J. Embryol. exp. Morph. Vol. 54, pp. 47-74, 1979 47 Printed in Great Britain © Company of Biologists Limited 1979

Quantitative staging of embryonic development of the , nitens

By DAVID BENTLEY,1 HAIG KESHISHIAN, MARTIN SHANKLAND AND ALMA TOROIAN-RAYMOND From the Department of Zoology, University of California, Berkeley

SUMMARY During development of the grasshopper embryo, it is feasible to examine the structure, pharmacology, and physiology of uniquely identified cells. These experiments require a fast, accurate staging system suitable for live embryos. We present a system comprising (1) subdivision of embryogenesis into equal periods, (2) expression of stage in percent of complete embryogenesis time, (3) characterization of stages by light micrographs (and descriptive text), and (4) illustration of stages at the egg, embryo, and limb levels of resolution. Advantages of a percent-system include communicability, flexibility in temporal resolution, accurate assignment of elapsed time in developmental processes, and uniform coverage of the period of embryogenesis. The stages described are at 5 % intervals with an estimated error of ± 1 %.

INTRODUCTION Recently it has become possible to investigate the physiology, pharmacology, and morphology of single, identified neurons, neuroblasts and other cell types during embryogenesis in (Bate, 197'6 a, b; Spitzer, 1979; Goodman & Spitzer, 1979; Goodman, O'Shea, McCaman & Spitzer, 1979; Bentley & Toroian-Raymond, 1979). The paucity of preparations in which these approaches are feasible has made grasshopper embryogenesis particularly attractive for analysing many problems in developmental neurobiology and developmental biology in general. To accurately characterize the time course of developmental events, it is necessary to have a precise, rapid staging system, applicable to unstained, living material and covering the entire period of embryogenesis. Such a staging has not been available. There is an extensive literature on grasshopper embryogenesis extending from the mid-nineteenth century (Packard, 1883; Wheeler, 1893; Slifer, 1932a; Roonwal, 1936; Johannsen & Butt, 1941; Anderson, 1972). Several systems for staging development have been described (Table 3). Although various features of these systems are excellent, no single one covers the entire course of embryo- genesis with the temporal and cellular detail now required. Most descriptions 1 Author's address: Department of Zoology, University of California, Berkeley, CA 94720, U.S.A. 48 D. BENTLEY AND OTHERS are illustrated by camera-lucida or free-hand drawings of whole eggs or embryos; many cover only a portion of embryogenesis. Most stages have been marked by easily observable changes in external morphology or orientation of the embryo; as these events are not distributed uniformly throughout development, relatively long, uncharacterized periods occur in these systems. In this paper, we present a new staging system. Its main features are: (1) the stages evenly subdivide the period of development; (2) stages are expressed as a percentage of total developmental time; (3) stages are illustrated by light photomicrographs; (4) stages are illustrated at three levels of increasing resolu- tion (egg; embryo; limb).

MATERIALS AND METHODS The experimental were , initially captured in 1963 and maintained in culture for approximately 50 generations. The animals were raised in small, crowded cages at 31 ± 1 °C, 16L/8D light cycle, and 60 ± 5 % relative humidity on a diet of freshly sprouted wheat supplemented by wheat- germ and dry dog-food. Generation time was 10-12 weeks. Egg-pods were deposited in 6 cm diameter/10 cm high paper cups containing cleaned no. 1 sand moistened by 15 % water by weight. Cups were capped during incubation. Pods contained from 25 to 100 eggs. While pods used for maintain- ing the culture could be left undisturbed, those intended for experiments had to be opened. Two methods were used for accomplishing this. In the first, the pod was placed on top of the moist sand, broken into several clusters of exposed eggs, and covered with damp cotton; alternatively, the pod was com- pletely dispersed and eggs were individually washed in distilled water and placed separately on filter paper kept at a constant moisture level by capillary wetting. Eggs were kept at various temperatures in an incubator accurate to ± 0-5 °C, and at a relative humidity of 60 ± 2 %. We wanted to express developmental stage as a percentage of total develop- mental time, and further, to place the stages at equal intervals throughout embryogenesis. To accomplish this, it was necessary to have a group of syn- chronously developing eggs whose total developmental time could be accurately predicted. Sample eggs could then be withdrawn from this group at equal intervals (5 % of total developmental time in this case) and described. Eggs within the same pod formed our synchronous groups. The total develop- mental time of each egg was determined with an accuracy of ± 45 min. Grass- hopper eggs are fertilized when deposited (McNabb, 1928; Slifer & King, 1934) and completion of a pod takes about 1 h. We selected only pods where deposition was observed, so the time of fertilization could be determined to ± 30 min. Hatching time was established by continuously observing each pod and counting and removing all the nymphs which hatched within each 30 min period; the hatching time of each nymph was consequently known to ±15 min. This Quantitative staging of embryonic development of Schistocerca 49 information was obtained from ten pods, and formed the data for a quantitative characterization of the synchrony of hatching within pods (Figs. 1, 2). Five percent staging required the selection of 20 equally spaced observation times during the complete period of development. This was done by determining the temperature at which the eggs developed in 20 days, and then observing the eggs each day at the time of initial deposition of the pod. The appropriate temperature was predicted by observing the development of 192 pods at temperatures ranging from 30 to 35 °C. The actual developmental time of each of the pods used for staging was determined as described above. The staging descriptions are designed to be pre-experimental, and therefore are based on features which can be seen in unstained, living embryos with a dissecting microscope. Correspondingly, all photomicrographs were made with a dissecting microscope (Wild M5A; a few additional features which can be seen in simple squashes in a compound microscope are noted). A text description of each stage is provided as well as light micrographs at three levels of resolution: (1) the whole egg showing the size, location, and orientation of the embryo; (2) the embryo; (3) detail of the metathoracic leg. Eggs were immersed in 3 % sodium hypochlorite for 1 min; this procedure clears the chorion but doesn't remove it (longer exposure clears better but causes osmotic changes which alter the shape of the embryo). Cleared eggs were photographed in dark-field illumination; embryos were photographed in transmitted illumination or, after they became opaque, in incident illumination. All colors are described from incident illumination. The saline in which embryos were examined comprised NaCl 140 mM, KC1 5mM, CaCl2.2H2O 4mM, MgSO4.7H2O 2mM, TES 2 HIM, dextrose 55-65 mM, pH 7-2, osmolarity; 310-325 milliosmol/kg. High osmolarity was crucial to maintaining physiological condition and had to be adjusted for age of the embryo (Carlson, 1961; Grellet, 1968). It was altered by varying dextrose concentration to maximize heart rate, peristalsis rate, neuroblast mitosis frequency or, in early stages, to prevent shrinking or swelling of the amniotic cavity. The final staging is based upon precisely timed observation of eggs from eight pods. At each observation period, descriptions of several eggs from each pod were made; for three pods, photographs of several eggs were made each day. Confirmatory observations have been made on many additional pods.

RESULTS The synchrony of hatching of eggs within the same pod is shown in Fig. 1. These four pods were maintained on sand under moist cotton (first method); the cotton was removed a few hours before the onset of hatching. The degree of synchrony can be expressed by calculating the percentage of eggs which hatched within a period equal to 1 % of total developmental time (Fig. 1). In three of four 50 D. BENTLEY AND OTHERS

10

Pod C Pod I) n = 50 7o= 100 '.'o= 100

•3 40

30

20

10 -

19 20 19 2G Developmental time (days) Fig. 1. Hatching synchrony of eggs within each of four pods (A, B, C, D; incubated on sand), n = the number of eggs hatching from each pod; % = the percentage of hatching eggs from each pod which hatched within a period equal to 1 % of the total development time. Quantitative staging of embryonic development of Schistocerca 51

20-5 Development time (days) Fig. 2. Hatching synchrony of eggs within each of four pods (E, F, G, H; incubated on filter paper), n = the number of eggs hatching from each pod; % = the per- centage of hatching eggs from each pod which hatched within a period equal to 1 % of the total developmental time. In pod H, seven eggs hatched when all the eggs were wetted with cool water at day 19; all of the remaining eggs hatched within the 9 h period shown, but the exact time of hatching was not noted. pods, all the eggs hatched within this period. This degree of synchrony indicates that the time of hatching of a subset of eggs from a pod is an accurate estimate of when eggs removed for staging descriptions would have hatched if they had been left undisturbed. Therefore, the percentage of developmental time experienced by embryos withdrawn for staging can be estimated within an error ofl%. An additional problem is the degree to which synchronous hatching indicates synchronous development. Mechanisms, such as pheromonal or mechanical stimulation, might be present which initiate simultaneous hatching in eggs that are actually at slightly different stages of development. To evaluate this factor, 52 D. BENTLEY AND OTHERS

Fig. 3. Developmental synchrony between eggs from the same pod at different stages in embryogenesis. Metathoracic limbs of four embryos from the same pod are shown at 55 % development and at 90 % development. The degree of similarity between these limbs is representative of the range of variation normally encountered and indicates that synchrony falls well within the 5 % range throughout embryo- genesis. Transmitted illumination. Quantitative staging of embryonic development of Schistocerca 53

Table 1. Features of 5 % developmental stages Stage Characteristics (distinguishing from previous stage)* 0 % Egg light yellow; lipid droplets uniform in size 5 % Egg brown; lipid droplets variable in size; energids in posterior yolk 10 % Disc-shaped blastoderm; energids throughout yolk 15 % Embryo cephalized; amnion present; early yolk cleavage 20 % Primary segmentation; embryo post-anatrepsis; yolk cleavage about two thirds length of egg 25 % Segmentation to third abdominal segment; neuroblasts visible; leg rudiments larger than mouthpart rudiments 30 % Embryo fully segmented; eye region delaminated; rudiments on firsttw o abdominal segments 35 % Rudiments on all abdominal segments; primary cuticle visible; proctodeum in tenth abdominal segment 40 % Separation of femur and tibia in metathorax; embryo half length of egg; proctodeum in ninth abdominal segment 45 % Separation of tibia and tarsus in metathorax; katatrepsis initiated; proctodeum in eighth abdominal segment 50 % Pigmentation of eye-plate; embryo post-katatrepsis; metathoracic tibia parallel to femur 55 % Antennal furrows; metathoracic tibia reaches base of femur; herringbone array of extensor tibia muscle fibers 60 % White line on eye-plate; rotation of embryo; embryo more than three quarters length of egg 65 % Double curvature of metathoracic tibia; leg twitching; synchronous contraction of median sinus 70 % Metathoracic tibia straightened; embryo pale green; white line close to anterior margin of eye 75 % Longitudinal rows of brown pigment spots line metathoracic femur; brown pigment on dorsal, caudal midline 80 % Brown pigment spots on all legs; apolysis of second embryonic cuticle completed 85 % Clearing of femoral crescent; embryo bright green; vertical eye stripes 90 % Tar sal claws are black; brown pigmentation covers compound eye; transverse, dark-green stripes on metathoracic femur 95 % Black hairs on antennae; integument opaque white; dark blue color within antennae and metathoracic legs 100 % Egg hatches; in saline, embryos near 100 % begin peristaltic contractions * Italics indicate most unequivocal feature. we examined groups of embryos from the same pod at various times in develop- ment. Inspection of rapidly changing features, such as the degree of differentia- tion of the metathoracic limb (Fig. 3), indicated that most embryos within a pod were at a very similar stage of development. Occasional out-of-step embryos were encountered. These were almost invariably behind their pod-mates, suggesting that they may have been moribund individuals which would have failed to hatch. Significant variability in the numbers of such individuals occurred between pods. We observed pods which hatched in total developmental times ranging from about 19-5 to 20-5 days. This gave a good sampling of the developmental rates 54 D. BENTLEY AND OTHERS surrounding and including 20 days. Daily sampling at these rates provided the basis for our characterization of 5 % developmental stages. The photomicro- graphs (Figs. 3-9) were made from pods characterized in Fig. 2. Key features delineating each 5% stage are briefly summarized in Table 1. The following is a more detailed description of each stage:

0% Freshly deposited eggs are light yellow. Within about 3 h, they tan to a dark brown. Eggs on the exterior of the pod tan first. The change in coloring is due to a darkening of the outer egg envelope or chorion, which is initially transparent and reveals the yellow, yolk-filled interior of the egg. Yellow, lipid yolk droplets (Mahowald, 1972) are initially uniform in size, about 20 ± 5 /mi in diameter (be aware that they will begin to fuse when ex- pressed from the egg into saline). As the egg develops they become much more heterogeneous. The poles and axes of the egg are assigned by convention with respect to the orientation of the egg in the maternal ovariole (Mahowald, 1972; Anderson, 1972). The posterior pole is marked by a prominent, opaque, cap-like etching of the chorion (Fig. 4-5). At the base of this cap are the micropyles through which the sperm enter the egg. The anterior pole lies at the opposite, more pointed, end of the egg. Eggs are usually curved. The concave side is dorsal and the convex side ventral (Fig. 4-50); however, uncurved or doubly curved eggs are regularly found.

5% The egg is brown. Lipid yolk droplets are highly variable in diameter, ranging from 10 to 100 /«n (Fig. 6 - 5). Dispersed among these droplets in the region of the posterior pole are colorless, oblong islands of cytoplasm, often containing round nuclei about 40 jam in diameter. These are the cleavage energids from which the embryo, embryonic membranes, and yolk cells will develop (Anderson, 1972). Energids cannot be detected through the cleared chorion; they are most easily seen by puncturing the appropriate part of the egg, squashing the expressed yolk under a coverslip, and viewing in a compound microscope.

10% At the posterior pole, located toward the ventral side of the egg, is a small cluster of cells floating on the surface of the yolk droplets (Fig. 6 - 10). The cluster is a disc-shaped monolayer about 300 /im in diameter and comprises the primitive blastoderm, from which the embryo will develop. Cleavage energids are now distributed throughout the entire length of the egg, although they are more numerous toward the posterior pole. Quantitative staging of embryonic development of Schistocerca 55

From the outside of a cleared egg, the embryo is now visible on the posterior/ventral surface (Fig. 4-15). Its dorsal side is apposed to the yolk and the anterior end faces the anterior pole of the egg. The embryo has differentiated a widened, anterior, cephalic region (proto- cephalon) extending about half its length (Fig. 6-15). Antennal and pre- antennal segments will arise here. Mouthpart, thoracic, and abdominal segments will develop from the narrow, posterior end of the embryo (protocorm). The embryo is one cell layer thick except for a band of cuboidal cells on the dorsal midline forming a second, inner layer. This inner layer extends about two thirds the length of the embryo, from the posterior end to the mid-proto- cephalon. At its anterior end, it expands to form two lobes. The inner layer arises from invagination of cells during gastrulation (Roonwal, 1936; Anderson, 1972) and comprises the presumptive mesoderm. A thin membrane (the amnion) is attached to a ridge of columnar cells at the perimeter of the embryo (Fig. 6-15) and covers the entire ventral surface. This membrane and the extra-embryonic membrane (the serosa) arise at a slightly earlier stage by the fusion of amniotic folds (Roonwal, 1936). The serosa eventually forms a sac enclosing the yolk (50 %; Fig. 7 - 55). The yolk remains a dispersion of variably sized droplets except at the extreme posterior pole where large, oblong cells, from 200 to 300 /*m in length, mark the onset of yolk cleavage. The yolk is gradually ingested by the formation of these cells in a posterior-anterior progression.

20% The location of the embryo has shifted due to an immersion and rotation within the yolk (anatrepsis; Wheeler, 1893). The embryo is now on the dorsal side of the egg, with its anterior end facing the posterior pole and its dorsal surface apposed to the yolk. The protocorm has elongated and differentiated into two distinct regions separated by a slight flexure and thickening of the embryo (Fig. 6 - 20). The anterior region contains the presumptive mouthpart tissue, while the posterior region will give rise to the thorax and abdomen. The protocephalon has a medially located depression on its ventral surface, the oral opening (stomodeum). A pair of small protruberances slightly posterior and lateral to the stomodeum mark the antennal rudiments. The cleavage of the yolk has progressed to approximately two thirds of the length of the egg, giving the yolk the appearance of a cellular mosaic (Fig. 4 - 20). Anterior to this area, the yolk still consists of variably sized droplets. 56 D. BENTLEY AND OTHERS

25% Mouthpart, thoracic, and the first three abdominal segments are now clearly delineated (Fig. 6 - 25). The segments in the mouthpart and thoracic regions each possess a pair of ventrally placed protruberances, the limb buds. The thoracic limb buds as a group are larger than the mouthpart limb buds, with the largest pair formed by the metathoracic segment. The metathoracic limb bud consists of an inner mass of cuboidal cells surrounded by a rind of tightly packed columnar cells (Fig. 8-25); all limb buds appear this way when they first differentiate. Anterior to the stomodeum is a small protruber- ance, the presumptive labrum. The antennal rudiments, lateral to the stomo- deum, have enlarged and turned medially. The abdominal segments at this stage lack limb buds. The mesoderm has become divided into a series of hollow, segmental blocks of tissue in both the mouthpart and thoracic regions. Mesodermal tissue in the abdomen is flattened. The unsegmented caudal tip of the embryo possesses a small invagination, the primitive anus (proctodeum). Along the ventral, median surface is a band of large spherical cells, neuro- blasts (Wheeler, 1893; Carlson, 1961; Bate, 19766). In the protocephalon they are distributed in packets anterior and lateral to the stomodeum, while in the protocorm they are arranged in a metamerically repeated pattern. Cleavage of the yolk has been completed.

30% The embryo is fully segmented (Fig. 6 - 30). The first two of the eleven abdominal segments possess limb buds (note that the lateral edge of the segment is easily mistaken for the limb bud, which lies more medially on the ventral surface). The metathoracic leg has elongated and has a slight, medially directed bend (Fig. 8 - 30). The mesothoracic and prothoracic legs are about the same size, but smaller than the metathoracic leg. The labial and maxillary rudiments are also of equal size, and are about twice as long as the mandibular rudiment. A thin membrane, the provisional dorsal closure, extends across the dorsal surface of the embryo. This membrane is supplanted later by a true, dorsal ectoderm. On the ventral side of the embryo, the amnion is stretched tightly over the segmental appendages, folding them medially. The head capsule has two prominent lobes which will eventually be occupied by the compound eyes. The posterior portion of these lobes has delaminated into two layers separated by a space. The outer layer is the eye plate (Roonwal, 1937) which comprises the presumptive retina and associated ommatidial structures; the inner layer will become the distal portion of the optic lobe of the brain. The eye plate is curved, and its posterior (medial) margin is thickened. Quantitative staging of embryonic development cf Schistocerca 57

35% Limb buds occur on all eleven abdominal segments. Those of the first segment, the pleuropodia, are multi-lobed and are much larger than buds of other segments. They are embryonic structures lost at hatching. Limb buds of segments 3-10 are ventral, while the remaining buds (1,2,11) extend laterally. The primary cuticle covers the surface of the embryo, and is visible as a clear film stretched over the appendages and into the opening of the proctodeum. This cuticle is later detached from the epidermis (apolysed) and replaced by a secondary cuticle which is shed after hatching (Mueller, 1963; Micciarelli & Sbrenna, 1972). A slight invagination of the columnar cell rind is visible at the tip of the metathoracic leg. It marks the beginning of differentiation of the claw retractor tendon (apodeme), one of three leg tendons (Snodgrass, 1929). Maxillary and labial limb buds have become trilobed. A pair of neurons, the pioneer fibers (Bate, 1976 a), can be seen within each antenna. The cell bodies are located at the tip of the antenna, adjacent to the columnar cell rind. Similar cells arise in other appendages. In each segment, neuroblasts have proliferated clusters (presumptive ganglia) of ganglion mother cells and undifferentiated neurons (Bate, 1916 b; Goodman & Spitzer, 1979). Fine, intersegmental fibers run longitudinally on both sides of the midline in segments anterior to the abdomen. They are viewed most easily from the dorsal aspect, and comprise early, axonal outgrowths of ventral cord neurons. The proctodeum has invaginated to the anterior border of the tenth abdominal segment (Fig. 6 - 35).

40% The embryo is visible as a wedge extending from the posterior pole for about half the length of the egg (Fig. 4 - 40). The metathoracic leg has a notch on its medial side (this will be the ventral side of the adult leg) which separates the femoral and tibial regions (Fig. 8 - 40). The claw retractor tendon has extended further proximally from its invagination, and the leg is heavily invested with spindle-shaped cells (probably myoblasts) and nerve fibers. These features also occur in the other thoracic and mouthpart appendages. In the central nervous system, intersegmental fiber bundles extend along the entire length of the embryo. Transverse, intrasegmental nerve fibers can be seen in all segments. The proctodeum has invaginated to the anterior border of the ninth abdominal segment (Fig. 6-40). 58 D. BENTLEY AND OTHERS

45% The embryo has begun moving around the posterior pole of the egg (Fig. 4 - 45). At the completion of this movement (katatrepsis; Wheeler, 1893; Slifer, 19326; Anderson, 1972), it will face the anterior pole of the egg and its ventral surface will appose the ventral side of the egg. During katatrepsis, the embryo generates a series of rhythmical, metachronal waves of posterior to anterior contractions. The waves can be seen in cleared eggs, and occur at 10-20 times per minute in 25 °C saline. They continue in less pronounced form after the completion of katatrepsis. The metathoracic leg has an additional constriction separating the tibial and tarsal regions (Fig. 8 - 45). Two tendon invaginations are just noticeable on the femur, the extensor tibia tendon on the lateral aspect and the flexor tibia tendon on the medial (Fig. 8 - 45, 50). The antennae show the onset of seg- mentation, with four thickenings of the outer, columnar rind. The eleventh abdominal appendages, cerci, have broadened into flattened lobes on either side of the proctodeum (Fig. 6 - 45). The posterior margin of the eye plate contains spindle-shaped cells extending across its full depth. These appear to be retinula cells of the differentiating ommatidia. Large number of fibers cross to the distal surface of the optic lobes. Clusters of white cells mark the appearance of the lateral ocelli. The fibers of the ventral nerve cord have formed a ladder-like arrangement, with a pair of distinct bundles of transverse fibers in each segment intersecting the longitudinal, intersegmental bundles on both sides of the midline. On either side of the midline, the dorsal surface of the embryo is flecked with white spots. These spots are composed of heavily pigmented cells in a tissue layer which becomes the fat body. The proctodeum has invaginated to the anterior border of the eighth abdominal segment (Fig. 6 - 45).

50% The embryo has completed its movement around the posterior pole (Fig. 4 - 50). Anterior and dorsal to the head is a plug of yolk enclosed by the serosa and occupying about half the volume of the egg. The metathoracic leg has flexed so that the tibia is parallel to the femur (Fig. 8 - 50). Spurs are visible at the distal end of the tibia. The tarsus has divided into two segments (this division has not yet occurred in the other legs). The cerci have enlarged and formed a nipple-like process at the tip. Sexual differences in the genitalia can be distinguished (Karandikar, 1942). The eye plate has an unlayered red-brown pigment along its posterior margin (eye axes will be given with respect to adult eye position). Differentiation of retinula cells has proceeded about half way to the anterior edge of the eye. The median ocellus is present. In the ventral nerve cord, a broadened, fibrous Quantitative staging of embryonic development of Schistocerca 59 region occurs at the intersection of the fiber bundles. This region is the incipient neuropil of the embryonic ganglia. There is marked apolysis of the primary embryonic cuticle, with the secondary cuticle forming underneath (Mueller, 1963; Micciarelli & Sbrenna, 1972). The proctodeum has invaginated to the anterior border of the seventh abdominal segment.

55% The ventral aspect of the embryo is still adjacent to the ventral (convex) side of the egg. The head extends between half and two thirds of the distance to the anterior pole (Fig. 5 - 55). The amnion now closes the dorsal surface up to the prothorax (Fig. 7 - 55). Yolk is continuous from the serosal sac through the open cervical dorsum into the midgut of the embryo. The yellow midgut yolk is encased by transparent fat-body tissue containing a profusion of white cells (Fig. 7 - 55, 60). A clear band along the dorsal midline of the embryo demarcates the median blood sinus, antecedent to the heart. Peristaltic, anteriorly directed constrictions of this sinus are a continuation of the rhythmical activity first seen at the 45 % stage. The metathoracic tibia reaches to the base of the femur (Fig. 8 - 55). The tip of the leg has a longitudinal furrow marking the tarsal claws. Well-differentiated muscle tissue is visible in transmitted light. A distinctive, herringbone array of muscle fibers, the incipient extensor tibia muscle (135a and b; Snodgrass, 1929; Fig. 8 - 55), occurs along the dorsal (previously medial; 45 %) tendon of the metathoracic femur. Localized muscle fiber contractions can be seen through the cuticle, but the leg does not twitch noticeably. The genital appendages (on the eighth and ninth abdominal segments of the female and the ninth and tenth of the male) shift toward the ventral midline (stage 3; Karandikar, 1942). No other abdominal limb rudiments remain between the pleuropodia and the cerci. Pronounced furrows in the epidermis divide the antennae into annular segments (Fig. 7 - 55). A brick-red band starts at the posterior margin of the compound eye and extends forward one fourth of its width (Fig. 7 - 55). This band comprises both a superficial and a deep layer of the same pigment. Rows of facets line the surface of the eye. The superficial pigment accumulates at the borders of the facets, lending a speckled appearance to its layer. The deeper pigment layer is unbroken, but bears a pattern of ommatidial silhouettes on its outer surface (in the adult ommatidium, there are red-brown pigments present in both the photoreceptors and in two layers of pigment cells that line the outside of each ommatidial cartridge; Roonwal, 1947).

EMB 54 60 D. BENTLEY AND OTHERS

60% Usually the embryo has expanded to fill the entire egg except for a small space at the anterior pole (Fig. 5 - 60). There is considerable variation both in the completion of expansion and in the amount of space left; about a third of our embryos did not finish this process until the next stage. Expansion is accompanied by a 180° rotation of the embryo about its longi- tudinal axis (Slifer, 19326; Bodenheimer & Shulov, 1951; Jones, 1956), leaving its ventral surface adjacent to the dorsal (concave) side of the egg (Fig. 5 - 60). This orientation is maintained for the remainder of embryogenesis. Rotation of the embryo was confirmed by marking the surface of the egg with wax. As the embryo fills the egg, the yolk is engulfed by the expanding midgut (Fig. 7 - 60). The regressing yolk sac transforms into a tubular protruberance of degenerating serosal cells (dorsal organ; Wheeler, 1893) which sinks into the midgut to allow the completion of dorsal closure. After closure, a pair of bilateral bladders, the cervical ampullae, lie between the head and the deeply wrinkled pronotum. They function during hatching and do not persist into the first instar (Bernays, 1971). Genital rudiments of the ninth abdominal segment have partially fused along the ventral midline in both sexes. A pair of narrow, flattened rudiments are barely distinguishable at the posterior margin of the female's eighth abdominal segment, and the tenth segment rudiments of the male have completely disap- peared underneath the fused ninth pair. The cercus has elongated into a cone which lies folded beneath the abdomen (Edwards & Chen, 1979). A white line extends dorso-ventrally along the anterior margin of the eye plate (Fig. 7 - 65) which is about halfway across the presumptive compound eye. An unpigmented zone divides this white line from the parallel posterior band of red pigment. White pigment cells in each ocellus begin to coalesce into a disc fronted by a lens primordium.

65% A narrow space persists between the embryo's head and the anterior pole of the egg (Fig. 5-65). The metathoracic tibia assumes a double curvature unique to this stage (Fig. 9 - 65). The proximal part of the tibia bends away from the femur, and the distal half gradually curves back toward it. This morphological charac- teristic is more reliable than behavioral features for identification of the stage (note: transitional forms obviously occur in the appearance and disappearance of this and other features; therefore it is important to rely also upon more subtle changes in the balance of form, such as the relative size of the abdomen and metathoracic leg, which can be apprehended by studying photographs of the different stages). The legs begin to twitch. These periodic jerks are easily distinguished from Quantitative staging of embryonic development of Schistocerca 61 the smooth limb displacement accompanying bodywall peristalsis. Individual limbs extend rhythmically at one or more joints, but there is no apparent co- ordination of frequency or phase between limbs. Limb movements are not evident until the embryo is removed from the egg. Peristaltic waves are replaced by simultaneous contraction along the entire length of the median sinus (Nelson, 1931). Apparently these contractions represent the beginning of normal heartbeat (Roonwal, 1937, described the formation of a definitive heart wall within this sinus at an equivalent stage in Locustd). White fat cells begin to appear beneath the transparent heart tube, reflecting its detachment from the midgut and the subsequent intrusion of the fat body. The labia have shifted medially and fused at the base. Other mouthparts remain laterally oriented and do not close over the stomodeum until hatching.

70% Prior to this stage, the embryonic tissue has been transparent or translucent. The embryo now becomes a very pale green, particularly the head and legs. The metathoracic tibia retains little of its prominent curvature from the previous stage (Fig. 9 - 70). The tibia will not become perfectly straight until after hatching. Brick-red pigment covers the posterior half of the compound eye (Fig. 5 - 70, 7 - 70). The superficial pigment layer extends further anterior than the deep layer, forming a speckled, light-red band along its leading edge. A colorless zone still separates the red region from the anterior white band (Fig. 7 - 70). This band has also moved anteriorly and approaches the frontal margin of the eye. In some individuals, brown pigment fringes the cuticular sleeve enveloping the mandible rudiment (this feature should not be mistaken for the extensive pigmentation of the mandible teeth that occurs at 90 % development).

75% This stage is distinguishable by the appearance of longitudinal rows of brown spots on the metathoracic femur (Fig. 7-75; 9 - 85). The rows contain about ten spots each, and mark incipient ridges along the dorsal and ventral edges of the femur. There are no spots yet on the tibia or on the other legs. Faint brown pigment also appears on the midline of the dorsal body plates (tergites) of the caudal-most segments. The anterior white line has reached the frontal margin of the compound eye (Fig. 7 - 75), and will remain throughout embryogenesis into the first instar. The red pigment region has turned dark brown and continued to move anteriorly. The facets are finely outlined in white, and this gives the eye a frosted appearance under incident illumination. White teeth begin to form on the medial edge of the mandible rudiment, within the cuticular sleeve. 5-2 62 D. BENTLEY AND OTHERS

80% For most embryos, the head is tightly pressed into the anterior end of the egg (Fig. 5 - 80). Brown pigment spots are present on the femur and tibia of every leg, but not on the head or thoracic body-wall. A faint, brown dorsal midline extends forward to the mesothorax. The second embryonic cuticle separates from the underlying epidermis over the entire body surface, and the third cuticle begins to form. The second cuticle remains intact throughout embryogenesis and is shed soon after hatching (Bernays, 1972). The third cuticle will be the cuticle of the first instar. Apolysis of the second cuticle has been described from histological sections for S. gregaria resembling our stages 70-75 (Micciarelli & Sbrenna, 1972), and can also be seen in S. nitens embryos of this stage with a compound microscope. The brown pigment band of the compound eye may encroach upon the posterior edge of the dorsal spot, a smooth, indistinctly faceted ellipse the size of an ocellus situated at the dorsal margin of the eye.

85% The entire embryo is bright green in color, although the midgut yolk still lends a yellow cast to the abdomen. Brown spots appear on the frontal head, pronotum, and posterior margins of the meso- and metathoracic tergites. The dorsal midline darkens. A large, transparent crescent develops at the distal end of the metathoracic femur (Fig. 7-85; 9-85), adjacent to the tibial articulation (Tyrer, 1970). This structure may sometimes be recognized at 80 % development as a small, indistinct outline that has not cleared. The brown eye band continues to expand toward the anterior, and now fills the posterior half of the dorsal spot. The brown color remains dark within the boundary of this structure, but pales elsewhere except for two vertical stripes which persist within the light-brown region (Fig. 7 - 85, 90). These stripes will disappear later and are not antecedent to the postembryonic striations described by Roonwal (1947). A circular, black image known as the pseudopupil appears to lie beneath the eye surface (Fig. 7 - 85, 90). The pseudopupil is not a structure but an optical illusion manifest in the organization of the eye (Horridge, 1977). Cereal sensory hairs have grown into the empty tip of the cuticular sheath in some individuals (they are not pigmented and must be viewed in transmitted light). The presence of sensilla demonstrates the deposition of a definitive first instar cuticle on the surface of the epidermis (Edwards & Chen, 1979).

90% Three broad transverse stripes of dark green appear on the metathoracic femur (Figs. 7-90; 9-90), accentuating the herringbone pattern on the Quantitative staging of embryonic development of Schistocerca 63 external surface. The pattern reflects a double oblique array of low ridges overlying the insertions of the extensor tibia muscle fibers on to the inner side of the cuticle. Brief cuticle indentations occur sporadically within the insertion area. They are reminiscent of the fiber twitches seen at stage 55, and imply that functional muscle insertions have been made on to the cuticle (Sharan, 1958). The indentations need not be correlated with whole leg movements. Light brown pigmentation covers the full width of the compound eye and is contiguous with the white anterior border. The entire dorsal spot is dark brown. The tarsal claws and tibial spurs turn black (Fig. 9 - 90), and the cereal hairs now appear black under incident illumination. The teeth of the mandible, and sometimes those of maxilla, are a lustrous brown (Fig. 7 - 90). Brown spotting has progressed to the antennae, tarsi, palps, abdominal tergites, and occipital area of the head.

95% The embryonic integument turns opaque white on much of the head and limbs and on small patches of the body. Brown and black markings are prominent against this background. The degree of white coloration on the abdomen varies greatly between individuals, so that the heart may be either exposed or obscured. The teeth of both the mandible (Fig. 7 - 95) and maxilla darken to black at the tips, as do the tibial spines which are pressed flat against the legs by the embryonic cuticle (Fig. 9 - 95). Black hairs appear on the legs, antennae, maxillary palps, head, and tergites. Dark blue tissue appears within the metathoracic legs and the antennae. This pigmentation is partly concealed by the white color, but the blue tint is pronounced around the femoro-tibial articulation, beneath the marginal ridges of the femur, and in the proximal segments of the antennae. Eye color progressively pales. The vertical stripes have faded, and the eye is an essentially uniform shade of brown except for the dark dorsal spot. In most individuals, the spontaneous cuticle indentations seen in the previous stage are no longer apparent.

100% No external morphological features unequivocally distinguish this stage from the previous. The most distinctive difference between the two is their respective competence to hatch. Hatching is effected by a series of powerful, anteriorly directed peristaltic contractions of the body wall which rupture the egg and liberate the embryo still encased in the second embryonic cuticle (vermiform larva; Bernays, 1971). Hatching competence was quantified by testing embryos at various stages between 90 % and 100 % from three pods (percentage develop- ment was calculated from the median length of embryogenesis for the remainder of each pod). Embryos were released from the egg under 24+1 °C saline; if 64 D. BENTLEY AND OTHERS

Table 2. Hatching competence of maturing embryos

95 % stage 90 % stage 100 % stage 90-92-5 92-5-95 95-97-5 97-5-100 Number tested 15 12 20 Number hatching" 0 2 2 16 Percent hatching 0% 13% 17% 80% * Showing rhythmical sequences of peristaltic waves (see text).

Fig. 4. Appearance of live eggs at 5 % developmental intervals during the first half of embryogenesis (eggs partially cleared in sodium hypochlorite). 5-25, ventral aspect; 30-35, dorsal aspect; 40-50, lateral aspect; posterior pole to right. Arrows: 15, embryonic disc; 30-40, posterior margin of embryo; 45, embryo turning within egg; 50, compound eye. Dark-field illumination. Quantitative staging of embryonic development of Schistocerca 65

Fig. 5. Appearance of live eggs at 5 % developmental intervals during the second half of embryogenesis (eggs partially cleared in sodium hypochlorite). 55-60, note that the convex side of the egg has changed from ventral (55) to dorsal (60); this reflects the 180° rotation of the embryo within the egg. Dark-field illumination. contractions did not appear, they could sometimes be elicited by brushing the legs with forceps. Only embryos with the 95 % - 100 % morphology produced rhythmical sequences of peristaltic waves (Table 2), and the behavior occurred much more reliably in embryos near 100 %. Newly hatched larvae normally must dig upward to reach the surface of the ground (Bernays, 1971). Larvae hatching on the surface begin molting the embryonic cuticle (first ecdysis) within 10 min. In the first ecdysis, the cuticular envelope is split along the dorsal midline and the first instar nymph emerges. Emergence is produced by a complex sequence of actions very similar to that described for S. gregaria (Bernays, 1972). As the embryonic cuticle is sloughed, numerous black and unpigmented hairs spring erect, the cervical ampullae deflate, and the mouthparts close medially for the first time. Following ecdysis, 66 D. BENTLEY AND OTHERS

Fig. 6. Appearance of live embryos at 5 % developmental intervals during the first half of embryogenesis. 5, site in yolk droplets where embryonic disc will appear; 10 (arrows), embryonic disc. 35, note amnion (membrane) spread out around the anterior end of the embryo. Transmitted illumination. the nymph is quiescent for ten to fifteen minutes before taking its initial steps. The first instar nymph is bright green with the same brown and black markings as the embryo. Blue pigment persists within the antennae, but fades from the metathoracic femur at the time of hatching. Faint vestiges of the embryonic eye stripes may be visible even after the broad, dark first instar stripe begins to grow down from the dorsal spot (Roonwal, 1947). Quantitative staging of embryonic development of Schistocerca 67

Fig. 7. Appearance of live embryos at 5 % developmental intervals during the last half of embryogenesis. 55, note yolk plug extending anterior and dorsal to embryo (transmitted illumination). 60-100, incident illumination.

DISCUSSION Three types of systems have been used previously for staging grasshopper embryogenesis. Stage has been based upon (1) absolute age of the embryo (age-staging), (2) distinctive changes in external morphology (event-staging), or (3) percentage of the total developmental time through which the embryo has passed (percent-staging). For present purposes, age-staging is unsuitable because of its inflexibility; it cannot be used when the same species is incubated at different temperatures, or for different species, or for species with individual variability in developmental rate. Although the event-staging system may be more useful for comparing similar stages among widely differing species, we 68 D. BENTLEY AND OTHERS

Fig. 8. Appearance of a metathoracic limb of live embryos at 5 % developmental intervals from 25 % to 60 % of embryogenesis. Note that there are marked differences between each stage, particularly involving segmentation, flexion, invagination of apodemes, and differentiation of musculature. Transmitted illumination. Quantitative staging of embryonic development of Schistocerca 69

Fig. 9. Appearance of a metathoracic limb of live embryos at 5 % developmental intervals from 65 % to 100 % of embryogenesis (65-80, transmitted illumination; 85-100, incident illumination). prefer the percentage system for studies of developmental processes for the following reasons: (1) the meaning of a percent-stage is readily apprehended by non-specialists. This greatly facilitates communication with developmental biologists who are not entomologists (in practice, the event-staging systems have been essentially species specific). (2) percent-staging allows greater flexibility in temporal resolution. If more temporal detail is required, it is straightforward to extend a 5 % level to a 1 % level of resolution. This kind of modification is a continuing problem with event-staging systems once the stages have been initially erected. (3) percent-staging allows an even distribution of stages (at whatever level of resolution is required) through the entire developmental 70 D. BENTLEY AND OTHERS

Table 3. Stagings of grasshopper embryogenesis

Genus Reference Schistocerca Jhingran (1947), Shulov & Pener (1963), Micciarelli-Sbrenna (1969), Tyrer (1970) Melanoplus Nelson (1931, 1934), Slifer (1932o), Salt (1949), Riegert (1961) Locusta Roonwal (1936, 1937), Shulov & Pener (1959), Salzen (1960) Chortoicetes Wardhaugh (1978) Ornithacris Chapman & Whitham (1968) Pyrgomorpha Chapman & Whitham (1968) Nomadacris Shulov (1970) Locustana Matthee (1951) Euprepocnemis Khalifa (1957) Dociostaurus Bodenheimer & Shulov (1951) Austroicetes Steele (1941) Aulocara Van Horn (1967) Acrida Kucukeksi (1964) period. The temporal placement of event-stages is dictated by the occurrence of easily recognized events and always results in a non-uniform distribution of stages. Since cellular and sub-cellular processes of great importance in differ- entiation may not be coupled to these easily recognized events, it is better to have a staging system which does not leave gaps in the complete course of embryogenesis. (4) percent-staging permits accurate assignment of elapsed time in developmental processes. Using event-staging, it is possible to order observa- tions but not to discuss the real or relative durations of processes; it is meaning- less to consider elapsed time between two observations unless both are made in the same . A percent system makes it possible to construct the develop- mental history of a process, using data collected from many individuals, with recognition of the time between succeeding steps. The percent-staging system presented here can be correlated with previously described systems (Table 3). Tyrer (1969, 1970) employed a percent-staging system at a 10 % level of resolution for the last 40 % of embryogenesis of S. gregaria. Shulov & Pener (1963) graph the location of their event-stages against percent of the total developmental period for S. gregaria, and Wardhaugh (1978) does the same for Chortoicetes. Chapman and Whitham (1968) and Wardhaugh (1978) have compiled tables showing the percent of development indicated by events in their staging systems, and in several other systems. Tables correlating homologous stages in most grasshopper embryo- genesis event-staging systems have been prepared by Chapman & Whitham (1968) and by Micciarelli-Sbrenna (1969). Age-staging and event-staging are correlated by Shulov & Pener (1959), Salzen (1960), Shulov & Pener (1963), Van Horn (1967), Chapman & Whitham (1968) and Shulov (1970) for several genera. Percent-stages described for one temperature or developmental rate in a Quantitative staging of embryonic development of Schistocerca 71 given species should be accurate for other temperatures or developmental rates. Shulov & Pener (1963) incubated batches of S. gregaria embryos at two different temperatures resulting in embryogenesis times of about 17 and about 50 days. They event-staged embryos developing at these markedly different rates and graphed stage against percent of total developmental time for the two temperatures. The two curves were overlapping throughout the whole period of development, showing that all stages were compressed or expanded propor- tionally according to developmental rate. This result has been confirmed for Locusta (Chapman & Whitham, 1968), Schistocerca (Tyrer, 1970) and Chortoicetes (Wardhaugh, 1978). Therefore, percent-staging should be independent of developmental rate. How applicable would these S. nitens percent-stages be to other species of grasshoppers? They will be inappropriate for species lacking continuous development (diapausing). Among non-diapausing species, Chapman & Whitham (1968) have made an extensive comparison of the percentage of developmental time required to reach a comparable morphological stage for all of the grasshoppers whose embryogenesis has been carefully described. While there appear to be some real differences of small magnitude, they conclude that in general all the non-diapausing species show 'remarkable uniformity' in the percentage of time taken to reach a given stage. Consequently, major adjustments would probably not be necessary to match the S. nitens stages to those of other species. We have placed our stages at 5 % intervals through embryogenesis. What percentage error can be expected in the accuracy with which the described stages match ideal 5 % stages ? The maximum cumulative error which would allow placement to the nearest 5 % would be 2-5 %, or 12 h at the end of a 20-day embryogenesis. There are several sources of error. The first is the accuracy with which the elapsed time of development is known; since egg deposition, when fertilization occurs, and hatching were directly observed (Materials and Methods), the maximum error introduced here is 0-75 h (±0-16 %). A second source of error is the deviation of developmental time of a pod from exactly 20 days. In pod-H, all normally hatching individuals appeared within ± 5 h of 20 days (±1-04 %). Several other pods hatched with slightly longer or shorter developmental times. Examination of embryos from this set of pods is the basis of our estimation of the appearance of true 20-day embryos, and of the difference in appearance introduced by slight deviation from the 20-day period. A third error is that introduced by asynchrony among embryos developing in the same pod. In S. nitens, this error is often less than ±0-5 % of total developmental time (Fig. 1). A similar degree of synchrony has been well documented in other species of grasshoppers (Bodine, 1925; Slifer, 1932a; Salzen, 1960; Shulov & Pener, 1963; Shulov, 1970; Tyrer, 1970; Wardhaugh, 1978). In S. gregaria, Tyrer (1970) reports that for 12 pods examined, 49 % of individuals hatched within 2 h of the first hatchling, and 82 % within 3 h. If our stage descriptions 72 D. BENTLEY AND OTHERS were based on individuals expressing errors of maximum size and the same sign, the cumulative error would be about ± 1-7 %. However, the morphs described were seen repeatedly, making it highly probable that they were near the mean, and not near the error extremes. Consequently, the stage descriptions should lie well within ± 1 % of the ideal 5 % stages (note that the types of errors discussed here are those involved in constructing the staging system, and are not a concern in employing it).

We thank Drs Corey S. Goodman and C. M. Bate for criticism of the manuscript. Support provided by NSF Grant BNS75-03450 and NIH Grant NS-9074-09.

REFERENCES ANDERSON, D. T. (1972). The development of hemimetabolous . In Developmental Systems. Insects, vol. I (ed. S. J. Counce & C. H. Waddington), pp. 96-165. New York, London: Academic Press. BATE, C. M. (1976a). Pioneer neurons in an embryo. Nature, Lond. 260, 54-55. BATE, C. M. (19766). Embryogenesis of an insect nervous system. 1. A map of the thoracic and abdominal neuroblasts in Locusta migratoria. J. Embryol. exp. Morph. 35, 107-123. BENTLEY, D. & TOROIAN-RAYMOND, A. (1979). Embryonic development of the dendritic arborization of an identified insect interneuron (unpublished). BERNAYS, E. A. (1971). The vermiform larva of Schistocerca gregaria (Forskal): form and activity (Insecta, ). Z. Morph. Okol Tiere 70, 183-200. BERNAYS, E. A. (1972). The intermediate moult (first ecdysis) of Schistocerca gregaria (Forskal) (Insecta, Orthoptera). Z. Morph. Okol Tiere 71, 160-179. BODENHEIMER, F. S. & SHULOV, A. (1951). Egg development and diapause in the Moroccan . Bull. Res. Coun. Israel. 1, 59-75. BODINE, J.H. (1925). Effect of temperature on rate of embryonic development of certain Orthoptera. J. exp. Zool. 42, 91-109. CARLSON, J. G. (1961). The grasshopper neuroblast culture technique and its value in radio- biological studies. Ann. N.Y. Acad. Sci. 95, 932-941. CHAPMAN, R. F. & WHITMAN, F. (1968). The external morphogenesis of grasshopper embryos. Proc. Roy. ent. Soc. Lond. {A) 43, 161-169. EDWARDS, J. S. & CHEN, S. (1979). Embryonic development of an insect sensory system, the abdominal cerci of Acheta domesticus. Wilhelm Roux' Arch. devl. Biol. 186, 151-178. GOODMAN, C. S., O'SHEA, M., MCCAMAN, R. & SPITZER, N. C. (1979). Embryonic develop- ment of identified neurons: temporal pattern of morphological and biochemical differentia- tion. Science. 204, 1219-1222. GOODMAN, C. S. & SPITZER, N. C. (1979). Embryonic development of identified neurons: differentiation from neuroblast to neuron. Nature, Lond. 280, 208-214. GRELLET, P. (1968). Milieu de culture pour embryons de gryllides. /. Insect Physiol. 14, 1735-1761. HORRIDGE, G. A. (1977). The compound eye of insects. Sci. Amer. 237, 108-120. JHINGRAN, V. G. (1947). Early embryology of the , Schistocerca gregaria (Forskal) (Orthoptera, ). Rec. Indian Mus. 45, 181-200. JOHANNSEN, O. A. & BUTT, F. H. (1941). Embryology of Insects and Myriapods. New York, London: McGraw-Hill. JONES, B. M. (1956). Endocrine activity during insect embryogenesis. Function of the ventral head glands in locust embryos {Locustana pardalina and Locusta migratoria, Orthoptera). /. exp. Biol. 33, 174-185. KARANDIKAR, K. R. (1942). External structures of the desert locust {Schistocerca gregaria, Forsk). /. Univ. Bombay 11:3B, 1-29. Quantitative staging of embryonic development of Schistocerca 73 KHALIFA, A. (1957). The development of eggs of some Egyptian species of grasshoppers, with special reference to the incidence of diapause in the eggs in Euprepocnemis plorans Charp. Bull. Soc. Entomol. Egypt. 41, 299-330. KUCUKEKSI, S. (1964). Contribution to the study of egg development in Acrida bicolor (Thunb.) (Orthoptera, Acrididae). Communs. Fac. Sci. Univ. Ankara (C) 13, 131-178. MAHOWALD, A. P. (1972). Oogenesis. In Developmental Systems. Insects, vol. i (ed. S. J. Counce & C. H. Waddington), pp. 1-48. New York, London: Academic Press. MATTHEE, J. J. (195.1). The structure and physiology of the egg of Locustana pardalina (Walk.). Sclent. Bull. Dep. Agric. S. Afr. 316, 1-83. MCNABB, J.W. (1928). A study of the chromosomes in meiosis, fertilization, and cleavage in the grasshopper egg (Orthoptera). /. Morph. Physiol. 45, 47-93. MICCIARELLT-SBRENNA, A. (1969). Gli stadi normali di sviluppo degli embrioni di Schistocerca gregaria Forskal (Orthoptera, Acrididae). Boll. Zoo I. 36, 77-95. MICCIARELLI, A. S. & SBRENNA, G. (1972). The embryonic apolyses of Schistocerca gregaria (Orthoptera). /. Insect Physiol. 18, 1027-1037. MUELLER, N. S. (1963). An experimental analysis of molting in embryos of Melanoplus differentialis. Devi Biol. 8, 222-240. NELSON, O. E. (1931). Life cycle, sex differentiation, and testis development in Melanoplus differentialis (Acrididae, Orthoptera). /. Morph. Physiol. 51, 467-525. NELSON, O. E. (1934). Segregation of the germ cells in the grasshopper, Melanoplus differentialis. /. Morph. 55, 545-575. PACKARD, A. S. (1883). The embryological development of the locust. Third Report of the U.S. Entomol. Comm. pp. 263-285. RIEGERT, P. W. (1961). Embryological development of a non-diapausing form of Melanoplus bilituratus Walker (Orthoptera, Acrididae). Can. J. Zool. 39, 491-494. ROONWAL, M. L. (1936). Studies on the embryology of the African Migratory Locust, Locusta migratoria migratorioides R. and F. I. The early development, with a new theory of multi-phased gastrulation among insects. Phil. Trans. R. Soc. Lond. B 226, 391-421. ROONWAL, M. L. (1937). Studies on the embryology of the African Migratory Locust, Locusta migratoria migratorioides R. and F. (Orthoptera, Acrididae). Phil. Trans. R. Soc. Lond. B 227, 175-244. ROONWAL, M. L. (1947). Variation and structure of the eyes in the desert locust, Schistocerca gregaria (Forskal). Proc. R. Soc. Lond. B 134, 245-272. SALT, R. W. (1949). A key to the embryological development of Melanoplus bivittatus (Say), M. mexicanus mexicanus (Sauss.), and M. packardii (Scudder). Can. J. Research. B 27, 233-235. SALZEN, E. A. (1960). The growth of the locust embryo. /. Embryol. exp. Morph. 8, 139-162. SHARAN, R. K. (1958). The embryonic cuticle of Locustana'pardalina. Annls Zoology 3,1-8. SHULOV, A. (1970). The development of eggs of the Red Locust, Nomadacris septemfasciata (Serv.), and the African Migratory Locust, Locusta migratoria migratorioides (R. and F.), and its interruption under particular conditions of humidity. Anti-Locust Bull. 48, 1-22. SHULOV, A. & PENER, M. P. (1959). A contribution to knowledge of the development of the egg of Locusta migratoria migratoriodes. Locusta 6, 73-88. SHULOV, A. S. & PENER, M. P. (1963). Studies on the development of eggs of the desert locust {Schistocerca gregaria Forskal) and its interruption under certain conditions of humidity. Anti-Locust Bull. 41, 1-59. SLIFER, E. H. (1932a). Insect development. IV. External morphology of grasshopper embryos of known age and with a known temperature history. /. Morph. 53, 1-21. SLIFER, E. H. (19326). Insect development. III. Blastokinesis in the living grasshopper egg. Biol. Zbl. 52, 223-229. SLIFER, E. H. & KING, R. L. (1934). Insect development. VII. Early stages in the development of grasshopper eggs of known age and with a known temperature history. /. Morph. 56, 593-601. SNODGRASS, R. E. (1929). The thoracic mechanism of a grasshopper, and its antecedents. Smithson. misc. Collns. 82, 1-111. SPITZER, N. C. (1979). Ion channels in development. Ann. Rev. Neurosci. 2, 363-397. 74 D. BENTLEY AND OTHERS STEELE, H. V. (1941). Some observations on the embryonic development of Austroicetes cruciata Sauss. (Acrididae) in the field. Trans Roy. Soc. Sci. Aust. 65, 329-332. TYRER, N. M. (1969). The time course of contraction and relaxation in embryonic locust muscle. Nature, Lond. 224, 815-817. TYRER, N. M. (1970). Quantitative estimation of the stage of embryonic development in the locust, Schistocerca gregaria. J. Embryol. exp. Morph. 23, 705-718. VAN HORN, S. N. (1967). Studies on the embryogenesis of Aulocara elliotti (Thomas) (Orthop- tera; Acrididae). External morphogenesis. /. Morph. 120, 83—114. WARDHAUGH, K. G. (1978). A description of the embryonic stages of the Australian Plague Locust, Chortoicetes terminifera (Walk.). Acrida 7, 1-9. WHEELER, W. M. (1893). A contribution to insect embryology. /. Morph. 8, 1-160.

{Received 6 March 1979, revised 4 April 1979)