1. In the latter half of the second instar the lymph glands of Drosophila larvae hypertrophy and release cells into the body-cavity, many of which move to the imaginal buds (Shatoury, 1955a). These cells have been referred to as oikocytes. The optic region of the cephalic disk becomes colonized by the large cells known as hexagons. During the first half of the third instar these increase considerably in number. At about the 85-hour stage, and simultaneously with similar changes in the cells still in the lymph glands, the oikocytes in the optic disk exhibit nuclear pycnosis and the deposition of melanin. Later the cells may become vacuolarized and disintegrate.

  2. The precursors of the ommatidia can be seen in the optic disk before the oikocytes become melanized.

  3. During the latter part of the first instar oikocytes are also released from the lymph glands, and hexagons accumulate around the end of the oesophagus where it enters the proventriculus. In the immediate neighbourhood of these oikocytes the radial layer of the proventriculus shows some cellular differentiation, forming an ‘upper ring’ of highly basophilic cells. It has usually been considered that this upper ring constitutes the primordia of the imaginal fore-gut, but it has been found that the lower radial cells will also contribute to this structure.

  4. The oikocytes of the proventriculus suffer melanization and degeneration at the same time as those of the optic disk and the lymph gland.

  5. In larvae aged between 85 and 96 hours two groups of degenerating cells can be found near the cephalic end of the testis and one group in the caudal end. These three groups of cells seem to be oikocytes, and can be traced back to the first instar, when they seem to enter the testis at the same time that the proventriculus is being colonized. Their appearance in the testis is followed by the first sign of the differentiation of spermatocytes.

  6. The appearances suggest that the oikocytes, when they arrive at the testis, the proventriculus and the optic disk, play some part in stimulating those processes of differentiation which bring these organs into the condition from which they can undergo metamorphosis. The simultaneous degeneration of these cells at the end of the third instar would seem to indicate that this stimulus is no longer needed after that time. Shatoury (1956) has suggested that the mechanism which brings about this degeneration involves the formation of auto-antibodies.

A CONSIDERABLE body of information has already been accumulated in which a careful study of the morphological effects of a gene in Drosophila has been used to throw light on the epigenetic processes which bring about development. Much of the earlier work of this kind (e.g. Goldschmidt, Waddington) has dealt with mutant genes which produced abnormal adults. More recently a great deal of attention has been paid to the developmental effects of lethal genes which cause the death of the individual before the adult stage is reached. In a recent monograph on this category of genes, Hadorn (1951), who has been one of the most active workers in this field, lays considerable stress on what he calls the phase specificity of the lethals, that is, on the fact that individuals homozygous for a particular lethal usually die at some rather definitely defined stage of their life history. These phases at which death supervenes are perhaps more a matter of the clinical medicine of Drosophila than an aspect of its fundamental biology. There are undoubtedly certain times in the life history at which a particularly complicated and ticklish set of processes have to be carried out if the animal is to survive. They have been spoken of as epigenetic crises (Waddington, 1941). Developmental alterations at some earlier time may not have any very drastic effects on the viability of the organism until such a crisis arises, when they may give rise to gross abnormalities which bring about the animal’s death. The fact that many lethals may be found to kill their bearers at the same phase (for instance at the metamorphosis from the larva into the prepupa) does not necessarily mean that the primary action of all these lethals has the same phase specificity. They may have begun to produce slight abnormalities at a much earlier stage. It is, indeed, by the investigation of these early stages of gene action that we may hope to throw light on the nature of the epigenetic crises and to discover the factors involved in them. One of the most important foundations for studies of this kind is a thorough knowledge of the development during the larval stages of the rudiments which will, after metamorphosis, give rise to the imaginal organs.

Shatoury (1955 a, b, c) has recently directed attention to the importance of the lymph glands in this connexion. Towards the end of both the second and third larval instars these glands become hypertrophied and eventually rupture, releasing into the haemocoele a number of freely dispersed cells. Histological evidence suggests that, at the end of the second instar, many of these cells accumulate around the imaginal buds and give rise to the imaginal mesoderm of the buds. This suggestion is confirmed by the fact that in one of the lethals (lethal nodifferentiation) studied by Shatoury abnormalities among the early third instar lymph glands are followed by a pathological behaviour of this imaginal mesoderm. In another lethal (lethal malignant) a considerable hypertrophy of the glands gives rise to an excessive number of free cells which behave in a way reminiscent of a malignant tumour. The present paper deals with certain other groups of cells which appear in normal wild-type larvae, being apparently derived from the lymph glands and taking part in the differentiation during the larval period of the future imaginal rudiments. The cells with which we shall be concerned enter into the formation of the cephalic imaginal bud, the foregut (proventriculus), and the testes. A later paper will deal with the fate of similar cells in the mid-gut or stomach. The evidence that these cells are derived from the lymph glands comes partly from the fact that they appear in the various organs shortly after the release of cells from these glands, partly from the histological similarity between them and the cells within the glands, and partly from the fact that in their later development they show histological changes which parallel exactly changes also proceeding in the lymph glands themselves. Shatoury has described the lymph gland cells as falling into three main classes— rather large cells known as hexagons, and two smaller types known as spheroids and platelets. It is convenient to have some name for the whole group of cells which, emerging from the lymph glands, move around the body and settle in particular localities. We shall refer to them as oikocytes (derived from the Greek oikizo, ‘to found a colony’).

The work to be described is based on the investigations of a wild-type OrK stock which had been inbred for a large number of generations and was obtained from the Institute of Animal Genetics, Edinburgh. Material was fixed at two-hourly intervals starting from 6 hours before the embryos hatch until the beginning of the pupal stage. Where accurate timing near the beginning of an instar was necessary, larvae were watched under the binocular and the time of moulting observed directly. Boiling Bouin was used as a fixative, and larvae were treated by Peterfi’s method of double embedding. Prepupae and pupae were dissected out from their puparia and were treated by the amyl acetate technique. Sections ranged from 2 to 5μ and were stained with haematoxylin, considerable care being devoted to appropriate differentiation of the stain.

Oikocytes in the cephalic imaginal bud

At roughly 5 hours before the end of the second instar, the first and only defined pair of glands hypertrophies and releases its cells into the body cavity. It has already been pointed out (Shatoury, 1955b) that some of these cells form the imaginal disk mesoderm, whose possible influence on the process of differentiation of the buds, leading to a segmentation pattern characteristic of each, was discussed.

No differential selection of the migrating lymph gland cells by the various imaginal buds was observed in the previous investigations. However, the present study has shown that all the imaginal buds, except that portion of the cephalic complex which later differentiates into the optic disk, receive the same type of cells, which are either spheroids or platelets, or possibly a mixture of both. The mesoderm cells of the presumptive optic bud are of an entirely different type, as they seem to be made up of hexagons. In the other buds the hexagons come to lie either on the peripodial membrane or somewhere in its vicinity.

Till shortly before the end of the second instar the cephalic complex is similar in structure to any other imaginal bud, being composed of a simple epithelium flanking a peripodial cavity which is surrounded by a thin peripodial membrane (Plate 1, fig. A). As shown in the figure, the migrating mesoderm cells accumulate in two groups. One lies underneath the presumptive eye-disk portion and the other lies on the opposite side. Three hours before the beginning of the third instar the cephalic bud becomes differentiated into an antennal and eye segment. Fig. B shows that during this process of differentiation the hexagons lying on the peripodial membrane have divided the cephalic bud into an anterior cylindrical segment and a posterior oval one, representing the antennal and eye-bud respectively. Although these appearances have been overlooked, Chen (1929), Steinberg (1943), and others have already reported that the cephalic complex remains undifferentiated till about the end of the second instar.

At the beginning of the third instar the cephalic complex is distinctly differentiated into its two major components, which still remain connected by a thin membrane. The hexagons of the optic bud proliferate and the oval posterior side of the bud becomes concave (Plate 1, fig. C). A new cavity, which is occupied by the multiplying hexagons, eventually appears. The increase in the size of the optic bud is accompanied by a rapid increase in the number of the hexagons so that later stages (up to the 80-hour stage) show a considerable number of them.

At the 85-hour stage, and simultaneously with the melanization of the hexagons in the lymph glands (Shatoury, 1955 a, b), those in the optic bud behave similarly; this is shown in Plate 1, fig. D, where the optic hexagons have become deeply melanized. In older larvae they become vacuolated and finally disintegrate, and appear later as a homogeneous undifferentiated mass. Melanin deposition was shown by Shatoury to be a post-mortem event, perhaps brought about by a complement reaction.

Before melanin deposition has taken place the optic bud is seen to have become differentiated into two major segments, a head segment which metamorphoses to the adult cephalic integument, and an eye segment (fig. D). The latter in turn is sub-differentiated into three layers and it is in the uppermost of these that the precursors of the ommatidia first appear. Literature on the time of the onset and completion of this process is still in a confused state. Krafka (1924) and Chen (1929) reported that this differentiation begins about 24 hours before pupation. Medvedev (1935) mentioned that the ommatidia could be seen in whole mounts of optic disks of mature larvae. On the other hand, Bodenstein (1938) and Robertson (1936) claim that the ommatidia begin to differentiate 36 hours after puparium-formation. Enzmann & Haskins (1938) state that it starts 18 hours after egg-laying, but this has been proved wrong by Steinberg (1941). According to Pilkington (1942), the optic disk of a 4-day-old larva consists of undifferentiated cells several layers thick and apparently homogeneous. This contradicts Steinberg’s report (1941) that the eye-disk begins to organize into clusters of four cells each at about 72 hours after hatching and that the clusters increase remarkably in both number and clarity in the 96-hour stage. The present investigations revealed that before the optic hexagons have disintegrated the eye segment of the optic bud has completely differentiated into ommatidia precursors. Further studies on the differentiation of these structures during the larval and post-larval stages are being carried out.

Oikocytes in the proventriculus

The hypertrophy of the lymph glands, followed by their rupture and the release of oikocytes into the haemocoele, occurs not only in the latter part of the second instar but at a similar stage in the first instar larva. Some of the cells released at this stage find their way into the proventriculus.

The oesophagus of the Drosophila larva is a narrow simple tube which passes between the brain hemispheres into the bulb-like proventriculus. In newly hatched larvae (Plate 2, fig. K), apart from the oesophageal component, the proventriculus is a compound structure composed of two walls. There is an outer layer of large cells of an intense affinity to basic dyes, which is regarded as an extension of the stomach. The inner layer is made up of faintly stained bow-like cells whose borders are poorly defined; the broad ends of the cells join, with the result that oval inter-cellular spaces occur. The nuclei are rectangular, and lie in the narrow central region of the cells with the major sides almost fused with the nearby cell-wall. The cytoplasm is reticular in texture, with thin strands connecting the nuclei with the cell-membrane. The internal structure of the nuclei is also important for our investigations. Nucleoli appear as rounded bodies which lie very close to the nuclear membrane. There are deeply stained granules of considerable size restricted to the periphery of the nuclei. This inner layer of cells is referred to here as the radial layer, and is usually regarded as an évagination of the oesophagus.

The oesophagus in the region where it joins the proventriculus is encircled by a ring of rather large cells whose function is uncertain; they were first described by Weismann and were named Weismann’s cell chaplets or wreath cells. Investigations of these cells in first instar material shows that they are amoeboid, and are grouped in a loose sheet of cells embedded in a homogeneous matrix. It seems that this tissue receives a nerve-supply from the adjacent hypocerebral ganglion which lies somewhat at the anterior side of the proventriculus.

When the hypertrophied lymph glands release their cells into the body-cavity towards the end of the first instar, changes occur within the mass of wreath cells whereby they leave the matrix and wander about near by. As a result of this migratory movement the anterior end of the proventriculus becomes exposed in the body cavity. Lymph gland cells then accumulate in this region and penetrate between the wall of the oesophagus and the radial layer (Plate 2, fig. L). A narrow cylindrical cavity is thus formed between these two layers of cells. Figure L clearly illustrates that the migrating cells, while creeping through the newly formed cavity, become flat and tightly compressed.

In the selection of the hexagons and the rejection of the spheroids and platelets an important role seems to be played by the phagocytic activity of the wreath cells. There is other evidence that these cells have a scavenging function. For instance in larvae kept on food containing silver nitrate they become heavily charged with opaque particles. It seems possible that the reason why it is the hexagons rather than the other lymph gland cells which colonize the proventriculus is simply that the hexagons are the largest type of cell and therefore the most difficult for the wreath cells to ingest.

As soon as the hexagons establish themselves in the proventriculus (i.e. approximately 2 hours before the beginning of the second instar), the radial layer shows signs of gradual differentiation into two distinguishable regions. A few cells in the uppermost part aggregate, with the eventual loss of both the intercellular spaces and the original bow-like outline. They also differ from the rest of the cells in that the cytoplasm has lost its reticular consistency and has become more positive to basic dyes than that of the other radial cells, which remain entirely non-basophilic.

By the end of the second instar the upper cells have become sharply differentiated into a compact ring of a larger number of cells with a stronger basophilic reaction (the ‘upper ring’). Nuclei are more positive to the stain than the cytoplasm and the individual cells have become spindle-shaped. The rest of the cells, which comprise the largest part of the radial layer, have also undergone changes, which in the main affect the nuclei; the basophilic granules have increased tremendously in number and instead of being restricted to the periphery, as in early first instar individuals, have become equally distributed throughout the whole of the nuclear material. The cytoplasm of these cells remains negative to haematoxylin.

It is not easy to decide whether the oikocytes which originated from late first instar lymph glands continue with differentiation processes during the third instar, or whether they die and are replaced by others derived from late second instar organs. However, microscopic examination suggests that they may undergo degenerative changes at the end of the second instar and that a fresh supply of hexagons from the hypertrophied second instar glands takes their place.

Although the hexagon component of the proventriculus is not particularly striking till the late phases of the second instar, it becomes a remarkable feature in the very early hours of the third. Plate 2, fig. M is from a 10-minute-old third instar larva and shows with clarity the exact location of the hexagons. It can also be seen that the upper ring of cells is proceeding to further differentiation; the cell number has increased and cell-walls have become more defined. Furthermore, the nuclei show a stronger basic reaction than the cytoplasm.

Successive stages in differentiation up to the 80-hour stage showed that the upper ring is undergoing progressive histological differentiation, with the nuclei still retaining a strong basophilic reaction. The remaining cells of the radial layer have also changed, in that the nuclei have become more organized and deeply stained. At the 85-hour stage the radial layer is clearly differentiated into the two types of cells. From now until the end of the third instar no further differentiation was observed in the radial layer.

After the differentiation of the radial layer has reached a state of completion, the hexagon cells of the proventriculus undergo changes of the same kind as those of the optic bud, that is, they become melanized (Plate 2, fig. N). This change is exactly simultaneous with the melanization of the optic bud hexagons and the similar cells remaining in the lymph glands. In later stages the hexagons of the proventriculus are displaced into the haemocoele (fig. O) where they disintegrate.

It has been believed by most previous investigators (Kowalevsky, 1887, on Muscids; Strasburger, 1932; Robertson, 1936, on Drosophila) that the imaginal fore-gut primordia is only represented by the upper ring. As a matter of fact, the remainder of the radial cells also take part in the formation of the imaginal organ.

At the beginning of metamorphosis the lower radial cells, which make up the greater part of the proventriculus, become attached by their inner ends to the wall of the oesophagus. At the white puparium stage the wall of the oesophagus becomes convoluted with the result that this organ contracts to less than half its original length. This process mechanically everts the radial layer of cells. Eversion of the two component primordia of the layer does not happen simultaneously but, owing to the fact that the lower cells have become attached to the oesophagus, they will be the first to be dragged forward (Plate 2, fig. P). Furthermore, in view of their characteristic pattern of distribution, it is natural to find that the everted cells come to lie side by side, forming a cylinder with their nuclei occupying its equator. Half-way through eversion, the cells of the main body of the radial layer have moved outside the proventriculus to constitute a thick ring (fig. Q). At this stage the upper ring has not been markedly affected by the process of eversion, but later it also is pulled out by the contracting oesophagus. When the process is completed both rings of cells form a compound waisted structure (fig. R) which is made up of two parts; an anterior region originating from the metamorphosis of the originally lower cells and a posterior region representing the everted upper ring of cells. Cellular proliferation of these ‘imaginal’ cells of the upper ring, as described by Strasburger and Robertson, has not been observed. This constricted structure will later metamorphose to the adult cardia.

Oikocytes in the testis

In the newly hatched larva of Drosophila the testis is undifferentiated and is only composed of spermatogonia. At the end of the first instar, which corresponds, according to the present investigation, to a few hours after the lymph glands have released their cells, the testis shows faint signs of differentiation to apical cells, spermatogonia, primary spermatocytes, and terminal cells. This remains the condition of the larval testis until shortly before pupation (Kerkis, 1933; Sonnenblick, 1941; Gloor, 1943; Geigy & Aboim, 1944).

In lethal malignant larvae, besides the principal malignant condition due to cells released from the lymph glands, Shatoury (1955b) has described a testis tumour tissue which invades the surrounding germinal layers, bringing about necrosis. In discussing the nature and origin of these cells, the hypothesis was suggested that they probably arise from cells of a pre-malignant type which lie at both ends of the gonads, which had been overlooked in previous investigations of its structure. A search for these hypothetical elements has now been made.

In larvae up to the 75-hour stage, preliminary efforts to identify them amongst the germ-tissue were unsuccessful. This was attributed to a possible equal stainability of these cells and the surrounding germ-cells. However, we have seen that the oikocytes in the proventriculus and optic disk exhibit a series of characteristic changes (pycnosis, melanization, or vacuolization) in the later stages of the third instar. It seemed possible that if similar cells occurred in the testis they would make their presence apparent at the same stage by processes of this kind. This was found to be the case. Plate 1, fig. I, which was taken from an 85-hour-old larva, shows two isolated groups of cells at the cephalic end of the testis in the neutral zone between the spermatogonia and the primary spermatocytes, which have become very striking in respect to their stronger affinity to basic stains, probably due to early symptoms of pycnotic degeneration. In 96-hour-old larvae (fig. J) these groups of cells have undergone lysis, so that in the prepupal testis their original place becomes represented by cavities. Similar cells (of undetermined number) with the same type of reaction were also easy to locate in the region between the posterior terminal cells and the primary spermatocytes. These three groups of cells appeared consistently in the hundreds of preparations investigated.

The exact location of these cells having been accurately determined, histological examination of them in stages as early as late first instar were carried out. Plate 1, fig. H was taken from a stage slightly earlier than fig. I and represents one of these groups of new cells in the spermatogonial region before it shows the type of change just described. As shown, the cells are morphologically distinguishable from the surrounding ones. Figure G was obtained from a late first instar larva after the lymph glands have released their cells. It shows clearly two groups of three cells which are different from the rest of the spermatogonia not only from the morphological point of view but also from their cytochemical reaction towards basic stains. These groups of cells are located somewhat terminally, occupying both a cephalic and a caudal position. Prior to this stage lymph gland cells are found to accumulate in the region of the testis, probably suggesting that these cells are derived from migrating lymph gland material.

In the proventriculus, the optic imaginal buds, and the testis the investigations just described have revealed a very similar sequence of rather peculiar phenomena. At a relatively early stage in larval life, groups of cells, which seem to be identical with the hexagons of the lymph glands, make a sudden appearance. In the testis and proventriculus this happens during the late first instar, in the optic bud, during the latter half of the second instar. In both cases the cells appear just after the lymph glands have hypertrophied and released their contents into the body-cavity. Subsequent to the appearance of the cells in the organs, signs begin to be visible of the changes which bring the imaginal rudiments into the state of readiness in which they await the onset of metamorphosis. Thus in the testis there is the beginning of a differentiation of spermatogonia into early spermatocytes. In the proventriculus the upper ring of radial cells acquires a more solid consistency and a higher basophilia. In the optic bud the cells of the uppermost layer of the epithelium become arranged into the precursors of the ommatidia. There is a clear suggestion that the oikocytes act in some way as stimulators to further development.

One might perhaps be tempted to compare them to the organizers so well known in vertebrates, but it must be pointed out that the evidence suggests rather that they stimulate development to proceed farther along a course already laid down for it, rather than deciding a choice between alternative types of development as in the case of vertebrate organizers. If one were to seek an analogy for them among the entities known to play a part in early embryonic development it is perhaps in the well-known ‘formation centre’ (Bildungszentrum) described by Seidel in Platycnemis that the closest parallel might be found. It would, however, be premature to press such analogies too far, since it is clear that much remains to be learnt about the stimulating action which appears to be exerted by these cells. It seems technically impossible to investigate their action by the normal experimental embryological techniques of extirpation and grafting. The most hopeful line of attack would appear to be the discovery of lethal mutations which interfere with the migration of the hexagons to one or more of the organs involved. If the above hypothesis is correct, an organ to which the hexagons fail to arrive should not show any further histological differentiation during the second and third instars but should retain the character which it had during the first. Some evidence which supports this suggestion will be described in a later paper (Shatoury & Waddington, 1957).

The results of the transplantation experiments of several workers might appear to contradict the above hypothesis. For example, Steinberg (1943) transplanted early second instar cephalic buds into late second instar hosts and reported that the transplants differentiated normally. At the time of transplantation these buds were probably free of mesoderm. There is, however, no reason why mesoderm should not have been supplied to them by the lymph glands of the host larvae, and thus their behaviour provides no evidence that their differentiation is possible in the absence of the mesoderm.

It is interesting to note that immediately following the presumed inductive action of the hexagons, the tissue which reacts to them exhibits an increase in cytoplasmic basophilic granules. This reminds one of the phenomena described by Kedrowski (1937) and Brachet (cf. 1950) in the amphibian embryo. According to both these authors, these basophilic granules increase during histological differentiation and later decrease again as that process is completed. In Drosophila the imaginal primordia generally retain their basophilia until the beginning of metamorphosis, but it can be observed that the cytoplasmic granules disappear entirely while the pupal differentiation of the primordia is proceeding. Cytochemical tests are now being made to discover whether these transient granules show the characteristics of sulphhydryl ribonucleoproteins as do those described by Brachet.

One of the most remarkable features in the behaviour of the oikocytes is their disappearance towards the latter part of the third instar. They suffer a series of degenerative changes which may include pycnosis and vacuolization and often the deposition of melanin. These changes occur simultaneously with similar phenomena which affect the hexagons in the lymph glands themselves. Shatoury (1956), arguing from phenomena seen in certain lethal stocks, has suggested that the degeneration may be due to the action of antibodies formed in relation to the oikocytes which become antigenic at this stage. It must remain for future work to decide whether this hypothesis of autoantigenicity will eventually be found valid. In the meantime it provides a stimulating working hypothesis. Whatever the mechanism may be by which the oikocytes are caused to degenerate, it is surely noteworthy that this happens after the imaginal rudiments have completed the whole of that part of their differentiation which they carry out in the larval period and have reached a state of readiness to undergo the more profound changes which will occur on metamorphosis.

This work was done while one of us (H. H. El S.) held a research studentship from the Egyptian Government. The work received financial support from the Agricultural Research Council, for which we are grateful.

Bodenstein
,
D.
(
1938
).
Untersuchungen zum Metamorphoseproblem. 1. Kombinierte Schnürungsund Transplantationsexperimente an Drosophila
.
Roux Arch. EntwMech. Organ
.
137
,
474
505
.
Bracket
,
J.
(
1950
).
Chemical Embryology
.
New York, London
:
Interscience
.
Chen
,
T. Y.
(
1929
).
On the development of the imaginal buds in normal and mutant Drosophila melanogaster
.
J. Morph
,
47
,
135
99
.
Enzmann
,
E. V.
, &
Haskins
,
C. P.
(
1938
).
The development of the imaginal eye in the larvae of Drosophila melanogaster
.
J. Morph
.
63
,
63
72
.
Geigy
,
R.
, &
Aboim
,
A. N.
(
1944
).
Gonadentwicklung bei Drosophila nach frühembryonaler Ausschaltung der Geschlechtszellen
.
Rev. suisse Zool
.
51
,
410
17
.
Gloor
,
H.
(
1943
).
Entwicklungsphysiologische Untersuchungen an den Gonaden einer Letalrasse (igl) von Drosophila melanogaster
.
Rev. suisse Zool
.
50
,
339
94
.
Hadorn
,
E.
(
1951
).
Developmental action of lethal factors in Drosophila
.
Advanc. Genet
.
4
,
53
85
.
Kedrowski
,
B.
(
1937
).
Uber die sauren (elektro-negativen) Kolloide des Protoplasmas. Studien an Larven von Rana temporaria
.
Z. Zellforsch
.
26
,
21
35
.
Kerkis
,
J.
(
1933
).
Development of gonads in hybrids between Drosophila melanogaster and Drosophila simulons
.
J. exp. Zool
.
66
,
477
509
.
Kowalevsky
,
A. Z.
(
1887
).
Beiträge zur Kenntnis der nachembryonalen Entwicklung der Musciden (Calliphora)
.
Z. wiss. Zool
.
45
,
542
94
.
Krafka
,
T.
, Jr.
. (
1924
).
Development of the compound eye of Drosophila melanogaster and its bar-eyed mutants
.
Biol. Bull. Wood’s Hole
,
47
,
143
8
.
Medvedev
,
N. N.
(
1935
).
Genes and development of characters. 1. The study of the growth of the imaginal discs of eyes of the wild type larvae and the three mutants lobe, glass and ‘eyeless’ in Drosophila melanogaster. Z. indukt. Abstammu
.
VererbLehre
,
70
,
55
72
.
Pilkington
,
R. W.
(
1942
).
Facet mutants of Drosophila
.
Proc. zool. Soc. London, A
,
111
,
199
222
.
Robertson
,
C. W.
(
1936
).
Metamorphosis of Drosophila melanogaster, including an accurately timed account of the principal morphological changes
.
J. Morph
.
59
,
351
99
.
Shatoury
,
H. H. EL
(
1955a
).
The structure of the lymph glands of Drosophila larvae
.
Roux Arch. EntwMech. Organ
.
147
,
489
95
.
Shatoury
,
H. H. EL
(
1955b
).
A genetically controlled malignant tumour in Drosophila
.
Roux Arch. EntwMech. Organ
.
147
,
496
522
.
Shatoury
,
H. H. EL
(
1955c
).
Lethal no-differentiation and the development of the imaginal discs during the larval stage in Drosophila
.
Roux Arch. EntwMech. Organ
.
147
,
523
38
.
Shatoury
,
H. H. EL
(
1956
).
An immunological hypothesis of tumour development in Drosophila
.
Roux Arch. EntwMech. Organ
.
148
,
391
402
.
Shatoury
,
H. H. E.
&
Waddington
,
C. H.
(
1957
).
The development of gastric tumours in Drosophila larvae
.
J. Embryol. exp. Morph
.
5
,
143
52
.
Sonnenblick
,
B. P.
(
1941
).
Germ cell movements and sex differentiation of the gonads in the Drosophila embryo
.
Proc. nat. Acad. Sci., Wash
.
27
,
484
9
.
Steinberg
,
A. G.
(
1941
).
A reconsideration of the mode of development of the Bar eye of Drosophila melanogaster
.
Genetics
,
26
,
326
46
.
Steinberg
,
A. G.
(
1943
).
The development of the wild type and Bar eyes of Drosophila melanogaster
.
Cañad. J. Res
.
21
,
277
83
.
Strasburger
,
M.
(
1932
).
Bau, Funktion, und Variabilität des Darmtraktus von Drosophila melanogaster Meigen. Z
.
wiss. Zool
.
140
,
539
649
.
Waddington
,
C. H.
(
1941
).
The pupal contraction as an epigenetic crisis in Drosophila
.
Proc, zool. Soc. Lond. A
,
111
,
181
8
.

Abbreviations: AS, antennal segment; BR, brain; CHX, cytolysing hexagons; CLGT, cytolysing lymph gland cells; DE, epithelium of imaginal region; DLC, displaced larval gastric epithelium; ES, eye-segment of bud; FO, oikocyte of fore-gut; HP, primordium of imaginal hind-gut; HS, head segment; LC, lower half of cardia; LFP, lower cells of fore-gut primordium; LGT, lymph gland cells; ME, imaginal mid-gut epithelium; MHX, melanized hexagon; OES, oesophagus; OHX, ‘Hexagons’ of optic segment; OP, outer layer; os, optic segment; PPM, peripodial membrane; RL, radial layer; SG, salivary gland; SGN, cellular masses resulting from hypertrophy of salivary gland imaginal primordium; SP, spermatogonia; SPC, primary spermatocytes; uc, upper half of imaginal cardia; UFP, upper ring of cells; wc, ‘Wreath cell’ of Weismann’s cell chaplet.

Plate 1

FIG. A. Late second instar cephalic bud. ×240.

FIG. B. Very late second instar cephalic bud, showing the differentiation of the antennal segment (AS) and optic segment (os), ×240.

FIG. C. Five-hour-old third instar bud, showing oikocytes underlying the optic region, x 440. FIG. D. Section of brain and optic bud in late third instar, showing melanization of hexagons, × 400.

FIG. E. Section of hind-gut, mid third instar.

FIG. F. Late second instar, section through anterior end of salivary gland, ×180.

FIG. G. Late first instar testis. Note that the bulk of the testis shows no differentiation into spermatogonia and spermatocytes, ×130.

FIG. H. Third instar, 80 hours; cephalic region of testis, × 400.

FIG. I. Eighty-five-hour testis, showing two groups of pycnotic lymph gland cells near cephalic end. × 220.

FIG. J. Ninety-six-hour-old testis, showing cytolysing lymph gland cells (CLGT). ×240.

Plate 1

FIG. A. Late second instar cephalic bud. ×240.

FIG. B. Very late second instar cephalic bud, showing the differentiation of the antennal segment (AS) and optic segment (os), ×240.

FIG. C. Five-hour-old third instar bud, showing oikocytes underlying the optic region, x 440. FIG. D. Section of brain and optic bud in late third instar, showing melanization of hexagons, × 400.

FIG. E. Section of hind-gut, mid third instar.

FIG. F. Late second instar, section through anterior end of salivary gland, ×180.

FIG. G. Late first instar testis. Note that the bulk of the testis shows no differentiation into spermatogonia and spermatocytes, ×130.

FIG. H. Third instar, 80 hours; cephalic region of testis, × 400.

FIG. I. Eighty-five-hour testis, showing two groups of pycnotic lymph gland cells near cephalic end. × 220.

FIG. J. Ninety-six-hour-old testis, showing cytolysing lymph gland cells (CLGT). ×240.

Plate 2

FIG. K. Proventriculus of newly hatched larva. ×220.

FIG. L. Late first instar proventriculus, × 440.

FIG. M. Ten-minute-old third instar; note large number of fore-gut oikocytes. × 220.

FIG. N. Late third instar, × 190.

FIG. O. Late third instar. Displacement of melanized hexagons into body-cavity, × 170.

FIG. P. Prepupa. Early stage in eversion of fore-gut primordium. × 330.

FIG. Q. Later stage in eversion of primordium. × 330.

FIG. R. Completion of eversion of fore-gut primordium. ×c. 130.

Plate 2

FIG. K. Proventriculus of newly hatched larva. ×220.

FIG. L. Late first instar proventriculus, × 440.

FIG. M. Ten-minute-old third instar; note large number of fore-gut oikocytes. × 220.

FIG. N. Late third instar, × 190.

FIG. O. Late third instar. Displacement of melanized hexagons into body-cavity, × 170.

FIG. P. Prepupa. Early stage in eversion of fore-gut primordium. × 330.

FIG. Q. Later stage in eversion of primordium. × 330.

FIG. R. Completion of eversion of fore-gut primordium. ×c. 130.