,
299
311
.
Gurdon
,
J. B.
(
1960b
).
The developmental capacity of nuclei taken from differentiating endoderm cells of Xenopus laevis
.
J. Embryol. exp. Morph
.
8
,
505
26
.
Gurdon
,
J. B.
(
1961
).
Transplantation of nuclei between two subspecies of Xenopus laevis
.
Heredity
,
16
,
305
15
.
Johnston
,
P. M.
(
1951
).
The embryonic history of the germ-cells of the large mouth Black Bass, Micropterus salmoides salmoides (Lacépède)
.
J. Morph
.
88
,
471
542
.
Mintz
,
B.
, &
Russell
,
E. S.
(
1957
).
Gene-induced embryological modifications of primordial germcells in the mouse
.
J. exp. Zool
.
134
,
207
37
.
Nieuwkoop
,
P. D.
(
1956a
).
In Normal Table of Xenopus laevis (Daudin). See below
.
Nieuwkoop
,
P. D.
(
1956b
).
Are there direct relationships between the cortical layer of the fertilized egg and the development of the future axial system in Xenopus laevisl Publ
.
Staz. zool. Napoli
,
28
,
241
9
.
Nieuwkoop
,
P. D.
Faber
,
J.
(
1956
).
Normal Table of Xenopus laevis (Daudin)
.
Amsterdam
:
North Holland Publ. Co
.
Parker
,
H. W.
(
1936
).
Reptiles and Amphibians collected by the Lake Rudolf Rift Valley Expedition
.
Ann. Mag. Nat. Hist
.
18
,
596
601
.
Raven
,
C. P.
(
1958
).
Information versus Preformation in embryonic development
.
Arch, néerl. Zool
.
13
,
185
93
.
Wallace
,
H.
(
1960
).
The development of anucleolate embryos of Xenopus laevis
.
J. Embryol. exp. Morph
.
8
,
405
13
.

FIG. A. Photomicrograph to show the appearance of the eggs of X. laevis laevis (blastula stage). The eggs were actually laid by a female of X. laevis victorianus but were indistinguishable from normal Xll eggs.

FIG. B. Photomicrograph to show the appearance of the eggs of Xlv (blastula stage). The eggs were actually laid by a young female of Xll but were indistinguishable from normal Xlv eggs. The photograph is presented at the same magnification as that of fig. A for direct comparison.

FIG. C. Photograph to show the rectal tube region of a stage 50 tadpole of Xlv (the shape of the limb bud gives the stage number). Melanophores are visible at this stage in the wall of the rectal tube.

FIG. D. Photograph to show the rectal tube region of a stage 54 tadpole of Xll (the shape of the foot gives the stage number). Melanophores are not visible at this stage in the wall of the rectal tube. HLB, hind limb-bud; HLF, hind limb-foot; VTF, ventral tail fin; LL, lateral line; RT, rectal tube; M, melanophores.

FIG. A. Photomicrograph to show the appearance of the eggs of X. laevis laevis (blastula stage). The eggs were actually laid by a female of X. laevis victorianus but were indistinguishable from normal Xll eggs.

FIG. B. Photomicrograph to show the appearance of the eggs of Xlv (blastula stage). The eggs were actually laid by a young female of Xll but were indistinguishable from normal Xlv eggs. The photograph is presented at the same magnification as that of fig. A for direct comparison.

FIG. C. Photograph to show the rectal tube region of a stage 50 tadpole of Xlv (the shape of the limb bud gives the stage number). Melanophores are visible at this stage in the wall of the rectal tube.

FIG. D. Photograph to show the rectal tube region of a stage 54 tadpole of Xll (the shape of the foot gives the stage number). Melanophores are not visible at this stage in the wall of the rectal tube. HLB, hind limb-bud; HLF, hind limb-foot; VTF, ventral tail fin; LL, lateral line; RT, rectal tube; M, melanophores.

1

In previous papers from this laboratory we have used the abbreviation N for nucleolarity. Since this has sometimes been confused with n, or N used in studies of chromosome ploidy, it has been decided to adopt the abbreviation nu in all future publications.