Regeneration and grafting experiments were carried out on the prothoracic leg of the cricket Teleogrylhts commodus (Walker) to examine the precision with which surface cuticular structures and internal epidermal derivatives are reformed. By comparing regenerated and grafted limbs with normal limbs it was found that the three-dimensional structure of epidermal derivatives is not restored. This is despite the fact that regenerated and grafted limbs appear similar in their external morphology to normal limbs. The implication of these results are discussed in the context of theories of pattern formation.

Regeneration has interested biologists for many years. Recently developmental biologists have been using regeneration, together with experiments involving grafting as a way of gaining an understanding into the events occurring in development. The result of these studies has been the general concept of positional information (Wolpert, 1969, 1972), that is, the concept that cells have access to information about their position in a developing field and differentiate accordingly.

One model using this concept is the polar co-ordinate or ‘clockface model’ (French, Bryant & Bryant, 1976). In this model, positional information and therefore the resultant pattern is held to be specific in two dimensions (French et al. 1976; Carlson, 1975). Structures having a three-dimensional shape emerge via folding and shaping of a two-dimensional cell sheet, similar to the process occurring in gastrulation.

Insect sensory structures offer an opportunity to study a three-dimensional arrangement in which positional information is held to be specified in two dimensions. One class of insect sensory structures, chordotonal organs, although internal in position, are derived from epidermal cell derivatives which become internalised during development (Moulins, 1976; Wigglesworth, 1953).

In regeneration experiments involving either amputation of a limb or removal of a strip of epidermis, the normal cuticular pattern is restored at subsequent moults (French, 1978). In view of the epidermal origin of chordotonal organs, one might expect that they would also regrow the normal three-dimensional arrangement.

This paper compares the three-dimensional arrangement of the tibial sensory complex (Eibl, 1978) with that of the cuticular surface in both regenerated and grafted limbs in the Australian field cricket Teleogryllus commodus (Walker).

Animals

Crickets Teleogryllus commodus (Walker) were collected in the field and were kept in culture in the laboratory. Cultures were supplied with water and commercial rat food, and kept under a 12 h day/night cycle at 25 °C. Under these conditions post-embryonic development took approximately 7–8 weeks.

Experimental animals were placed in individual containers (10×6×6 cm), supplied with food and water. All operated crickets were examined periodically. Those which showed graft rejection were discarded (Table 1).

Table 1.

Details of experiments

Details of experiments
Details of experiments

Grafts

Two types of grafting experiments, using nymphal crickets, were carried out by transplanting a prothoracic coxa (Table 1). These were:

(a) Non-congruent graft-host junction

Either a left or right (L or R) coxa was grafted onto a R or L host coxa respectively in order to reverse the anterior–posterior (A–P) polarity.

(b) Congruent graft-host junction

As a control, a donor coxa was removed and transplanted onto a like-handed prothoracic stump. In some cases the same animal was used as both donor and host. In such cases all axes were congruent.

In both types of grafts both host and donor nymphs were matched by means of a staging system developed by Ball & Young (1974). All grafts were performed on cold-immobilized animals. To ensure grafts remained in position until the haemolymph hardened, a strip of insect wax (25 % violin resin, 75 % bees wax) was applied to the graft-host junction using a micro-soldering iron. Crickets were allowed to recover at room temperature.

Regeneration

Either a R or L prothoracic leg was cut off at mid-coxal level and allowed to regenerate. In all cases the contralateral leg was allowed to grow normally.

Criteria for examination of tibial sensory complex

In order to facilitate comparisons between the two types of graft, leg regeneration and the normal condition, one must have some consistent criteria for comparing animals after these different treatments. Otherwise, differences could be due to different stages of development. In order to overcome this problem, only those duplicated and regenerated limbs showing full segmentation and having a tibial length 70 % or greater compared to the contralateral limb were included in the examination. Contralateral limbs were normal.

Histology

Immediately after the last moult, the tibial segment was removed and placed in fixative (alcoholic Bouin) for 2 h. Tibial segments were then dehydrated in graded alcohols, passed through xylene and vacuum embedded in paraffin wax (m.p. 56 °C). Sections were stained with Ehrlich’s haematoxylin and Eosin or Baker’s modification of Masson’s trichrome stain (Pantin, 1946). Some material was embedded in glycol methacrylate (Polysciences) and stained with Lee’s methylene blue-basic fuchsin (Bennet, Wynick, Lee & McNeil, 1976). Sections were photographed with a Leitz Orthomat microscope/camera system using Kodak Pan F or Ilford FP4 film.

External cuticular pattern

External cuticular anatomy of both normal and regenerated and graft limbs was studied by scanning electron microscopy. Material was either critical-point dried or air dried, gold coated and examined using a Cambridge Stereoscan electron microscope. A total of 33 normal limbs from various instars was examined, while 8 regenerated and transplanted limbs were examined also. Some limbs were photographed with a Zeiss-Tessovar microscope/camera system.

(1) General morphology of experimental limbs

Congruent grafts and regenerates produced single outgrowths. In contrast, the non-congruent grafts produced multiple outgrowths. Of the 100 non-congruent grafts attempted 17 were successful in producing multiple outgrowths (Table 1). Of these 17, eight produced triple outgrowths consisting of two laterals from the sites of axial incompatibility and a regenerate lying between the two as expected (Bohn, 1965). The remaining nine produced only a single fully segmented lateral; the other lateral was either reduced to a small unsegmented bud or was absent altogether.

In all cases the outgrowths showed segmentation developing in a proximodistal order, i.e. tarsal segments were last to be differentiated.

(2) Comparing the experimental series

In order to make comparisons of the end point of development two criteria were chosen (Materials/Methods). Though arbitrary, such criteria are useful because a normal limb at a similar morphological stage contains the full complement of sensilla in the subgenual organ and the tympanal organ is almost complete (Ball & Young, 1974). The campaniform sensilla pattern is also nearly complete.

(3) Tympanal organ

The structure of the normal tympanal organ has been described earlier, so has its post-embryonic development (Ball & Young, 1974; Young & Ball, 1974). It consists of 70 sensilla (scolopidia) which can be divided into five distinct anatomical groups (Fig. 1). These sensilla, like the others in the tibial complex, must arise de novo in regenerates and grafts, since the distal segments of the leg are removed during the initial operation.

Fig. 1.

Line drawing of tibial sensory complex in a normal limb. In (a) the position of the complex is shown in relation to the tibia (Ti) and femur (Fe) while in (b) the sensory complex is shown in detail. The tibia has been drawn as if the cuticle (C) and epidermis (E) on its anterior face has been removed exposing the internal structures. Muscle has been eliminated for clarity. The right-hand side of the figure corresponds to the distal portion of tibia. Note tympanal organ (TO) in association with expansion of anterior (Atr) and posterior trachea (Ptr). Its two groups of neurones a proximal (PN) and distal (DN) are indicated. The proximal neurones are in close association with the neurones of subgenual organ (SON). The subgenual organ scolopoles (SoSc) are also shown. The campaniform sensilla (CS) lie on the dorsal cuticle. Modified from Young & Ball (1974). PT, posterior tympanum ; SN, subgenual nerve; TN, tympanal nerve. Calibration = 0·1 mm.

Fig. 1.

Line drawing of tibial sensory complex in a normal limb. In (a) the position of the complex is shown in relation to the tibia (Ti) and femur (Fe) while in (b) the sensory complex is shown in detail. The tibia has been drawn as if the cuticle (C) and epidermis (E) on its anterior face has been removed exposing the internal structures. Muscle has been eliminated for clarity. The right-hand side of the figure corresponds to the distal portion of tibia. Note tympanal organ (TO) in association with expansion of anterior (Atr) and posterior trachea (Ptr). Its two groups of neurones a proximal (PN) and distal (DN) are indicated. The proximal neurones are in close association with the neurones of subgenual organ (SON). The subgenual organ scolopoles (SoSc) are also shown. The campaniform sensilla (CS) lie on the dorsal cuticle. Modified from Young & Ball (1974). PT, posterior tympanum ; SN, subgenual nerve; TN, tympanal nerve. Calibration = 0·1 mm.

In the course of normal development a limb of 70 % of adult tibial length would contain between 50 and 60 of the total complement of sensilla. All the five anatomical groups would be represented. In contrast in only five experimental cases (four regenerates, one congruent graft) could a structure homologous with the tympanal organ be identified. The homology was based on similar proximo-distal and circumferential positions within the tibia, being distal to and separate from the subgenual organ. In four out of the five cases this structure contained five neurones or less, with no associated scolopales or attachment cells. These neurones were in close association with the epidermis (Fig. 2).

Fig. 2.

Line drawing of tympanal organ in experimental cases (a) regenerate (Instar 1) and (b) congruent graft (Instar 1). Note neurones (N) contained in an envelope located between epidermis (E) and trachea (Tr). Scolopales and attachment cells are absent. Top margin of drawing corresponds to dorsal side of tibia, the left hand corresponds to anterior side of tibia. Drawings traced from 10 μm transverse wax sections stained with haematoxylin and eosin. The cuticle has been lost in sectioning. Calibration = 25 μm.

Fig. 2.

Line drawing of tympanal organ in experimental cases (a) regenerate (Instar 1) and (b) congruent graft (Instar 1). Note neurones (N) contained in an envelope located between epidermis (E) and trachea (Tr). Scolopales and attachment cells are absent. Top margin of drawing corresponds to dorsal side of tibia, the left hand corresponds to anterior side of tibia. Drawings traced from 10 μm transverse wax sections stained with haematoxylin and eosin. The cuticle has been lost in sectioning. Calibration = 25 μm.

In the remaining case, using a newly emerged instar 1 (regeneration), the regenerate tympanal organ in the adult was considerably more advanced, consisting of a dorsal mass of attachment cells, ten neurones and associated scolopales (Fig. 3). These sensilla however could not be allocated to any of the groups described by Young & Ball (1974). No similar structure either of only neurones or sensilla was seen in non-congruent grafts. Thus a regenerate limb and congruent graft showed the rudiments of a tympanal organ internally while a non-congruent graft did not, even though both were similar in external morphology.

Fig. 3.

Line drawing of tympanal organ in (a) most advanced experimental case (regenerate, Instar 1) and for comparison in a normal limb at the same level. Note both cases contain neurones (N), scolopales (SC) and attachment cells (ac) while each differ in mode of attachment to epidermis (E) and number of sensilla. Top of drawing corresponds to dorsal portion tibia, the right side the anterior side of the tibia. Drawings traced from 5 μm transverse glycol methacrylate sections stained with Lee’s methylene blue/basic Fuschin. Atr, anterior trachea; C, cuticle; Ptr, posterior trachea. Calibration = 50 μm.

Fig. 3.

Line drawing of tympanal organ in (a) most advanced experimental case (regenerate, Instar 1) and for comparison in a normal limb at the same level. Note both cases contain neurones (N), scolopales (SC) and attachment cells (ac) while each differ in mode of attachment to epidermis (E) and number of sensilla. Top of drawing corresponds to dorsal portion tibia, the right side the anterior side of the tibia. Drawings traced from 5 μm transverse glycol methacrylate sections stained with Lee’s methylene blue/basic Fuschin. Atr, anterior trachea; C, cuticle; Ptr, posterior trachea. Calibration = 50 μm.

(4) Subgenual organ

The subgenual organ is a fan-like chordotonal organ located in the dorsal haemolymph space. It lies midway between the campaniform sensilla and the tympanal organ (Figs. 1, 4 b). Its normal adult structure has been described in detail elsewhere (Young & Ball, 1974; Eibl, 1978). The organ is supplied by two nerves, the tympanal nerve which also supplies the tympanal organ sensilla and the subgenual nerve which also supplies the campaniform sensilla (Young & Ball, 1974). It contains about 20 sensilla, the neurones of which are divided into two distinct groups – a larger anterior group containing about 15 neurones and smaller posterior group containing about five neurones (Fig. 4). Unlike the neurones the scolopales are not divided into two distinct groups and it is unclear which scolopales correspond to the posterior neurones. In both graft experiments and regenerates, scolopidia are differentiated in two distinct groups: a posterior and anterior group (Fig. 4). Posterior and anterior are taken with respect to stump axes. However, the precise ordered structure of the normal adult subgenual organ was never re-established even using instar 1 animals for experiments (Fig. 4). Cell counts from three experiments (two non-congruent laterals, one regenerate) show that the posterior group contains between five and nine scolopodia, while the anterior group has between three and seven scolopodia. Each group receives separate innervation by nerves considered to be equivalent to the subgenual and tympanal nerves respectively. Grafts using older instars indicate that these anterior and posterior groups of scolopodia originated from different regions of epidermis.

Fig. 4.

Line drawing of tibial sensory complex in (a) regenerate Instar (1) and for comparison a normal limb (b). In each the tibia has been drawn as if the cuticle (C) and epidermis (E) on the dorsal surface has been removed. The underlying trachea and muscle has been omitted for clarity. Right-hand side corresponds to posterior side of tibia, the top of the figure corresponds to the distal portion of tibia. In (b) the distal portion of tympanal organ has also been omitted. Note presence of subgenual organ (SO) in both normal and regenerated limbs while tympanal organ is only present in regenerated limbs as a small group of neurones (TON), unlike the normal limb in which distinct proximal (PN) and distal neurones (DN) are present. Drawings reconstructed from 10 μm wax sections cut in a horizontal longitudinal plane. Sections stained with haematoxylin and eosin, (a) Is a composite reconstruction from five regenerated limbs. AtN, anterior neurones; N, neurones; PtN, posterior neurones; Sc, scolopales; SN, subgenual nerve; TN, tympanal nerve. Calibration = 0·1 mm.

Fig. 4.

Line drawing of tibial sensory complex in (a) regenerate Instar (1) and for comparison a normal limb (b). In each the tibia has been drawn as if the cuticle (C) and epidermis (E) on the dorsal surface has been removed. The underlying trachea and muscle has been omitted for clarity. Right-hand side corresponds to posterior side of tibia, the top of the figure corresponds to the distal portion of tibia. In (b) the distal portion of tympanal organ has also been omitted. Note presence of subgenual organ (SO) in both normal and regenerated limbs while tympanal organ is only present in regenerated limbs as a small group of neurones (TON), unlike the normal limb in which distinct proximal (PN) and distal neurones (DN) are present. Drawings reconstructed from 10 μm wax sections cut in a horizontal longitudinal plane. Sections stained with haematoxylin and eosin, (a) Is a composite reconstruction from five regenerated limbs. AtN, anterior neurones; N, neurones; PtN, posterior neurones; Sc, scolopales; SN, subgenual nerve; TN, tympanal nerve. Calibration = 0·1 mm.

(5) Tympana

A normal limb has two tympana, a larger posterior and smaller anterior one (Fig. 5). Tympana were formed in all experimental series. In the majority of experiments only the posterior tympanum differentiated and an area of lighter cuticle was formed at the site where the anterior tympanum would normally be found. The posterior tympanum approached the normal condition in terms of size and shape (Fig. 5), whereas the anterior tympanum, if present, was far more variable (Figs. 5, 6). To form a tympanum under regeneration conditions nymphs instar 4 or earlier were required, while graft experiments had to be done at earlier instars to achieve a similar result.

Fig. 5.

Photomicrographs ot normal and regenerated tibia (Instar 1) viewed from (a) posterior surface and (b) anterior surface showing larger posterior and smaller anterior tympana, respectively. These stand out against the background of much darker leg cuticle. In each case the normal limb is located on the right. In (a) note the slightly smaller size at regenerated posterior tympana. In (b) the regenerated anterior tympana is present only as an area of slightly lighter cuticle (arrowed). Calibration = 0.5 mm.

Fig. 5.

Photomicrographs ot normal and regenerated tibia (Instar 1) viewed from (a) posterior surface and (b) anterior surface showing larger posterior and smaller anterior tympana, respectively. These stand out against the background of much darker leg cuticle. In each case the normal limb is located on the right. In (a) note the slightly smaller size at regenerated posterior tympana. In (b) the regenerated anterior tympana is present only as an area of slightly lighter cuticle (arrowed). Calibration = 0.5 mm.

Fig. 6.

Photomicrographs of experimental tibia to show tympana. Same orientation as in previous figure, (a) posterior tympana; regenerate (Instar 3). Note that it runs to a point distally, around which the leg cuticle is devoid of hairs; (b) anterior tympana congruent graft (Instar 1). Note its irregular shape compared to normal anterior tympana in previous figure. Calibration = 0·1 mm

Fig. 6.

Photomicrographs of experimental tibia to show tympana. Same orientation as in previous figure, (a) posterior tympana; regenerate (Instar 3). Note that it runs to a point distally, around which the leg cuticle is devoid of hairs; (b) anterior tympana congruent graft (Instar 1). Note its irregular shape compared to normal anterior tympana in previous figure. Calibration = 0·1 mm

Experiments carried out using older instars consistently lacked an anterior tympanum, and the posterior tympanum was reduced in size. Its shape, unlike the normal oval configuration, ran to a point distally. The cuticle around the distal margin of the posterior tympanum, unlike the rest of the leg cuticle was devoid of hairs (Fig. 6). In all cases the circumferential and proximo-distal positions conformed to the host axes.

(6) Trachea

In contrast to the consistent pattern of tracheal growth seen in normal limbs (Ball & Young, 1974) tracheal growth was far more variable and never conformed to the adult pattern (Fig. 7).

Fig. 7.

Schematic line drawing of successive transverse sections through region of tibia normally occupied by tibial sensory complex showing different patterns of tracheal growth in normal limbs and experimental limbs. The outer circle represents the outline of the limb cuticle, while the inner circles represent tracheal profiles. Arrows indicate position of tympana. The right-hand side of each figure corresponds to posterior side of limb. The top of each figure corresponds to the dorsal side of tibia, (a) normal limb; (b) non-congruent graft-anterior lateral (Instar 1); (c) congruent graft (Instar 2); (d) regenerate (Instar 3). The normal tracheal pattern is consistent from specimen to specimen (compare with Fig. 1). In contrast the experimental limbs show considerable variability in tracheal pattern. Calibration = 0·3 mm.

Fig. 7.

Schematic line drawing of successive transverse sections through region of tibia normally occupied by tibial sensory complex showing different patterns of tracheal growth in normal limbs and experimental limbs. The outer circle represents the outline of the limb cuticle, while the inner circles represent tracheal profiles. Arrows indicate position of tympana. The right-hand side of each figure corresponds to posterior side of limb. The top of each figure corresponds to the dorsal side of tibia, (a) normal limb; (b) non-congruent graft-anterior lateral (Instar 1); (c) congruent graft (Instar 2); (d) regenerate (Instar 3). The normal tracheal pattern is consistent from specimen to specimen (compare with Fig. 1). In contrast the experimental limbs show considerable variability in tracheal pattern. Calibration = 0·3 mm.

In experimental limbs the only consistent pattern was the appearance of two main trachea. Their size and shape, unlike the normal situation, was variable between experimental limbs. In addition, a number of small trachea was present.

Tympana were associated with internal expansion of the tracheal system, although no causal relationship can be inferred from this study.

(7) Campaniform sensilla

A group of 13–15 campaniform sensilla is located in a mid-dorsal position on the tibia in all three pairs of legs. In the prothoracic leg, they lie in the proximal part of the tibia 0·3–0·4 mm from the femoro-tibial joint (Fig. 1). The sensilla are normally found in a distinct pattern (Fig. 8d). The normal development and the pattern arising after regeneration was examined in order to compare cuticular differentiation. Within the normal pattern, three groups of sensilla can be designated; a proximal group of the two largest sensilla and a distal group of three slightly smaller sensilla arranged in the form of a triangle, the apex of which points towards the femur-tibia joint. Located between these two groups is an intermediate group of eight to ten sensilla arranged in an arc which is open distally (Fig. 8). Within this intermediate group the size of the sensilla varies in a systematic way: the largest are located on the posterior side of the arc and their size decreases toward the anterior side (Fig. 8b). Hence the arrangement of campaniform sensilla in the adult has both proximo-distal polarity designated by the proximal and distal groups, and circumferential polarity (anterior-posterior) due to size differences in the intermediate sensilla. Each sensillum has its own intrinsic polarity due to the presence of a distally pointing protuberance (Fig. 8a). In normal development the two proximal, three distal and the most posterior (largest) of the intermediate sensilla are present at hatching (Fig. 8a). At successive instars the other intermediate sensilla are added in a strict sequence around the arc until the adult condition is achieved (Fig. 8b, c, d).

Fig. 8.

Scanning electron micrographs showing successive stages in formation of the normal pattern of campaniform sensilla (a) Instar 1 (left leg); (b) Instar 4 (right leg); (c) Instar 7 (left leg); (d) adult (left leg). In each micrograph the top corresponds to the distal portion of the tibia. Note in (a) the distally pointing protuberance on a proximal sensillum (arrowed). In (b) the larger posterior sensillum of the intermediate group is arrowed. In the adult condition (d) the Proximal Group (PG) and Distal Group (DG) sensilla are indicated. Between them lie the intermediate group. Magnification: (a) 1500 × : (b) 1200 × ; (c) 700 × ; (d) 820 ×

Fig. 8.

Scanning electron micrographs showing successive stages in formation of the normal pattern of campaniform sensilla (a) Instar 1 (left leg); (b) Instar 4 (right leg); (c) Instar 7 (left leg); (d) adult (left leg). In each micrograph the top corresponds to the distal portion of the tibia. Note in (a) the distally pointing protuberance on a proximal sensillum (arrowed). In (b) the larger posterior sensillum of the intermediate group is arrowed. In the adult condition (d) the Proximal Group (PG) and Distal Group (DG) sensilla are indicated. Between them lie the intermediate group. Magnification: (a) 1500 × : (b) 1200 × ; (c) 700 × ; (d) 820 ×

Five experimental limbs (two regenerates, two congruent, one non-congruent grafts) were examined. Although in these experimental limbs approximately the same numbers of sensilla as normally found differentiated, the precise ordered adult pattern was never reformed (Fig. 9). This was despite the fact that sensilla preserved their intrinsic polarity in such cases and that the three size classes corresponding to proximal, distal, and intermediate sensilla appeared to be present also.

Fig. 9.

Scanning electron micrographs of experimental limbs showing campaniform sensilla pattern. Same orientation as in previous figure, (a) congruent graft (Instar II; (b) non-congruent graft-middle limb (Instar 1). Note presence of a distally pointing protuberance, arrowed in (a) indicating each sensillum preserved its intrinsic polarity. By comparing this figure with the previous one it can be seen that approximately the same numbers of sensilla are present but in experimental limbs the precise ordered pattern is not reformed. Magnification : (a) 900 × ; (b) 900 ×.

Fig. 9.

Scanning electron micrographs of experimental limbs showing campaniform sensilla pattern. Same orientation as in previous figure, (a) congruent graft (Instar II; (b) non-congruent graft-middle limb (Instar 1). Note presence of a distally pointing protuberance, arrowed in (a) indicating each sensillum preserved its intrinsic polarity. By comparing this figure with the previous one it can be seen that approximately the same numbers of sensilla are present but in experimental limbs the precise ordered pattern is not reformed. Magnification : (a) 900 × ; (b) 900 ×.

Close examination of regenerated and grafted limbs which appear superficially normal show that pattern formation is never perfect. The sensory structures reveal a marked disparity between the observed results and what one would expect to find in a normally growing limb of similar dimensions. In the case of the two chordotonal organs, the subgenual and tympanal organs, clearly regeneration does not re-establish the adult condition even though these structures are epidermal derivatives.

Subgenual organ

In the case of the regenerated and graft subgenual organ, polarity is preserved in the sense that the neurones are located proximally with proximal running axons, and scolopales distally. In addition two sensilla populations, an anterior and posterior one, were present. In the case of duplicated limbs, the middle limb consistently had the larger sensilla grouping on the anterior side. On this criterion, together with a posteriorly located larger tympana such regenerates appear of stump handedness, indicating the anterior-posterior axis with respect to these structures is capable of reversal, a result not predicted by the polar co-ordinate model (French et al. 1976). Unfortunately, epidermal markers do not exist for other epidermal structures to confirm this result generally.

Tympanal organ

The tympanal organ was in the normal position in terms of circumferential and proximo-distal position within the tibia. The very poor and somewhat variable representation of the tympanal organ may be due to two factors. Firstly, it normally develops considerably later than the subgenual organ (Ball & Young, 1974) and so it seems reasonable that the same should occur during regeneration. Hence insufficient time may have existed for it to differentiate. Secondly, the apparent variable nature of the presence or absence of the tympanal organ itself might have been due to an inability to identify the small population of tympanal organ sensilla. In view of the close anatomical association between subgenual organ sensilla and proximal sensilla of the tympanal organ in normal legs (Ball & Young, 1974) a slight change in proximo-distal position might have led to fusion of the two groups and an inability to detect the tympanal organ sensilla. In grafts and regenerates differentiation of the tympana appears to be independent of the tympanal organ itself since a nearly normal posterior tympanum and the beginnings of the anterior tympanum were evident in cases where internal differentiation of tympanal organ sensilla was apparently absent. The timing and degree of tympanal differentiation was similar to that found in other studies involving regeneration and transplantation in the cricket (Ball, 1979).

Campaniform sensilla

Unlike the two chordotonal organs the campaniform sensilla do not become internalized during development but remain in an epidermal position. As such, they can be treated ideally as being on a two-dimensional cylinder and can be fitted more easily into previous work on pattern formation (French et al. 1976). Clearly each regenerated sensillum preserved its intrinsic polarity since the distinct protuberance always pointed distally. However, the anterior/posterior polarity produced by the size class arrangement within the intermediate group was lost in regeneration. This was because the precise spatial pattern of the adult condition was not re-established. However, the number and diversity of sensilla in terms of the three broad size classes appeared to be retained. This particular system indicates to some degree the independence of polarity and spatial pattern. This system requires further study to establish how the precise normal pattern is lost in regeneration. In normal development the sequential appearance of sensilla of the intermediate group probably controls both the size and distribution of sensilla by mechanisms previously proposed (Lawrence, 1973). What is required is a study of the time sequence appearance of sensilla during various sorts of experimental treatments.

Regeneration and normal development

The assumption implicit and indeed the rationale in most regeneration studies is that they in some way duplicate the events occurring during normal development. This approach is understandable in view of the difficulties in working with embryos, especially where experimental interference is required. However, some recent evidence suggests that a regenerating system does not re-establish the adult condition, at least in insects. O’Farrell & Stock (1954) noted that tracheation in regenerated cockroach limbs is different from that normally found.

In this study tracheation in experimental limbs was variable. The reasons for this variability are unclear. In normal limbs tracheal patterns in the region of subgenual and tympanal organ are consistent from specimen to specimen. In graft and regenerated limbs the variability may be due to different responses to functional demand. However, this is difficult to reconcile with the consistent pattern found in normal limbs. -

Differences in muscle development, its biochemical properties and nerve root branching pattern found in cockroach regenerated limbs led Denberg and Whitington (1978) to suggest that regeneration may not be a good model system in which to examine normal development. This study confirms this conclusion, showing that the three-dimensional arrangement of chordotonal organs is lost and the two-dimensional pattern of the campaniform sensilla is only partly restored, while the gross morphology of the limb appears normal.

I am grateful to Dr David Young for his advice and criticism. Thanks to Daphne Hards and Linda Crosby for technical assistance. Jane Doolan and Ralph MacNally commented on the manuscript. Support was provided by the Commonwealth Postgraduate Research Award (CPRA) and Victorian Education Department Studentship.

Ball
,
E.
(
1979
).
Development of the auditory tympana in the cricket Teleogryllus commodus (Walker): Experiments on regeneration and transplantation
.
Experientia
35
,
324
325
.
Ball
,
E.
&
Young
,
D.
(
1974
).
Structure and development of the auditory system in the prothoracic leg of the cricket Teleogryllus commodus (Walker). II. Post-embyronic development
.
Z. Zellforsch
.
147
,
313
324
.
Bennett
,
H. S.
,
Wyrick
,
A. D.
,
Lee
,
S. W.
&
McNeil
,
J. H.
(
1976
).
Science and art in preparing tissues embedded in plastic for light microscopy, with special reference to glycol methacrylate, glass knives and simple stains
.
Stain Technol
.
51
,
71
97
.
Bohn
,
H.
(
1965
).
Analyse der Regenerationsfâhigkeit der Insektenextremitat durch Amputations und Transplantation versuche un Larven der Afrikanischen Schabe Leucophaea moderae. Febr (Blattaria) II. Mitt Achsendetermination
.
Arch. Entw Mech. Org
.
156
,
449
503
.
Carlson
,
B. M.
(
1975
).
The effects of rotation and positional change of stump tissues upon morphogenesis in the regenerating axolotl limb
.
Devl Biol
.
47
,
269
291
.
Denburg
,
J. L.
&
Whitington
,
P. M.
(
1978
).
Regeneration of entire legs in cockroaches as a model for developmental events
.
Experientia
34
,
252
253
.
Eibl
,
E.
(
1978
).
Morphology of the sense organs in the proximal parts of the tibia of Gryllus campestris L. and Gryllus bimanculatus (de Geer) - Insecta Ensifera
.
Zoomorphologie
89
,
185
205
.
French
,
V.
(
1978
).
Intercalary regeneration around the circumference of the cockroach leg
.
J. Embryol. exp. Morph
.
47
,
53
84
.
French
,
V.
,
Bryant
,
P. J.
&
Bryant
,
S. V.
(
1976
).
Pattern regulation in epimorphic fields
.
Science
193
,
969
980
.
Lawrence
,
P. A.
(
1973
).
Polarity and patterns in the post-embryonic development of insects
.
Adv. Insect. Physiol
.
7
,
197
225
.
Moulins
,
M.
(
1976
).
Ultrastructure of chordotonal organs
.
In Structure and Functions of Proprioceptors in the Invertebrates
(ed.
P. J.
Mill
).
London
:
Chapman & Hall
.
O’Farrell
,
A. F.
&
Stock
,
A.
(
1954
).
Regeneration and the moulting cycle in Blatella germánica L. II. Successive regeneration of both metathoracic legs
.
Aust. J. biol. Sci
.
7
,
525
536
.
Pantin
,
C. F. A.
(
1946
).
Notes on Microscopical Techniques for Zoologists
.
Cambridge University Press
.
Wigglesworth
,
V. B.
(
1953
).
The origin of sensory neurones in an insect Rhodinius proxilus (Hemiptera)
.
Quart. Jl. micros. Sci
.
94
,
93
112
.
Wolpert
,
L.
(
1969
).
Positional information and the spatial pattern of cellular differentiations
.
J. theor. Biol
.
25
,
1
47
.
Wolpert
,
L.
(
1972
).
The concept of positional information and pattern formation
.
In Towards a Theoretical Biology
(ed.
C. H.
Waddington
).
Edinburgh University Press
.
Young
,
D.
&
Ball
,
E.
(
1974
).
Structure and development of the auditory system in the prothoracic leg of the cricket Teleogryllus commodus (Walker. I. Adult structure
.
Z. Zell forsch
.
147
,
293
312
.