The adult abdominal epidermis of the flesh fly, Sarco-phaga agryostoma, arises from small nests of diploid histoblasts which grow out and replace the polytene larval epidermal cells during metamorphosis. Extirpation experiments are used to investigate the roles of interactions with neighbouring nests or larval epidermis in determining the fates of the ventral histoblast nests. The extirpation of both left and right ventral histoblast nests deletes the sternite in the corresponding adult segment, but when only one nest is extirpated the hemisternite on that side of the adult is regenerated by the remaining nest. Regenerated hemisternites are smaller and bear fewer bristles than normal hemisternites. Extirpation of the larval epidermis between the two ventral nests leads to the production of small duplicate hemisternites, which resemble the regenerated hemisternites produced by single ventral nests. These midventral extirpations were also found to delay ventral nest fusion.

The results suggest that each ventral nest has an autonomous capacity to form both hemisternites, but that interactions between the nests after fusion normally limit their development and prevent pattern duplication. The possible nature of such interactions is discussed and related to models of pattern regulation and growth control in imaginal disc development.

Studies of regeneration in the abdominal segments of hemimetabolous insects (Wright & Lawrence, 1981) and the imaginal discs of Drosophila (Haynie & Bryant, 1976) have established that, after wound healing, interactions with their new neighbours induce cells to divide and intercalate intermediate pattern elements. The ability of cells to respond appropriately to positional cues from their neighbours is probably a common feature of insect epidermis and may be important in normal development, as well as in the regeneration of damaged structures.

The adult abdominal epidermis of DÍptera arises from small nests of diploid ‘histoblast’ cells set aside at specific dorsal and ventral locations in each hemisegment. Unlike the imaginal discs, which form the rest of the adult epidermis, these nests do not invaginate but remain an active, cuticle-secreting part of the abdominal epidermis throughout their development. Following pupariation the histoblasts begin to divide rapidly and spread out, progressively replacing the poytene larval cells around them. The nests fuse and eventually form a continuous sheet of imaginal cells, which differentiate to secrete the adult abdominal cuticle (Madhavan & Madhavan, 1980; Bautz, 1971). It is not clear whether the final pattern produced by each nest is determined autonomously, or through interactions with other nests, or is generated in response to positional cues from the larval cells encountered during nest outgrowth.

Single histoblast nests can be ablated by microcautery or by extirpation at appropriate sites in late 3rd instar larvae and, in some species, the remaining nests appear to regenerate structures normally formed by an ablated nest. For example, in Calliphora microcautery of both the ventral nests of a posterior abdominal segment completely deletes the sternite of that segment, but when one ventral nest is cauterized the corresponding hemisternite is regenerated by the remaining nest (Emmert, 1972). It is possible that regeneration results from interactions between histoblasts from the regenerating nest and larval cells or dorsal histoblasts in the contralateral hemisegment. An alternative explanation is that each ventral nest has an inherent ability to produce the entire sternite but is normally inhibited from doing so by the presence of the other nest. In this work the regulation of ventral histoblast development is studied in the flesh fly Sarcophaga agryostoma, and the role of instructive interactions with other epidermal cells tested by investigating the effects of extirpations of both histoblasts and larval cells.

Stocks of Sarcophaga agryostoma were maintained at 25 °C. Larvae were fed on a milk and yeast agar (Saunders, 1971) until late 3rd instar when they left the food and were collected as they began to pupariate, that is to lose their mobility and contract to form barrel-shaped sclerotized prepupae. As in other higher DÍptera, metamorphosis takes place inside the old larval cuticle, which is not shed in the pupal moult (Fraenkel & Bhaskaran, 1973). Extirpation of histoblast nests was carried out on semisclerotized pupariating larvae. Square or rectangular fragments of cuticle plus epidermis were removed with razor blade knives. These fragments were dipped in 100 % alcohol for 30 s to kill the attached cells, then washed in insect Ringer for 60 s before being replaced in the wound site. After the fragments had become firmly sealed in place by clotted haemolymph, animals were transferred to a 25°C incubator until their eclosion.

Whole mounts of the adult cuticle were made by dissecting out the abdomen, soaking in 10 % KOH to remove the tissue, washing in water and then mounting in Euparal. For whole mounts of the epidermis of either late 3rd instar larvae or young pupae, the abdomen was cut open along the dorsal midline and the internal organs removed with forceps. Epidermis and cuticle were fixed for 1 h (40 parts 95 % ethanol: 5 parts glacial acetic acid:4 parts 20% formalin), washed in 70% ethanol, heated for 10min in 2M-HC1 at 60°C, and washed in water. The preparations were then stained for 60 s in a 1% solution of basic fuschin in 2 ·5% acetic acid, destained as required in 5 % acetic acid, washed in water, dehydrated through a series of ethanols, cleared in xylene and mounted in DPX.

Normal sternite morphology

The abdominal sternites of adult Sarcophaga are small ventral plates of sclerotized cuticle surrounded by flexible pleura. Each consists of mirror-symmetrical left and right hemisternites, which are fused at the ventral midline. In females, the 4th and 5th sternites are oval in shape and bear evenly spaced bristles on either side of the midventral region (Fig. 1A). Male 4th sternites are more rectangular in shape than their female counterparts (Fig. IB), while male 5th sternites have a complex modified morphology, quite unlike that of the other sternites (Fig. 1C).

Fig. 1.

Ventral morphology in the 4th and 5th abdominal segments of male and female Sarcophaga showing sternites (S4, 55), pleura (P) and edges of tergites (T4, T5). (A) Female 5th and 4th sternites: note the bilaterally symmetrical distribution of bristles. (B) Male 4th sternites. (C) Male 5th sternite showing the modified morphology of the bristlebearing ‘pads’ separated by the deep indentation between the two hemisternites. Bar, 0·5 mm.

Fig. 1.

Ventral morphology in the 4th and 5th abdominal segments of male and female Sarcophaga showing sternites (S4, 55), pleura (P) and edges of tergites (T4, T5). (A) Female 5th and 4th sternites: note the bilaterally symmetrical distribution of bristles. (B) Male 4th sternites. (C) Male 5th sternite showing the modified morphology of the bristlebearing ‘pads’ separated by the deep indentation between the two hemisternites. Bar, 0·5 mm.

The positions of the ventral histoblast nests in the larval segment

Examination of whole mounts of the abdominal epidermis of newly pupariated (white) 3rd instar larvae reveals that each hemisegment contains two dorsal and one ventral nest of diploid histoblasts and a spiracle anlage amongst the polytene larval epidermal cells (Fig. 2A). The ventral nests lie next to distinctive >-shaped muscle insertions, which are visible as indentations in the cuticle surface (Fig. 2B). Extirpation at this site (Fig. 3) completely removed the ventral histoblasts in all 18 animals examined.

Fig. 2.

Mapping the location of the ventral histoblast nests in pupariated larvae. (A) Stained whole mount of the abdominal epidermis of a late 3rd instar larvae showing the relative positions of the two dorsal histoblast nests (ad and pd), the ventral histoblast nest (vh) and the spiracle anlage (sa) in the 4th left hemisegment: note the >-shaped muscle insertion site, marked by the dashed outline, next to the ventral histoblast nest. (B) Pupariated larva, segments 2-4. The intersegmental folds lie in the ventral denticle belts (d). Note the >-shaped indentations (v) marking the sites of the muscle insertions next to the ventral histoblast nests. Bars, 0-5 mm.

Fig. 2.

Mapping the location of the ventral histoblast nests in pupariated larvae. (A) Stained whole mount of the abdominal epidermis of a late 3rd instar larvae showing the relative positions of the two dorsal histoblast nests (ad and pd), the ventral histoblast nest (vh) and the spiracle anlage (sa) in the 4th left hemisegment: note the >-shaped muscle insertion site, marked by the dashed outline, next to the ventral histoblast nest. (B) Pupariated larva, segments 2-4. The intersegmental folds lie in the ventral denticle belts (d). Note the >-shaped indentations (v) marking the sites of the muscle insertions next to the ventral histoblast nests. Bars, 0-5 mm.

Developmental fate of the ventral histoblast nests

Extirpation of both ventral nests in the 4th larval segment completely deleted the 4th sternites of all 31 adults which emerged following this operation. The midventral area was instead occupied by enlarged 4th segment pleura and tergite or the enlarged sternites of the adjacent segments (Fig. 4). The ventral nests therefore normally give rise to the sternites and symmetry further suggests that each ventral nest gives rise to the corresponding hemisternite, as has been concluded in other Diptera (Roseland & Schneiderman, 1979; Bhaskaran, 1973; Emmert, 1972).

Regeneration of the hemisternite following extirpation of one ventral nest

In the non-modified sternites (female 4th/5th and male 4th), the border between the two hemisternites is marked by a gap in the bristle pattern, dividing it into right and left halves. Extirpation of the left ventral nest of the 4th or 5th segment had different effects on the left and right hemisternites. Bristles were either absent or reduced in number on the left side (Fig. 5A-C), while on the right, bristle number was not significantly lower or was even slightly higher than in the hemisternites of unoperated flies (Table 1). As bristle spacing appeared normal, this increase in bristle number indicates an enlargement of the right hemisternite, similar to the enlargement of structures in adjacent segments observed following severe reductions in left hemisternite size (Fig. 5C). Ventral nest extirpations had obvious effects on morphology as well as bristle distribution in the modified male 5th segment sternites. In all 14 males that survived extirpation of the 5th left ventral nest, the right hemisternites were normal, while left hemisternites were reduced in size (Fig. 5D) and in extreme cases the pads lacked bristles.

TABLE 1.

Bristle number on right and left sides of sternites formed after extirpation of one ventral nest

Bristle number on right and left sides of sternites formed after extirpation of one ventral nest
Bristle number on right and left sides of sternites formed after extirpation of one ventral nest

These results show that removal of the left ventral nest only damages the left hemisternite, which is smaller and bears fewer bristles than normal but is, in most cases, not completely deleted. This implies that the right ventral nest is able to compensate for the absence of the left by (partially) regenerating the left hemisternite. The following experiments test whether this could be due to instructive interactions with larval cells in the left hemisegment.

Effects of larval cell extirpation on ventral nest development

Following extirpation of the 5th left ventral nest plus a large square of ventral larval epidermis just to the left of the midline (Fig. 3), all 10 males that emerged bore 5th sternites consisting of a normal right hemisternite fused to a small left hemisternite (Fig. 6A). These results are indistinguishable from those of just extirpating the histoblast nest (Fig. 5D), so it seems that hemisternite regeneration does not depend on contacts between the histoblasts and the ventral larval cells. The next experiment investigates whether interactions with these larval cells are necessary for the normal development of the sternite from both ventral nests.

Fig. 3.

Diagram showing the relative positions of the ventral histoblast nest and the muscle insertions in an abdominal hemisegment shortly after pupariation. The sites at which extirpations were made to remove ventral histoblast nests (vh), or ventral larval epidermis, (mid-V) are marked, isf, intersegmental fold; vm, ventral midline.

Fig. 3.

Diagram showing the relative positions of the ventral histoblast nest and the muscle insertions in an abdominal hemisegment shortly after pupariation. The sites at which extirpations were made to remove ventral histoblast nests (vh), or ventral larval epidermis, (mid-V) are marked, isf, intersegmental fold; vm, ventral midline.

The extirpation of ventral larval epidermis in the 5th segment (Fig. 3) had a striking effect on the development of both right and left ventral nests. In 7 of the 15 males, one additional hemisternite had been formed (Fig. 6B) while in 5 both right and left hemisternites had been duplicated (Fig. 6C). Duplicate hemisternites were generally smaller and bore fewer bristles than normal hemisternites and thus resembled the regenerated hemisternites formed after the extirpation of one ventral nest (Fig. 5D). Examination of the pupal epidermis following ventral extirpations revealed that, although such operations left both ventral nests intact, their rates of outgrowth were reduced. In four of the eight pupae examined at 68h after pupariation, the left ventral nest had grown out less than those of the unoperated segments and, in a further two pupae, the outgrowth of both left and right nests appeared to have been retarded in the 5th segment (Fig. 7). A quantitative assessment of the effect on ventral outgrowth was made by counting the number of larval cells still separating the two ventral nests in each segment. In the 5th segment, this was significantly greater following extirpation than in unoperated control pupae, whereas no such difference between operated and control animals was found in the unoperated 3rd segment (Table 2). The delayed fusion of the 5th ventral nests could not be attributed to any obvious physical barrier to their outgrowth. The epidermis had healed and though haemolymph clots were observed, they were not present in all hemisegments in which the ventral nest outgrowth was retarded (see also Fig. 7B). The outgrowth of other histoblast nests in the vicinity of wounds is, however, often similarly retarded (Smith, 1988).

TABLE 2.

Separation of the ventral nests at 68 h after pupariation following ventral larval cell extirpations in segment 5 (mV5) and in unoperated control animals

Separation of the ventral nests at 68 h after pupariation following ventral larval cell extirpations in segment 5 (mV5) and in unoperated control animals
Separation of the ventral nests at 68 h after pupariation following ventral larval cell extirpations in segment 5 (mV5) and in unoperated control animals

These results show that the extirpation of ventral larval cells, rather than causing a corresponding deletion in the adult pattern, leads to the formation of additional pattern elements which resemble those produced after the removal of one of the ventral nests. This may result directly from the delay in ventral nest fusion which gives both nests an extended period of separate development similar to that experienced by the remaining ventral nest after its partner has been removed.

In Diptera each of the adult abdominal sternites arises from the two ventral histoblast nests in the correspond-ing larval segment. Analysis of the deletions that follow ventral nest ablation in Drosophila (Roseland & Schneiderman, 1979), Musca (Bhaskaran, 1973) and in Calliphora (Emmert, 1972) shows that each ventral nest normally gives rise to the corresponding hemisternite, although, in Calliphora, one nest is able to form the entire sternite if the other is ablated. The present experiments show that removing both the ventral nests in Sarcophaga deletes the corresponding sternite, whereas the removal of a single ventral nest leads to a reduction in the size and the number of bristles in the corresponding hemisternite (Fig. 5) but, in most cases, does not completely delete it. These results indicate that, as in Calliphora, each ventral nest is able to regenerate the hemisternite normally made by its partner.

Unlike cells in the invaginated imaginal discs, histoblasts are always in close contact with the larval cells they eventually replace and there is some evidence that the larval cells could provide positional cues to direct histoblast development. After severe irradiation of Drosophila larvae, the histoblast nests do not develop and the larval cells persist through metamorphosis and eventually secrete a pattern of hairs similar to that seen in the adult (Madhavan & Madhavan, 1984). Thus larval cells can carry the positional information to form an adult pattern. Larval cell instruction of histoblast development could also provide a mechanism for the regeneration of the contralateral hemisternite following the extirpation of one ventral nest, as histoblasts from the remaining nest then spread across the ventral midline and contact larval cells in the contralateral hemisegment (Smith, 1988). However, the extirpation of ventral larval cells in this hemisegment has no effect on hemisternite regeneration when one ventral nest has been removed (Fig. 6A) and, in otherwise intact animals, frequently leads to hemisternite duplication (Fig. 6B,C), rather than the deletions which would be expected if histoblasts developed according to positional cues from the larval epidermis. The conclusion that pattern formation in Sarcophaga histoblasts is independent of positional instruction from the larval cells they encounter during outgrowth is consistent with findings in other Diptera that microcautery, extirpation and rotation of larval epidermis also has no corresponding effects on adult pattern (Roseland & Schneiderman, 1979; Roseland, 1976; Pearson, 1977). The duplications of anterior or posterior tergite that follow the microcautery of larval epidermis close to the dorsal histoblast nests (Roseland & Schneiderman, 1979) are probably associated with damage to histoblasts as well as to larval cells (Smith, 1988). Defects in the pattern of hemisegment fusion in the larva following microcautery or X-irradiation of Drosophila or Calliphora embryos often lead to corresponding defects in adult segmentation (Bownes, 1976; Pearson, 1973), but this may be due to mechanical effects on nest orientation and outgrowth rather than on pattern generation. No such correspondence between the pattern eventually produced by the histoblasts and effects on anteroposterior patterning within larval segments was found in costal mutants of Drosophila (Simpson & Grau, 1987).

One alternative explanation of hemisternite regeneration following the extirpation of the corresponding ventral nest is that the remaining nest eventually forms contacts with dorsal histoblasts in the contralateral hemisegment and this stimulates intercalary regeneration of the hemisternite The hemisternite duplications observed following ventral larval cell extirpations cannot, however, be due to interactions between histoblast nests which do not normally meet as, although these operations lead to a delay in ventral nest outgrowth and fusion (Table 2), they do not prevent the normal pattern of contacts between the nests from ultimately taking place.

The formation of extra hemisternites is, however, consistent with a third hypothesis, namely that the ventral histoblast nests each have an autonomous capacity to generate both right and left hemisternites but are normally prevented from doing so by interactions between the two nests after they fuse. The extirpation of ventral larval epidermis delays ventral nest fusion (Table 2), thus increasing the time available for each nest to develop autonomously. In many cases, this increase could be sufficient to enable either or both ventral nests to produce a complete sternite rather than just the corresponding hemisternite. The effects of extirpation on histoblast nest development in Sarco-phaga are very similar to those of analogous operations on the female genital discs of Musca and Calliphora in both species, the right and left lateral genital discs each give rise to the corresponding half of various internal and external structures in the 8th adult abdominal segment. When one of the discs is removed, its adult derivatives are regenerated by the remaining disc (Emmert, 1972) and single discs also produce a complete set of the structures normally formed by the right and left discs together following in vivo culture (Due-bendorfer, 1971). Thus, like the ventral histoblast nests, these discs have an inherent ability to generate pattern elements normally made by their partners. Such an ability is, however, very unusual: in almost all other cases undamaged imaginal discs show no sign of being able to generate additional structures in response either to culture or to the extirpation of their neighbours (Bryant, 1978).

Fragments of Drosophila wing and leg discs are able to exceed their normal fate when cultured in vivo and generally stop developing once they have generated either the complete pattern normally produced by the whole disc or a duplication of that part normally formed by the fragment in situ. Fragment healing brings together cells from normally nonadjacent regions of the disc, interactions between which are thought to stimulate shortest-route intercalary regeneration to remove the discontinuity thus created (French et al. 1976). A similar process may normally limit pattern generation by histoblast nests (or by the lateral genital discs).

Following the fusion of two ventral nests, each would be able to respond to positional cues from the other and stop developing once a continuous, and therefore stable, pattern had been produced in the midventral region. Such interactions between adjacent nests/discs would prevent gaps being produced in the final adult pattern but could not correct all types of abnormal pattern development. Thus, the final form of extra pattern elements generated prior to a delayed nest fusion might be modified by intercalation (for example, note the abnormal symmetry of the single extra hemisternite in Fig. 6B) but once produced such pattern elements could not be deleted.

In Drosophila, the interactions that regulate pattern development within each disc also seem to control its growth (Bryant & Simpson, 1984). For example, regenerating or duplicating disc fragments cultured in adult females stop growing once pattern continuity has been restored, although all the hormonal and nutritive conditions necessary for growth are still present. No such link between size and pattern has, however, been established in histoblast development. In Sarcophaga, the deletion of the sternite, which follows the extirpation of both the ventral nests, causes compensatory increases in the size of adjacent structures (Figs 4A,B, 5C) and when one nest is removed, the hemisternites normally formed by the remaining nests can be significantly larger than those of controls (Table 1). Growth control therefore cannot be strictly coupled to pattern formation in histoblast development, as even nests with the capacity to regenerate structures may also produce larger versions of those they normally give rise to. Such expansive growth may normally be limited by competition between histoblasts from neighbouring nests to fill the space occupied by the remaining larval cells. Thus, interactions between histoblasts and larval cells, although without direct effects on adult pattern formation, may be necessary to regulate the growth of developing histoblast nests.

Fig. 4.

Ventral abdominal morphology in the adult following extirpation of both ventral nests in the 4th segment. (A) The 4th sternite is deleted and its place taken by pleura. Note also the ventralward expansion of the tergites in the 4th segment. (B) The sternite is deleted and in this case the sternites of the adjacent segments have expanded to take its place.

Fig. 4.

Ventral abdominal morphology in the adult following extirpation of both ventral nests in the 4th segment. (A) The 4th sternite is deleted and its place taken by pleura. Note also the ventralward expansion of the tergites in the 4th segment. (B) The sternite is deleted and in this case the sternites of the adjacent segments have expanded to take its place.

Fig. 5.

The effects of extirpation of the animal's left ventral nest on the corresponding hemisternite (on righthand side in cuticle whole mounts). (A) Male 4th segment. Relatively few bristles have been produced on the extirpated side of the sternite. (B) Female 5th segment. Bristle number is severely reduced on the extirpated side of the sternite. (C) Female 4th segment. The hemisternite has been deleted on the extirpated side. Note the anteriorward expansion of the left hemisternite in the adjacent 5th segment. (D) Male 5th segment. On the extirpated side, the hemisternite is small and bears relatively few bristles on the ‘pad’ while on the control side the hemisternite is normal.

Fig. 5.

The effects of extirpation of the animal's left ventral nest on the corresponding hemisternite (on righthand side in cuticle whole mounts). (A) Male 4th segment. Relatively few bristles have been produced on the extirpated side of the sternite. (B) Female 5th segment. Bristle number is severely reduced on the extirpated side of the sternite. (C) Female 4th segment. The hemisternite has been deleted on the extirpated side. Note the anteriorward expansion of the left hemisternite in the adjacent 5th segment. (D) Male 5th segment. On the extirpated side, the hemisternite is small and bears relatively few bristles on the ‘pad’ while on the control side the hemisternite is normal.

Fig. 6.

The effects of ventral larval cell extirpation on the male 5th sternite. (A) Following extirpation of the left 5th ventral nest and of left hemisegment ventral larval epidermis (see Fig. 3) one normal and one small hemisternite have been produced. (B) Duplication of one of the hemisternites, following midventral extirpation. Note the abnormal bilaterally symmetrical structure of the middle hemisternite pad. (C) Duplication of both hemisternites following extirpation of the left midventral larval epidermis. Like the sternites formed after ventral nest extirpation each ‘sternite’ consists of one large and one small (regenerated) hemisternite.

Fig. 6.

The effects of ventral larval cell extirpation on the male 5th sternite. (A) Following extirpation of the left 5th ventral nest and of left hemisegment ventral larval epidermis (see Fig. 3) one normal and one small hemisternite have been produced. (B) Duplication of one of the hemisternites, following midventral extirpation. Note the abnormal bilaterally symmetrical structure of the middle hemisternite pad. (C) Duplication of both hemisternites following extirpation of the left midventral larval epidermis. Like the sternites formed after ventral nest extirpation each ‘sternite’ consists of one large and one small (regenerated) hemisternite.

Fig. 7.

Stained whole mount of pupal epidermis 68 h after pupariation following the extirpation of left ventral epidermis in the 5th segment. (A) The ventral nests (V4L, V4R, V5L and V5R) are close to fusion but in the 5th segment the histoblasts have grown out less toward the ventral midline than in the 4th segments. Note also the haemolymph clot (arrows) at the healed wound site in the left hemisegment. (B) Enlargement of the 5th left hemisegment (bottom left in A, note the marked haemolymph clot; h), showing the ventral border of the histoblast nest which has not yet reached the haemolymph clots. Bars, A, 0-5mm; B, 0-25mm.

Fig. 7.

Stained whole mount of pupal epidermis 68 h after pupariation following the extirpation of left ventral epidermis in the 5th segment. (A) The ventral nests (V4L, V4R, V5L and V5R) are close to fusion but in the 5th segment the histoblasts have grown out less toward the ventral midline than in the 4th segments. Note also the haemolymph clot (arrows) at the healed wound site in the left hemisegment. (B) Enlargement of the 5th left hemisegment (bottom left in A, note the marked haemolymph clot; h), showing the ventral border of the histoblast nest which has not yet reached the haemolymph clots. Bars, A, 0-5mm; B, 0-25mm.

I am grateful to Vernon French for his advice and helpful comments on the manuscript, David Saunders for the provision of Sarcophaga stocks, Denis Cremer for technical assistance and the SERC for financial support.

Bautz
,
A. N.
(
1971
).
Chronologie de la mise en place de I’hypoderme imaginai de 1’abdomen de Calliphora erythrocephala Meigen (Insecte diptere, Brachycere)
.
Archs. Zool. exo. gén
.
112
,
157
178
.
Bhaskaran
,
G.
(
1973
).
Developmental behavior of the abdominal histoblasts in the housefly
.
Nature, Lond
.
241
,
94
96
.
Bownes
,
M.
(
1976
).
Larval and adult abdominal defects resulting from microcautery of blastoderm staged Drosophila embryos
.
J. exp. Zool
.
195
,
369
392
.
Bryant
,
P. J.
(
1978
).
Pattern formation in imaginai discs
.
In The Genetics and Biology of Drosophila
, vol.
2C
(ed.
M.
Ashburner
&
T.
Wright
), pp.
229
335
.
New York
:
Academic press
.
Bryant
,
P. J.
&
Simpson
,
P.
(
1984
).
Intrinsic and extrinsic control of growth in developing organs
.
Q. Rev. Biol
.
59
,
387
415
.
Duebendorfer
,
A.
(
1971
).
Untersuchung zum Anlageplan und Determinationszustand der weiblichen Genital und Analprimordien von Musca domestica L
.
Wilhelm Roux Arch. EntwMech. Org
.
168
,
142
168
.
Emmert
,
W.
(
1972
).
Entwicklungleistungen abdominaler Imaginalscheiben von Calliphora erythrocephala (Insecta DÍptera). Experimentelle Untersuchungen zur Morphologie des Abdomens
.
Wilhelm Roux’ Arch. EntwMech. Org
.
169
,
87
133
.
Fraenkel
,
G.
&
Bhaskaran
,
G.
(
1973
).
Pupariation and pupation in cyclorraphous flies (DÍptera): terminology and interpretation
.
Ann. Ent. Soc. Amer
.
66
,
418
422
.
French
,
V.
,
Bryant
,
P.
&
Bryant
,
S.
(
1976
).
Pattern regulation in epimorphic fields
.
Science
193
,
969
981
.
Haynie
,
J. L.
&
Bryant
,
P. J.
(
1976
).
Intercalary regeneration in imaginai wing disc of Drosophila melanogaster
.
Nature, Lond
.
259
,
659
662
.
Madhavan
,
M. M.
&
Madhavan
,
K.
(
1980
).
Morphogenesis of the epidermis of the adult abdomen of Drosophila
.
J. Embryol. exp. Morph
.
60
,
1
31
.
Madhavan
,
M. M.
&
Madhavi
,
K.
(
1984
).
Do larval epidermal cells possess the blueprint for adult pattern in Drosophila?
J. Embryol. exp. Morph
.
82
,
1
8
.
Pearson
,
M.
(
1973
).
The relation between larval and adult abnormalities in the abdominal segmentation of Calliphora erythrocephala (DÍptera)
.
J. Em’bryol. exp. Morph
.
32
,
533
555
.
Pearson
,
M.
(
1977
).
Pattern and polarity of sclerites in adult abdominal segments of Calliphora erythrocephala (DÍptera): Anlage rotation experiments
.
J. Embryol. exp. Morph
.
37
,
91
104
.
Roselond
,
C. R.
(
1976
).
Regulation and pattern formation in the abdominal histoblasts of Drosophila melanogaster and Musca domestica
.
Ph.D. Thesis, University of California, Irvine
.
Roselond
,
C. R.
&
Schneiderman
,
H. A.
(
1979
).
Regulation and metamorphosis of the abdominal histoblasts of Drosophila melanogaster
.
Wilhelm Roux’ Arch, devl Biol
.
186
,
235
265
.
Saunders
,
D. S.
(
1971
).
The temperature-compensated photoperiodic clock “programming” development and pupal diapause in the flesh fly Sarcophaga agryostoma
.
J. Insect Physiol
.
17
,
801
812
.
Simpson
,
P.
&
Grau
,
Y.
(
1987
).
The segment-polarity mutant costal-2 in Drosophila. The origin of the imaginai pattern duplications
.
Devi Biol
.
122
,
201
209
.
Smith
,
H. K.
(
1988
).
Development of the adult abdomen in Sarcophaga agryostoma (DÍptera)
.
Ph.D. thesis, University of Edinburgh
.
Wright
,
D.
&
Lawrence
,
P. A.
(
1981
).
Regeneration of the segment boundary in Oncopeltus
.
Devi Biol
.
85
,
317
327
.