/. Embryol. exp. Morph. Vol. 59, pp. 113-130, 1980
Printed in Great Britain © Company of Biologists Limited 1980
\\ 3
Somite formation in cultured embryos of the
snapping turtle, Chelydra serpentina
By DAVID S. PACKARD, JR. 1
From the Department of Anatomy,
Upstate Medical Centre, New York
SUMMARY
A simple, reliable method for the in vitro cultivation of snapping turtle embryos was demonstrated. This technique was used to study somite formation in explants containing
segmental plates. Segmental plates formed a full complement of somites whether the neural
tube or the neural tube and notochord was present. Explanted snapping turtle segmental
plates formed an average of 6-6 ± 1-2 somites. Removal of the node region or tail bud from
cultured intact embryos led to a cessation of somite formation after an additional 6-1 ±1-8
somites had formed. These results indicate the number of somites the snapping turtle segmental
plate will form. Also, the number of somites formed by explanted segmental plates showed
only slight variation over a wide range of segmental plate lengths. It was concluded that while
snapping turtle segmental plates formed significantly fewer somites than chicken or Japanese
quail segmental plates, they were similar to the avian explants in their ability to form a
consistent number of somites regardless of the length of the segmental plate.
INTRODUCTION
Since metameric pattern is fundamental to vertebrate anatomy, the specification and morphogenesis of the original somite pattern is of interest. It had
been suggested that the neural tube and/or notochord adjacent to the unsegmented somite mesoderm played an essential role in somitogenesis (Fraser,
1960; Menkes, Miclea, Elias & Deleanu, 1961) but subsequent studies found
that avian and amphibian somitic mesoderm could segment without further
contact with the neural tube or notochord (Bellairs, 1963; Sandor & Amels,
1970; Lanot, 1971; Christ, Jacob & Jacob, 1972; Brustis & Gipouloux, 1973;
Packard & Jacobson, 1976). The evidence now indicates that in avian embryos
the commitment of mesoderm to form somites is closely associated with
Hensen's node (Nicolet, 1970; Hornbruch, Summerbell & Wolpert, 1979) and
does not occur in the node's absence (Butros, 1962; Bellairs, 1963; Butros,
1967; Packard & Jacobson, 1976). Consistent with this view of the role of the
node region in somite formation is the concept that the unsegmented somite
mesoderm is committed to somite formation by the time it condenses into a
1
Author's address: Department of Anatomy, State University of New York, Upstate
Medical Center, Syracuse, New York 13210, U.S.A.
114
D. S. PACKARD
segmental plate. The segmental plate mesoderm will not only segment in the
absence of the axial structures, but also without the tissues normally lateral
or anterior to it (Packard & Jacobson, 1976; Packard, 1978) without the
ectoderm or hypoblast (Sandor & Amels, 1971), or when grafted to various
heterotopic locations (Nicolet, 1971; Christ et al 1972, 1974; Menkes &
Sandor, 1977). The passage of some physical signal along the unsegmented
mesoderm has been suggested as an essential part of somite formation (Lanot,
1971) but experimental manipulations that would have interrupted the passage
of such signals failed to prevent further somite formation (Menkes et al. 1961;
Christ et al. 1974; Menkes & Sandor, 1977; Packard, 1978). This still leaves
the possibility that a kinematic wave of some kind might occur (Pearson &
Elsdale, 1979).
Packard and Jacobson (1976) reported that chick segmental plates contained
10 to 12 prospective somites. They further demonstrated that if a cultured
chick segmental plate was cut into two pieces, each piece formed the number
of somites it would have formed had the cut not been made (Packard, 1978).
One interpretation of this result was that the segmental plate contained a
committed somite pattern. This interpretation was supported when Meier
(1979) showed that chick segmental plates contained a subtle, but demonstrable,
metameric pattern along their complete length. This pattern consisted of 10
to 11 repeating, bilaterally paired mesodermal cell domains. Meier suggested
that these ' somitomeres' became the somites.
Another observation made by Packard & Jacobson (1976) was that cultured
segmental plate explants from older chick embryos with longer segmental
plates did not tend to form more somites than explants from younger embryos.
It was suggested that some phenomenon, possibly similar to those demonstrated
by Cooke (1975) in Xenopus embryos and by Flint, Ede, Wilby & Proctor
(1978) in mouse embryos, might be adjusting the somite pattern to the available
length of segmental plate. More recently, this phenomenon was studied in.
Japanese quail embryos (Packard, 1980). It was found that, much like chick
segmental plates, explanted quail segmental plates formed 11-3 ±2-9 somites.
The present experiments were designed to find if the above observations
were peculiar to avian embryos by investigating somitogenesis in embryos of
the snapping turtle. Somite formation in these animals appears to be similar
to avian somite formation (Yntema, 1968). Yntema (1970) showed that unilateral extirpation of three somites and the associated ectoderm and neural tube
led to deficiencies of the ribs and carapace. He concluded that by the time the
somites of the turtle are segmented from the dorsal mesoderm, they behave
as a mosaic rather than an equipotential system. It was found in the study
described here that snapping turtle segmental plates formed 6-6 ± 1-2 somites.
The number of somites formed varied only slightly over a wide range of
segmental plate lengths.
Somite formation in the snapping turtle
115
MATERIALS AND METHODS
Gravid females of the snapping turtle, Chelydra serpentina, were collected
in the late spring of 1977-1979 from various locations in the upstate region
of New York State. Eggs were collected and stored as described elsewhere
(Yntema, 1964). The eggs were rinsed three times with sterile tap water and
stored at 20 °C in sterile finger bowls. Within 24 h of collection, the eggs
were candled and a portion of the shell and its membrane overlying the embryos
was removed (Yntema, 1964). Eggs thus prepared were incubated in finger
bowls at 20 °C until taken for in vitro culture.
Intact Chelydra embryos and tissue explants derived from them were cultured
on agar medium according to the procedure for chick embryos described by
Packard & Jacobson (1976). Initially, supernatant derived from the centrifuged contents of turtle eggs was successfully substituted for the chicken egg
supernatant of the original technique. This was, however, very wasteful of
the few turtle eggs available for study. It was found that Chelydra embryos
develop on the agar substrate and fluid medium made from fertile, unincubated
chicken eggs. The only modification to the chick technique was that 1 ml of
a mixture of penicillin (10000 units/ml) and streptomycin (10000 mcg/ml
Grand Island Biological Company) and 0 1 g of glucose were added to each
100 ml of fluid medium.
Each embryo was prepared for culture by first wiping the egg with 70 %
ethanol and making an equatorial cut through the shell and subjacent yolk.
Yolk was washed from the shell's dorsal hemisphere with Howard's (1953)
saline and the embryo was carefully peeled from the vitelline membrane. The
embryo was transferred to a 35 mm plastic petri dish containing the nutritive
agar substrate and about 1 ml of fluid culture medium. The embryo was placed
on the substrate with its endodermal surface against the agar and excess fluid
medium was removed. Cuts appropriate to each experiment were made through
all three germ layers of the embryo with tungsten wire needles and the plate
was cultured in a humidified atmosphere of 95 % O2, 5 % CO2 at 30 °C. Most
cultures were maintained for three days. On the second and third days the
cultures were 'fed' by adding about 1 ml of fresh fluid medium to each dish,
gently swirling the dish for about 30 sec and drawing the excess medium
off.
The lengths of segmental plates were measured with an ocular micrometer
following explantation but before operations were performed. To insure that
the segmental plates were not distorted during explantation, the lengths of
the segmental plates of several embryos were measured immediately before
and after explantation. No significant alterations were found.
The methods of histological analysis (Packard & Jacobson, 1976) and
photography of cultured tissues (Packard, 1978) have been described.
116
D. S. P A C K A R D
Fig. 1. (a) 16-somite snapping turtle embryo explanted onto agar medium. A
portion of the blastoderm covering the head has been removed. Bar = 10 mm.
(b) Higher magnification photomicrograph of the same embryo to illustrate the
segmental plate and its relationship to the forming somites (top) and tail bud
(bottom). Bar = 0-2 mm.
Correlation coefficients between sets of data, calculations of t and other statistical analyses were performed on a Monroe 1930 preprogrammed statistical
calculator. Values ofP < 001 were considered significant.
RESULTS
Somite formation in the snapping turtle appears much like somite formation
in avian embryos. A 16-somite turtle embryo (Stage 8 + ; Yntema, 1968) is
Fig. 2. (a) Photomicrograph of transverse section through an 18-somite embryo
at the level of a recently formed somite pair. Somite morphology and their relationship to the neural tube and notochord are very similar to that seen in avian embryos.
The notochord is often limited by a layer of eosinophilic material, (b) Parasaggital
section through recently formed somite (on the right) and anterior segmental
plate of a 12-somite embryo. Cells of the next somite to form appear to arrange
themselves radially and separate from the anterior end of the segmental plate.
Bar = 0-1 mm.
Somite formation in the snapping turtle
#*£sJ*4*k"*<
111
118
D. S. PACKARD
I
1
'-
»
> '
V*.
i
-1
• T
Fig. 3. Diagram to illustrate cuts made through all three germ layers of the embryo
to produce two explants each of which contained a segmental plate and nearly
one-half of the adjacent neural tube. Only one of the explants contained notochord.
shown explanted onto agar in Fig. 1 a. The somites form in pairs in an anterior
to posterior sequence from paraxial mesoderm condensed into a segmental
plate (Fig. Ib). A transverse section through a recently formed pair of somites
in an 18-somite embryo is shown in Fig. 2a. Somites cleave from the anterior
end of each segmental plate as shown in Fig. 2b. Frequently the cells of the
segmental plate seemed to be arranged in a double columnar configuration
(not shown).
The initial series of experiments were intended to demonstrate the capacity
of explanted Chelydra embryo tissues to form somites under experimental
conditions. It was possible to investigate simultaneously whether continued
contact with the notochord is essential for segmental plates to form their full
complement of somites. The embryos were cut so as to make two tissue
explants each of which contained a segmental plate, associated endoderm and
ectoderm and all embryonic tissues lateral to the segmental plate. Since the
intent was to assess the ability of the segmental plate present at the time of
surgery to form somites, the node region or tail bud posterior to the chorda
bulb was excluded from the explant to inhibit regression movements and thus
prevent formation of additional segmental plate.
Embryos were explanted onto agar medium and the cuts diagrammed in
Somite formation in the snapping turtle
119
Fig. 4. Photomicrographs of living explants, produced by cuts indicated in Fig. 3,
following approximately 48 h of culture, (a) Explant with neural tube, (b) Explant
with neural tube and notochord. The explants formed similar numbers of somites.
Bar = 0-2 mm.
Fig. 3 were made. The first cut was made perpendicular to the embryonic
axis at the level of the last visible intersomitic furrow. The second cut was
made parallel to the embryonic axis, passing slightly to one side of the notochord and thus nearly bisecting the neural tube and including the notochord
with one of the explanted segmental plates. Finally, cuts were made to exclude
the node region or tail bud posterior to the chorda bulb. Thus, two tissue
explants were made each of which contained a segmental plate. These explants
differed in that only one of them contained notochord and the ventral part
of the neural tube. The anterior portion of the embryo was pulled to the side
of the dish. The excised node region or tail bud was discarded. While the
explants appeared to complete somite formation within 20 to 30 h, the cultures
were maintained for three days. In 10 of the 19 experiments, the notochord
was in the right explant. The notochord was in the left explant in the remaining
9 experiments. The embryos ranged in developmental age from 6 to 19 pairs
of somites. In all but two cases, both explants made somites. In those two
120
D. S. P A C K A R D
Fig. 5. Photomicrograph of transverse section through explant, similar to the
explant shown in Fig. 4 a, that contained a portion of the neural tube but no
notochord. The somites formed in these explants (arrow) had a typical morphology.
Bar = 01 mm.
cases the explants that did not form somites appeared to be partially disorganized and so the cases were not included in subsequent analyses. In the
remaining 17 experiments, the explants containing notochord made a mean of
6-5+10 somites while the explants that did not contain notochord made a mean
of 6-2+ 1-0 somites. The two results were not significantly different. Further
analysis of the results from paired explants (dependent /-test) also showed
no significant difference. The explants thus readily formed somites in culture
and the number of somites formed was not influenced significantly by the
presence or absence of the notochord. A typical pair of explants from these
experiments is shown in Fig. 4. The somites appeared much as they did in
intact embryos. Histological analysis of such explants revealed typical somite
morphology even in the absence of the notochord (Fig. 5).
A series of experiments similar to the preceding was carried out to determine whether the continued presence of the neural tube was required for the
segmental plate to form the maximum number of somites. Embryos were
placed on the agar substrate and the cuts diagrammed in Fig. 6 were made.
These cuts were identical to those made in the previous experiments with the
exception that the cut parallel to the embryonic axis at the level of the segmental
plate was made to pass immediately to the left or the right of the neural tube.
The neural tube and notochord remained with only one of the explants; the
Somite formation in the snapping turtle
*&;
3
121
... *
^
Fig. 6. Diagram to illustrate cuts made to produce two explants each of which
contained a segmental plate but only one contained the adjacent neural tube
and notochord.
other explant contained a segmental plate but no axial tissue. A total of 18
such experiments were performed. The embryos ranged in stage of development
from 8 to 16 pairs of somites. The nine explants that contained neural tube
and notochord formed a mean of 7-3 ±1-7 somites, while the nine explants
without neural tube and notochord formed 6-5 ±1-3 somites. These means
were not significantly different. Thus Chelydra segmental plates were able to
form a full complement of somites without further contact with either the
neural tube or the notochord. A typical pair of explants from this experiment
is shown in Fig. 7. The appearance of the somites was similar to those formed
in intact embryos. However, upon histological analysis of explants without axial
structures, the somites were found to be smaller with a lower epithelium (Fig. 8).
It is important to determine how many somites the turtle segmental plate will
form. This can be determined by counting the number of somites such segmental
plates form in culture. Since in the previous experiments no significant differences
were found in the number of somites formed by the variously treated segmental plates, an estimate of the number of somites that a Chelydra segmental
plate will form may be obtained by combining all of the previous data. When
this is done an estimate of 6-6 ±1-2 somites per segmental plate is obtained.
122
D. S. P A C K A R D
Fig. 7. Photomicrographs of living explants, produced by cuts indicated in Fig. 6,
following approximately 48 h of culture, (a) Explant without neural tube and
notochord. (b) Explant with neural tube and notochord. The explants formed
similar numbers of somites. Bar = 0-2 mm.
Since the segmental plates used in the above experiments varied in length,
the possibility that the number of somites formed varied with segmental plate
length had to be considered. Prior to operating on any embryos, the number
of somite pairs they possessed was noted and the length of their segmental
plates was measured. These data are summarized in Fig. 9. It can be seen that
while there was considerable variation in the length of Chelydra segmental
plates, there appeared to be a tendency towards shorter segmental plates in
older embryos. Did shorter segmental plates form fewer somites? Since the
length of each segmental plate had been measured prior to the operations,
these data were compared with the number of somites that each segmental
plate ultimately formed. As shown in Fig. 11, there was no evidence of a
correlation between segmental plate length and the number of somites that
the segmental plate formed (r = 0-135, n = 72). Rather, the number of somites
formed by segmental plates showed only a slight variation over a wide range
of segmental plate lengths.
Before the estimate of the number of somites formed by the Chelydra
Somite formation in the snapping turtle
123
Fig. 8. Photomicrographs of transverse sections through explants similar to those
shown in Fig. 7. (a) Explant without neural tube and notochord contained easily
discernible somites (arrow) but with lower epithelial walls, (b) Explants with
neural tube and notochord contained somites (arrow) with typical morphology.
Size differential between somites shown is due to the more anterior location of
somite in (b). Bar = 01 mm.
segmental plate was accepted, one additional point was investigated. When
somites were counted in the cultured explants it occasionally appeared that
the most posterior portion of the segmental plate had not segmented even
after three days in culture (Fig. 7). This raised the possibility that due to the
conditions of culture, explanted segmental plates might form fewer somites
than they would have if left in the donor embryo. Since it was observed that
most explants had completed somite formation within 30 h of culture and it
was known that cultured intact embryos continued to form somites for at
least 70 h, it was decided to examine the number of somites that segmental
plates would form in cultured, intact embryos. Again, it was necessary to
distinguish somites derived from segmental plates present at the time of initial
cultivation from those derived from segmental plate that formed subsequently.
A number of marking techniques were tried with unsatisfactory results. Ultimately, the node region or tail bud was removed from the embryo, much as
in the previous experiments, to stop segmental plate formation. The region
that was removed is shown in Fig. 11. If the operation prevented further
segmental plate formation as it did in the explant experiments, then any
somites formed after the beginning of culture must be derived from segmental
plate present at the time of the operation. Seven cultured embryos were
124
D. S. PACKARD
l'4r
1-2 -
10
•a
0-8
o
•S 0-6
0-4
0-2
10
15
Number of donor somites
20
25
Fig. 9. Summary of segmental plate lengths measured in embryos of various ages
as indicated by number of somite pairs. Segmental plates varied significantly in
length from embryo to embryo with a tendency for older embryos to have shorter
segmental plates.
operated on in this way. Four additional embryos were cultured as controls.
A typical result is shown in Fig. 12. The last few somites to form were noticeably
narrower. Posterior to the last somite pair no further somite mesoderm was
visible, although the neural tube and notochord appeared to extend for a few
tenths of a millimetre further. Coronal sections of a similar embryo are shown
in Fig. 13. The abrupt halt in somite formation is seen in Fig. 13 a. Fig. 13 b is
a photomicrograph of a section taken a few sections more ventrally in the
same embryo. It shows that the neural tube and notochord usually extended
beyond the last somite pair. The fact that the axial structures appear in crosssection indicates that they have bent ventrally posterior to the last somite
pair. The results are summarized in Fig. 14. The control embryos continued
to make somites throughout the 70 h culture period, making an average of
125
Somite formation in the snapping turtle
1-2
10
E
•E
0-8
1 0-6
J3
04
0-2
5
10
15
20
Number of somites formed by explant
25
Fig. 10. Summary of comparisons between length of segmental plate at time of
explantation and the number of somites ultimately formed by the segmental plate.
The number of somites formed showed little variation over a wide range of segmental plate lengths.
120±0-8 somites. The operated embryos made somites at a similar rate for
the first 43-44 h. After that only one additional somite was counted. It appeared
that removing the node region or tail bud was effective in preventing further
segmental plate formation. The operated embryos formed an average of
6-1 ±1-8 somites during the culture period. This result constituted another
estimate of the number of prospective somites that the Chelydra segmental
plate represents and, as such, it was not significantly different from the estimate
derived from the preceding tissue explant experiments.
DISCUSSION
This report describes a simple and reliable method for the in vitro cultivation
of embryos of the snapping turtle. The method saves turtle eggs for experimental
use, since it utilizes a nutritive agar substrate and a fluid medium based upon
the supernatant from centrifuged, unincubated chicken eggs. Somite-stage
turtle embryos and tissue explants derived from them may be cultivated for
9
EMB
59
126
D. S. P A C K A R D
Figure 11
Figure 12
Fig. 11. Diagram of cuts made to remove node region from cultured embryos. In
older embryos, the tail bud was cut off at the level of the forming notochord.
Fig. 12. Photomicrograph of embryo cultured for 72 h after removal of the tail
bud. Somite formation has stopped abruptly with a slight decrease in somite
width. Neural tube and notochord have lengthened posterior to the level of the
cut. Bar = 0-2 mm.
at least three days at 30 °C. Presumably, the method could be modified for
other reptilian species and for longer periods of culture.
The explant experiments demonstrated that the ability of turtle segmental
plate mesoderm to form somites in culture was not dependent on further
contact with the adjacent neural tube or notochord. This result was identical
with that previously obtained in similar experiments using embryos of the
chicken (Packard & Jacobson, 1976) or the Japanese quail (Packard, 1980).
The segmental plate of the snapping turtle embryo formed 6-6 ± 1-2 somites.
This number was significantly less than the number of somites previously found
to be formed by segmental plates of the chicken (11 -9 ± 1 • 1; Packard & Jacobson,
1976) or of the Japanese quail (11-3 + 2-9; Packard, 1980). The number of
somites formed by snapping turtle segmental plates in culture showed little
127
Somite formation in the snapping turtle
Fig. 13. Photomicrographs of coronal sections through an embryo similar to that
shown in Fig. 12. (a) Somite pairs stopped abruptly at the level of the cut. Neural
tube and notochord appeared in transverse profile. (Jb) Section from same embryo
taken more ventrally to illustrate that lengthened neural tube and notochord
turned ventrally and were not accompanied by somitic mesoderm. Bar = 0-2 mm.
variation over a wide range of segmental plate lengths. Thus, despite the fact
that snapping turtle segmental plates formed significantly fewer somites in
culture than chicken or quail segmental plates, they were similar to the avian
segmental plates in that they tended to form the same number of somites
regardless of variations in the length of the unsegmented somite mesoderm.
Do these results make a case for a pattern-regulating mechanism that adjusts
the somite pattern to the available length of segmental plate in embryos of
the turtle, chick and quail? An obvious problem with such a suggestion is
that the segmental plates of birds and the snapping turtle are not static entities;
as mesoderm condenses to form new segmental plate posteriorly, segmental
plate is being converted into somites anteriorly. Somites in these animals are
not specified in groups of 7 or 12. Rather, they resemble amphibian embryos
in that somite specification appears to be a continuous process (Cooke, 1975).
A pattern-regulating mechanism would have to be continuously adjusting to
segmental plate length to specify the length of the next somite to form. However, Packard (1978) showed that if a chick segmental plate explant was cut
transversely into two pieces, each piece formed the number of somites it
would have formed if the cut had not been made. One interpretation of this
9-2
128
D. S. PACKARD
30
r
20 -
30
40
Hours of culture
Fig. 14. Summary of results of tail bud removals. Number of somite pairs were
counted at various intervals in cultured control embryos (open circles) and
embryos with node region or tail bud removed (closed circles). Control embryos
continued to form somites for at least 70 h. Operated embryos stopped forming
somites after 40 to 50 h. The number of somites formed by operated embryos from
time of operation (6-1 ±1-8) was taken as the number of somites formed by the
segmental plate present at that time.
finding was that a pattern of prospective somites existed along the length of
the segmental plate. Furthermore, since the experiment involved culturing
short pieces of segmental plate, one might have expected that any patternregulating mechanism would adjust the number of somites formed by the
pieces. Such regulation did not occur. I suggested elsewhere (Packard, 1980)
that segmental plates would form a consistent number of somites if the rate
at which prospective somites are specified, presumably in the node region, is
similar to the rate at which somites segment. A pattern of prospective somites
of constant element number would then be present at any moment in the
segmental plate despite variations in its length.
Recently Meier (1979) has shown that the chick segmental plate does exhibit
a metameric pattern consisting of paired somitomeres. It is quite possible
that the prospective somites demonstrated in the above experiments may be
physically manifested by structures similar to chick somitomeres. These
morphological and experimental results, taken together, support the notion
that the mesodermal cells composing the segmental plates of embryos of the
Somite formation in the snapping turtle
129
chicken, Japanese quail and the snapping turtle are committed to form a
certain number of somites. It is likely that the mechanism for forming the
metameric pattern in these animals initially functions at an earlier stage in the
developmental history of the somitic mesoderm than that represented by the
segmental plate.
The author is grateful to Professors D. R. Robertson and C. L. Yntema and to Syamala
Murti for many helpful discussions. I wish to thank Marisa Villani and Nancy Steinberger
for their patient technical assistance. The drawings were expertly prepared by Mr David
Factor. This study was supported by Grant HD 03484 (C.L.Y.) and in part by Grant
HD 13396 (D.S.P.) from the U.S. National Institutes of Health.
REFERENCES
BELLAIRS, R. (1963). The development of somites in the chick embryo. /. Embryol. exp.
Morph. 11, 617-714.
BRUSTIS, J.-J. & GIPOULOUX, J.-D. (1973). Potentialites d'organisation et de differenciation
du mesoderme somitique non segmente associe aux tissus axiaux (corde dorsale et tube
nerveux) chez les amphibiens anoures. C. r. hebd. Seanc. Acad. Sci. Paris, 276,
85-88.
BUTROS, J. (1962). Studies on the inductive action of the early chick axis on isolated postnodal fragments. /. exp. Zool. 149, 1-20.
BUTROS, J. (1967). Limited axial structures in nodeless chick blastoderms. /. Embryol. exp.
Morph. 17, 119-130.
CHRIST, B., JACOB, H. J. & JACOB, M. (1972). Experimented untersuchungen zur somitenentstehung beim huhnerembryo. Z. Anat. EntwGesch. 138, 82-97.
CHRIST, B., JACOB, H. J. & JACOB, M. (1974). Die somitogenese beim huhnerembryo. Zur
determination der segmentierungsrichtung. Verh. anat. Ges., Jena 68, 573-579.
COOKE, J. (1975). Control of somite number during development of a vertebrate, Xenopus
laevis. Nature, Lond. 254, 196-199.
FLINT, O. P., EDE, D. A., WILBY, O. K. & PROCTOR (1978). Control of somite number in
normal and amputated mutant mouse embryos: an experimental and a theoretical analysis.
/. Embryol. exp. Morph. 45, 189-202.
FRASER, R. C. (1960). Somite genesis in the chick. III. The role of induction. J. exp. Zool.
145, 151-167.
HORNBRUCH, A., SUMMERBELL, D. & WOLPERT, L. (1979). Somite formation in the early
chick embryo following grafts of Hensen's node. /. Embryol. exp. Morph. 51, 51-62.
HOWARD, E. (1953). Some effects of sodium chloride concentration on the development of
early chick blastoderms in culture. /. cell. comp. Physiol. 41, 237.
LANOT, R. (1971). La formation des somites chez l'embryon d'oiseau: etude experimentale.
/. Embryol. exp. Morph. 26, 1-20.
MEIER, S. (1979). Development of the chick embryo mesoblast. Formation of the embryonic
axis and establishment of the metameric pattern. Devi Biol. 73, 24-45.
MENKES, B., MICLEA, C, ELIAS, S. & DELEANU, M. (1961). Researches on the formation
of axial organs. I. Study on the differentiation of somites. Acad. RPR Baza Trimisoara,
Stud. Cere. St. Med. 8, 7-33.
MENKES, B. & SANDOR, S. (1977). Somitogenesis: regulation potencies, sequence, determination and primordial interactions. In Symp. Soc. Devi Biol. 3. Vertebrate Limb and Somite
Morphogenesis (ed. Ede, Hinchliffe and Balls), Cambridge University Press.
NICOLET, G. (1970). Is the presumptive notochord responsible for somite genesis in the
chick? /. Embryol. exp. Morph. 24, 467-478.
NICOLET, G. (1971). Avian gastrulation. Adv. Morphogen. 9, 231-262.
PACKARD, D. S., JR. (1978). Chick somite-determination: the role of factors in young somites
and the segmental plate. /. exp. Zool. 203, 295-306.
130
D. S. PACKARD
D. S., JR. (1980). Somitogenesis in cultured embryos of the Japanese quail,
Coturnix coturnix japonica. Am. J. Anat. (In the press.)
PACKARD, D. S., JR. & JACOBSON, A. G. (1976). The influence of axial structures on chick
somite formation. Devi Biol. 53, 36-48.
PEARSON, M. & ELSDALE, T. (1979). Somitogenesis in amphibian embryos. I. Experimental
evidence for an interaction between two temporal factors in the specification of somite
pattern. /. Embryol. exp. Morph. 51, 17-50.
SANDOR, S. & AMELS, D. (1970). Researches on the formation of the axial organs. VI. The
role of the neural tube in somitogenesis. Revue Roumaine d'' Embryologie et de Cytologie,
Serie Embryologie 7, 49-57.
SANDOR, S. & AMELS, D. (1971). Researches on the formation of the axial organs in the
chick embryo. VII. Revue Roumaine d'Embryologie et de Cytologie, Serie Embryologie
8, 37-41.
YNTEMA, C. L. (1964). Procurement and use of turtle embryos for experimental procedures.
Anat. Rec. 149, 577-586.
YNTEMA, C. L. (1968). A series of stages in the embryonic development of Chelydra serpentina.
J. Morph. 125, 219-252.
YNTEMA, C. L. (1970). Extirpation experiments on embryonic rudiments of the carapace of
Chelydra serpentina. J. Morph. 132, 235-244.
PACKARD,
(Received 29 January 1980, revised 27 March 1980)