ABSTRACT
Although considerable progress has been made with the chemistry of juvenile hormone (Dahm, Roeller & Trost, 1968), studies on its mechanism of action in immature insects are still in a preliminary stage. Much of the recent work has been interpreted as showing an effect of juvenile hormone on the morphogenetic program through which an insect passes in the course of its ontogeny (Williams, 1961). It is the purpose of this paper to describe three studies which illustrate the complex nature of this developmental program in saturniid moths.
MATERIALS AND METHODS
Experimental animals
The saturniids (Antheraea polyphemus, Samia cynthia and Hyalophora cecropia) used in the present study were reared or purchased from dealers in the United States and England. Staging of animals was carried out by examining the state of the epidermis and the differentiation of adult structures through the pupal cuticle as described by Schneiderman & Williams (1954). Partially purified extracts of juvenile hormone (Williams, 1956) were injected into polyphemus pupae shortly before the onset of adult development. The amount injected varied with the preparation and purification of hormone used. Pupae which receive extra, active corpora allata or juvenile hormone at the onset of adult development, moult into another pupal instar or a form with some pupal and some adult characteristics. All pupal-adult intermediates which had a pupa-like abdomen and a score of + + + or higher in the Williams assay (Williams, 1961) will hereafter be designated as deuteropupae instead of the more confusing term ‘second pupae’ which has been used previously.
The pupa-deuteropupa transformation is more rapid than the pupa-adult transformation (Gilbert & Schneiderman, 1960). Thus, in this study, deutero-pupae have been compared with pharate adults as well as with adult moths because any reduction in size or maturity of an organ in the deuteropupa may be due to an insufficient growth period and not to a more fundamental difference in the hormonal environment (see Williams, 1968).
Histological preparations
Whole mounts of cuticle were prepared by removing the adhering scales and epidermal cells, dehydrating in ethanol, clearing in xylene and mounting on a slide in Canada balsam or ‘HSR’. When the epidermis was to be studied in whole mounts, it was fixed and stained with Feulgen or hematoxylin prior to mounting.
Cuticle and compound eyes were fixed in Bouin’s or Zenker’s fluid; passed successively through ethyl cellosolve, methyl benzoate, and benzene; embedded in paraffin wax or ‘Paraplast’; and sectioned at 6 μ. Cason’s stain (Cason, 1950) was principally used. The eyes of six polyphemus deuteropupae and several normal polyphemus and cynthia pupae and developing adults were examined in serial section. .
Autoradiography was performed on sections cut from tissues generously pro-vided by Dr Blair Bowers from animals used in previous work (Bowers & Williams, 1964) and their methods were followed.
RESULTS
The presence of ‘larval’ tubercles on untreated pupae
The fifth-instar cecropia larva is elaborately decorated with orange, yellow, and blue tubercles. A remnant of these larval tubercles can be found in all cecropia pupae in the form of cells which correspond in position and color to the tubercles of the larva. These colored cellular patches are especially obvious in the newly moulted, untanned pupa but small clusters of the blue cells have also been observed in fully tanned, diapausing pupae.
Occasionally cecropia pupae are found which bear raised knobs of brown pupal cuticle in the same position in each segment as is occupied by a tubercle in the larva. This resynthesis of a larval character in a pupa was seen in a most dramatic form in a batch of cecropia purchased from Mr S. E. Ziemer, a Wisconsin dealer, in 1959. Fourteen per cent of his 491 pupae possessed tubercles, and in this respect resembled a larval-pupal intermediate. The abnormal pupae varied from those which had a few brown tubercles to some which had a full complement of pigmented tubercles.
These protuberances were an integral part of the heavily sclerotized pupal cuticle, for after the formation of the moth they were present on the pupal exuviae. From a study of the cast skins of the final larval instar recovered from the cocoons of the abnormal pupae, it was evident that these larvae had had tubercles identical to those found on normal final instar animals.
There was no obvious cause of these abnormalities. No evidence for any simple genetic basis was provided by the equal distribution of the abnormality between the sexes or by a cross between one pair of adults reared from abnormal pupae. In response to a questionnaire, Mr Ziemer could report no conditions of diet or environment which might have been responsible for this phenomenon.
Pupae with’larval’tubercles have subsequently been produced experimentally, either by infecting larvae with Nosema (C. M. Williams, personal communica-tion), a microsporidian which produces juvenile hormone (Finlayson & Walters, 1957; Fisher & Sanborn, 1962), or by injecting larval tubercles with juvenile hormone or active corpora allata (Staal, 1967). Thus, although we have no direct evidence, it seems reasonable to suspect that the abnormal Ziemer pupae were caused by Nosema infection, and thus indirectly by an excess of juvenile hormone.
Many of the adults which emerged from the tubercle-bearing pupae had one significant abnormality: minute scaleless patches on the dorsal side of the abdomen in the same location as the tubercles of the pupae. No adult from a normal pupa has ever been observed to have such scaleless areas. A microscopic examination of whole mounts of the adult cuticle of these moths revealed that some of these patches consisted of scaleless thickened, tanned ‘pupal’ cuticle surrounded by a thin adult scale-bearing cuticle (Fig. 4D). In one case a heavily tanned tubercle-like protuberance was found on the adult cuticle of an animal which had had a complete set of tubercles in the pupal stage.
Development of the compound eye under the influence of juvenile hormone
Detailed descriptions of the histology and developmental significance of the various parts of the lepidopteran compound eye are available in the literature (Umbach, 1934; Wolsky, 1949). I have examined the histology of the eyes of normal diapausing and adult cynthia and polyphemus, and pharate adults of cynthia on days 3, 7, 8, 11, 17 and polyphemus on day 13 after the initiation of adult development.
The normal pupal eye
In Lepidoptera, bilateral crescents of smooth cuticle exhibiting rudimentary facets are located on the head of the pupa and are partially or completely covered by the antennae. These crescents face outward and are limited on their antero-medial borders by a dark unfaceted furrow (Fig. 3 A). In the diapausing pupa, the cells which underlie the crescent near the furrow are already organized into clusters of undifferentiated cells, each cluster destined to form an individual ommatidium (Umbach, 1934) (Fig. 1A). The rest of the presumptive eye is present as a multilayered epithelium which lies anterior to the crescent and which is not organized into cell clusters (Wolsky, 1949).
Normal eye development A. Diapausing polyphemus pupa. Cross-section of the eye showing the cell clusters (arrows) which give rise to individual ommatidia. B. Cynthia: third-day adult development. The crystalline cones are forming. C. Autoradiograph of a cynthia eye on day 2 of adult development. The arrow indicates the silver grains at the border of the eye. The region to the left of the silver grains had begun differentiating. D. Polyphemus adult. Cross-section through rhabdoms with surrounding tracheoles of tapetai layer. E. Cynthia: seventh day of adult development. The crystalline cone vacuoles have fused. The arrow designates the groove which delineates the edge of the eye. F. Polyphemus: thirteenth day of adult development. The eye has the adult con-figuration but has not yet reached full size.
Normal eye development A. Diapausing polyphemus pupa. Cross-section of the eye showing the cell clusters (arrows) which give rise to individual ommatidia. B. Cynthia: third-day adult development. The crystalline cones are forming. C. Autoradiograph of a cynthia eye on day 2 of adult development. The arrow indicates the silver grains at the border of the eye. The region to the left of the silver grains had begun differentiating. D. Polyphemus adult. Cross-section through rhabdoms with surrounding tracheoles of tapetai layer. E. Cynthia: seventh day of adult development. The crystalline cone vacuoles have fused. The arrow designates the groove which delineates the edge of the eye. F. Polyphemus: thirteenth day of adult development. The eye has the adult con-figuration but has not yet reached full size.
Eye development in the pharate adult
The cells underlying the crescent are not only morphologically more advanced in their differentiation than the rest of the presumptive eye, but, as Wolsky (1949) has shown by extirpation and microcautery, this region also serves as a ‘differentiation center’ which is required for the peripheral regions to differen-tiate. Umbach (1934) reports that approximately half-way through adult development the peripheral regions of the developing eye ‘catch-up’ with the differentiation center and from then on all of the ommatidia differentiate syn-chronously. I have found that the same pattern of development is present in cynthia where synchronous differentiation is established by the seventh day of adult development, at a time when the cornea has not yet appeared (Fig. 1E). By the thirteenth day of adult development the eyes of polyphemus have assumed their final form, although considerable enlargement by virtue of increase in the size of the individual cellular elements still occurs during the remaining week before emergence (Fig. 1F).
The adult eye
Although the development and morphology of the adult saturniid eye are similar to that described by Umbach (1934) for Ephestia, a brief illustrated (Fig. 1A-F) account of my findings is included in order to facilitate a considera-tion of the abnormalities found in the deuteropupa. Umbach’s terminology is used throughout.
The adult eye is covered with a transparent cornea composed of hexagonal facets each of which covers an individual ommatidium (Fig. 3C). Directly beneath the cornea lie Semper’s cells, which secrete the cuticular crystalline cone. The cornea is formed by the corneal pigment cells whose nuclei lie lateral and proximal to Semper’s cell nuclei. The retinula cell nuclei lie in a cluster connected distally to the crystalline cone by an axial strand and proximally to the rhabdom by another thin filament (Achsenfaden). The rhabdom is surrounded distally by other (Neben) pigment cells and proximally by the tapetai layer, a group of tightly packed tracheoles (Fig. 1 D). At the base of each ommatidium lies the basal retinula cell. The eye is covered at its base with a basement mem-brane and laterally is separated from the rest of the head by the cuticular eye capsule.
The eye of the deuteropupa
(1) Gross structure
It is an interesting curiosity, first reported by Williams (1959), that no matter how much juvenile hormone is injected into a pupa and no matter how com-pletely pupal most of the characteristics of the resulting deuteropupa are, their eyes never fail to assume some of the characteristics of adult eyes. The most obvious difference between the compound eye of the adult and the eye of the deuteropupa is the smaller size of the latter. The size of the compound eye is correlated with the degree of formation of pupal structure (Gilbert & Schneider-man, 1960; Williams, 1961), so that, in cases where the deuteropupa is extremely ‘pupal’, the compound eye may appear no larger than the initial pupal eye crescent. The severity of internal abnormalities was also positively correlated with the retention of general pupal characteristics.
(2) General organization and number of elements
By the seventh day of adult development in cynthia, presumptive eye tissue is set off from the adjacent epidermis by a deep fold of cells which subsequently is covered with cuticle which forms the eye capsule (Fig. 1 E). In the deuteropupa, the sharp distinction between eye and non-eye is lacking, the eye region of the deuteropupa consisting of a central area composed of recognizable pigmented ommatidia, a lateral zone of clustered poorly differentiated cells organized into presumptive ommatidia (Fig. 2B) and a peripheral zone of a multilayered epithelium. A shallow furrow can be recognized peripherally to these regions (Fig. 2A). This pattern suggests that those cells which were organized into ommatidial groups in the pupa (the cells underlying the pupal eye crescent) cannot be completely blocked in their course of differentiation by juvenile hormone. By contrast, the peripheral regions (which were not organized into pre-ommatidial groups in the pupa) either remain as they were or, in the case of deuteropupae with large eyes, differentiate with some cells only reaching the stage normally reached by the central region immediately following pupation.
Deuteropupa eye A. Note the extensive growth zone (between arrows) at the edge of the eye. B. The growth zone from another deuteropupa. The arrow indicates a cell cluster. A maximally formed crystalline cone is present immediately adjacent to the growth zone. C. Central region from a more adult-like deuteropupa. The major elements of the eye are recognizable.
Deuteropupa eye A. Note the extensive growth zone (between arrows) at the edge of the eye. B. The growth zone from another deuteropupa. The arrow indicates a cell cluster. A maximally formed crystalline cone is present immediately adjacent to the growth zone. C. Central region from a more adult-like deuteropupa. The major elements of the eye are recognizable.
Thus, in the formation of a second pupal instar, the central and peripheral regions each take but one step forward in development.
The pupal eye consists of approximately 2000 facets, the adult eye of about 8000 ommatidia. The smallest deuteropupal cornea examined (Fig. 3B) had roughly 4000 modified hexagonal corneal lenses; the region which appeared most like the crescent of the normal pupa contributed an additional 2000 rudi-mentary facets. The correspondence in number of rudimentary facets in the crescent of the deuteropupa and the normal pupa may be due either to a mechanism which governs crescent size or to the recruitment of new cells to form the ommatidia of the deuteropupa leaving the original crescent cells unaltered. The latter interpretation is highly unlikely in view of the experimental work which has shown that the ‘differentiation center’ must undergo some differentiation before the other areas of the eye (Wolsky, 1949).
Whole mounts of cuticle A. The pupal eye. The unfaceted furrow lies to the left and the area of presumptive ommatidia to the right. B. The deuteropupal eye. The equivalent of the pupal eye crescent is in the top of the picture, the region with abnormal facets lies below. C. The cornea of an adult polyphemus. D. The cornea of a deuteropupa. Note the abnormality in the center of each irregular facet. E. Polygonal field zone of a pupa. F. Polygonal field zone of a deuteropupa at the same magnification as E.
Whole mounts of cuticle A. The pupal eye. The unfaceted furrow lies to the left and the area of presumptive ommatidia to the right. B. The deuteropupal eye. The equivalent of the pupal eye crescent is in the top of the picture, the region with abnormal facets lies below. C. The cornea of an adult polyphemus. D. The cornea of a deuteropupa. Note the abnormality in the center of each irregular facet. E. Polygonal field zone of a pupa. F. Polygonal field zone of a deuteropupa at the same magnification as E.
A. The polygonal field zone of a pupa in cross section. B. The epidermis from the same region. C. The polygonal field zone and adhering epidermis from a deuteropupa. D. Whole mount of cuticle from a moth which had tubercles as a pupa. Note the central structure which resembled tanned pupal cuticle and the surrounding socketless region.
A. The polygonal field zone of a pupa in cross section. B. The epidermis from the same region. C. The polygonal field zone and adhering epidermis from a deuteropupa. D. Whole mount of cuticle from a moth which had tubercles as a pupa. Note the central structure which resembled tanned pupal cuticle and the surrounding socketless region.
(3) Histology of the ommatidium
(a) Length of the ommatidia
There was considerable variation in the length of the ommatidia in the deuteropupae, and the length of the ommatidia was positively correlated with the size of the eye. One deuteropupa’s ommatidia were but one-quarter the length found in an adult, and the least affected deuteropupa had ommatidia which were the same length as are normally encountered two-thirds of the way through adult development (Table 1; Fig. 2 A, C).
(b) The cornea
The most adult-like deuteropupal corneal facets were rounded hexagons and contained a central protuberance; their diameter was smaller than the diameter of the hexagons of the normal adult eye (Fig. 3D). In many of the deuteropupae a distinct pupal eye crescent was present in addition to the abnormal facets (Fig. 3B).
(c) The crystalline cone
Frequently the only indications of the crystalline cone in the deuteropupa were unfused vacuoles of secretion in Semper’s cells (Fig. 2 A). Even when the crystalline cone was present as a fused single structure, it was only about half the length normally reached by the thirteenth day of adult development (Table 1). The irregularities and small size of the cones cannot be due to premature emergence alone, for by the seventh day of adult development in cynthia complete fusion of the vacuoles into a single element has occurred (Fig. 1 E).
(d) The retinular cells and rhabdom
Frequently, there was no indication of the rhabdom in the deuteropupa, and when it was present it generally extended only part way to the basement membrane (Fig. 2 A, C). The mature form of the rhabdom (Fig. 1D) was not evident in cross-sections until the fourteenth day of adult development. In the deuteropupa the rhabdom was never represented by more than a dense rod in cross-section.
The clustered retinula cell nuclei are apparent in the eye of the deuteropupa but the basal retinula cell could not be located in many of the preparations (Fig. 2C). Presumably, it had not yet migrated away from the distal nuclear cluster, although this migration is normally complete by the eleventh day of adult development in cynthia (Fig. IE). Furthermore, in the deuteropupa the mass of retinula cell nuclei was located just below the crystalline cone. In the few cases where they were at a slight distance from the cone, the axial strand was recognizable.
(e) The tapetum and outer pigment cells
There was no indication of a tapetai layer in the eyes of the deuteropupa. The area between the crystalline cone and basement membrane was filled only by pigment cells (Fig. 2C). These cells, which appear as a matrix of triangles with pigmented sides when viewed in cross-section, were normally restricted to the area between the crystalline cone and the distal end of the retinula cell nuclear cluster (Fig. 1F). The tapetai layer was not evident in normal cynthia until the fourteenth day of adult development.
(f) Nervous elements
Nerves extend to the basement membrane across the entire width of the deuteropupal eye including the peripheral area, which had only differentiated as far as the cell-cluster stage (Fig. 2 A).
(4) Incorporation of tritiated thymidine into developing eyes
An examination was made of autoradiographs of normal eyes from a cynthia on the second day of adult development which had been injected with tritiated thymidine and fixed 18 h later. Although there was general background labelling over the entire eye, none of the eye elements which showed obvious signs of differentiation had any conspicuous labelling over their nuclei. By contrast, there was heavy labelling at the peripheral borders of the eye in the regions where the stratified epithelium had not yet formed cell clusters (Fig. 1C). Thus, while the peripheral borders of the eye incorporate thymidine and presumably syn-thesize DNA prior to forming ommatidia, the previously organized central region, which is insensitive to juvenile hormone, has already completed DNA synthesis.
Abdominal cuticle
A more direct analysis of the progressive differentiation of individual cells has been possible by utilizing the epidermis of the abdomen. The virtues of the insect epidermis for developmental analysis are its uniformity in cell type, single-layered nature, and well-documented developmental activities (Wigglesworth, 1959). Especially well suited for analysis is the anterior region of the interseg-mental membranes of lepidoptera, where each epidermal cell secretes a discrete ‘polygonal field’ which is, in fact, a ‘cast’ of the surface of the underlying cell (Kühn & Piepho, 1940) (Figs. 3E, F, 4A-C). The area and density of the poly-gonal fields therefore reveal the surface area and density of the underlying epidermal cells.
When a deuteropupa is formed following the injection of juvenile hormone, it remains within the exuviae of the old pupal cuticle. One may punch a small area of cuticle from the deuteropupa along with the overlying bit of cuticle from the first (normal) pupa. Marcus (1962) reports that in Galleria some of the cells of this region of the intersegmental membrane die, but that there is no cell division between the pupal and adult stages. Consequently, it seems likely that in the present work the two cuticles are secreted by the same group of cells.
Analysis of these double discs of cuticle revealed that there were approxi-mately half as many cells per unit area in the deuteropupal cuticle as there were in the original (Fig. 3 E, F). The surface of the polygonal fields of the deuteropupa looks more irregular in texture than that of the pupa.
A comparison of cross-sections of cuticle with its epidermis in diapausing pupae (Fig. 4A, B) and deuteropupae (Fig. 4C) also revealed an increased cell size in the latter. The height of the epidermal cell layer in the diapausing pupa was only about 13 while in the deuteropupa the cell layer may reach 47 in height. The cuticle was thicker in the diapausing pupa (up to 78 μ) than in the deuteropupa (maximum 50 μ). The nuclei of the epidermal cells were also significantly larger in the deuteropupa than in the diapausing pupa (Fig. 4B, C). The intense Feulgen staining, the presence of extra nucleoli, and the greater nuclear size all suggest that these cells were polyploid in the deuteropupa.
DISCUSSION
Action of juvenile hormone
(1) Theories
Ever since the discovery by Wigglesworth (1936) of the profound morpho-genetic control exerted by juvenile hormone there has been keen interest in its mode of action. Current hypotheses as to the primary site of its activity favor the nucleus. It has been postulated that juvenile hormone selects a set of genes which then direct the appropriate syntheses. Present conjecture is centered on whether the hormone activates a set of ‘larval genes’ (Wigglesworth, 1959) or blocks derepression of ‘new’ genes thereby enforcing rereading of a set of ‘previously used genes’ (Williams, 1961). Since it has been shown (Nayar, 1954) that larval cuticle grafted on to pupae will form scales and not pupal cuticle when the pupa moults to an adult, neither hypothesis can require that all steps in the program be morphologically expressed.
Reversal of metamorphosis, i.e. reversion to the larval condition at the pupal adult moult, has never been demonstrated in any saturniid. However, there are rare but well-documented cases of just such reversals in a few other systems (Wigglesworth, 1940; Piepho & Meyer, 1951; Lawrence, 1966). The “status quo’ hypothesis requires that the developmental program contains some mechanism whereby the cells keep track of their progress through the program. These cases of reversal of metamorphosis are not, however, a deciding factor in distinguish-ing between the Wigglesworth and Williams hypotheses. It is possible to account for the reversals reported in the literature by postulating that the cells have ‘lost track’ of their temporal progress through the developmental program in much the same manner that imaginai disc cells of Drosophila have been ‘transdeter-mined’, i.e. lost their state of spatial determination, by prolonged culture in adult hosts (Hadorn, 1966). It is significant that the same conditions which are necessary for transdetermination, namely prolonged culture with considerable cell multiplication, are identical to the conditions which have produced reversal of metamorphosis.
(2) Evidence for a program
The results presented in this paper constitute evidence for a developmental program intrinsic to the individual cells. Thus the larval tubercles which were found on pupae remained out of phase with the rest of the organism; in short, the cells in question secreted larval cuticle on a pupal background, and pupal cuticle on an adult background. The study of the eye demonstrated that the cells of the pupal eye field are relatively insensitive to juvenile hormone while the anterior region, which was not yet organized into pre-ommatidia in the pupa, was held in check and could not complete all of adult differentiation. Indeed the eyes of a deuteropupa resemble those of a hemimetabolous insect such as Notonecta (Liidtke, 1940), which has a growth zone around the periphery of the nymphal eye which contributes to the new facets added at each instar. The recognition of the stepwise differentiation of the eye explains some of contradic-tory results in the literature. Gilbert & Schneiderman (1960) attribute to species differences their finding that the pupal eye of polyphemus is insensitive to juvenile hormone whereas Piepho (1942) had shown that the eye discs of Galleria were one of the most sensitive tissues when corpora allata were implanted into last instar larvae. An alternative explanation is that certain regions of the pupal eye have reached a stage in their program which is relatively insensitive to the juvenile hormone whereas the imaginai disc of the larval eye is still at a state in its program which is highly sensitive. This interpretation is strengthened by the findings of Tsao, Jou & Chiang (1963), who showed that injection of juvenile hormone extracts into non-diapausing Samia ( = Attacus) ricini pupae within 12 h after pupation resulted in eyeless deuteropupae. In addition to the obvious completion of the cuticle, Bowers & Williams (1964) have demonstrated that there is DNA synthesis in the epidermal cells of the head during this period. These findings suggest that there must be some progress in the developmental program of the eye prior to the onset of diapause.
(3) What is the sensitive step in the program?
It is thus evident that individual tissues can proceed in the developmental program independently of adjacent tissues, and with differential sensitivities to juvenile hormone at different stages in their development. Furthermore, the program is responsible for a dissimilar array of syntheses in different tissues. In the general body epidermis, the presence of excess juvenile hormone favors the resynthesis of highly pigmented thin larval cuticle at the larval-pupal moult and heavily tanned pupal cuticle at the pupal-adult moult. Sensitive regions of the eye are prevented from carrying out diverse synthetic acts which are performed in the identical hormonal environment by cells which are more advanced but still morphologically undifferentiated.
Although the completion of DNA synthesis appears to be one factor which makes the central region of the eye relatively insensitive to juvenile hormone, both DNA synthesis and mitosis can be ruled out as sites of that hormone’s action. The present study shows that post-mitotic tissues such as the inter-segmental membrane are sensitive to juvenile hormone; and Wigglesworth (1959) demonstrated that juvenile hormone increases mitotic activity in specific regions of Rhodnius. The polygonal field zone apparently becomes polyploid under the influence of juvenile hormone; and Krishnakumaran, Berry, Ober-lander & Schneiderman (1967) could detect no differences between tritiated thymidine incorporation during the pupal-adult and pupal-deuteropupal moults.
(4) Can the action of juvenile hormone be delayed?
Recent studies with embryos (Riddiford & Williams, 1967) and with develop-ing aphids (White, 1968) have shown that juvenile hormone applied at one stage may not show an effect until several instars later. Riddiford and Williams applied juvenile hormone preparations to developing saturniid embryos and found that larval development was normal but that the pupae frequently had ‘larval’ tubercles and were similar to those described in the first section of this paper. They interpreted their data to mean that the epidermal cell program had been ‘reset’. There is an alternative explanation to this delayed action of the respon-sive cells, namely that extra juvenile hormone in the embryo might prevent the slowing down of the activity of the corpora allata which normally precedes pupation. In that case abnormal pupae would be produced because the moult took place in the presence of excess juvenile hormone. The permanent alteration of endocrine activity by prenatal treatment with hormones is well documented in vertebrates (Barraclough, 1967).
The significance of the abnormalities in the deuteropupal eye
Although the deuteropupal eye bears a superficial resemblance to the adult eye it does not show normal morphology. Thus the tapetai layer is not formed and the rhabdom is not completely differentiated. The failure of the ommatidia to differentiate completely may be due in part to the precocious maturation of the deuteropupa—especially to the precocious secretion of new cuticle. It is unlikely that the abnormalities are caused by any failure of the central nervous system, for Wolsky (1949) has shown that the eyes of Bombyx adults deprived of brains as young pupae are usually completely normal. Furthermore, corneae from cecropia adults debrained as pre-diapausing pupae are also normal (un-published observations). The abnormalities of the deuteropupal eye are essen-tially different from the temporally graded abnormalities produced by the implantation of 5-fluorouracil in Ephestia (Imberski, 1967). Possible insight into the nature of the abnormalities might be obtained by comparing the eye of true deuteropupae with those from the abnormal adults obtained by Williams (1968) following injections of excessive amounts of ecdysone and some of its analogues into pupae. He has attributed the resulting abnormalities to a speed-up in developmental rate and premature cuticle deposition.
The finding that certain elements of the pupal eye, such as the protuberance in the center of each corneal facet, are maintained while other elements dif-ferentiate towards the adult structure, is reminiscent of the findings of Wiggles-worth (1940) that juvenile hormone applied at successively later times in the nymphal-adult transformation of Rhodnius affected a decreasing number of characters. These findings show that juvenile hormone does not exert an all-or-none action, but that it can control a developmental sequence within an indi-vidual cell. Juvenile hormone’s precise control over such small steps in the developmental program constitutes strong evidence for the status quo hypothesis.
SUMMARY
Cecropia pupae with larval tubercles and their transformation into adults with patches of scaleless or pupal cuticle were described.
Eyes of deuteropupae produced by the injection of juvenile hormone into polyphemus pupae were analysed. The deuteropupal eye was smaller than the normal adult eye, was bordered by a ‘growth zone’, and individual elements of the ommatidia had not differentiated normally or completely.
The polygonal field zone of the intersegmental membrane of the deutero-pupa consisted of fewer and larger cells than the corresponding region of the pupa.
These results were interpreted as supporting the ‘status quo’ theory of juvenile hormone’s action.
ZUSS A MENFASSUNG
Der Ablauf der Differentiation und seine Kontrolle durch das Juvenilhormon bei Saturniiden
Cecropia-Puppen mit larvalen Tuberkeln und deren Umformung in adulte Tiere mit Flecken von schuppenloser oder pupaler Kutikula wurden beschrieben.
Untersucht wurden die Augen der Deutero-Puppen, die durch die Injektion des Juvenilhormons in Polyphemus-Puppen entstanden. Das deutero-pupale Auge war kleiner ais das normale Auge eines adulten Tieres, es war begrenzt durch eine ‘Wachstums-Zone’, und einzelne Teile der Ommatidien hatten sich nicht in der normalen Weise oder vollstândig ausdifferenziert.
Die polygonale Feld-Zone in der intersegmentalen Membran der Deutero-Puppen bestand aus weniger und grdsseren Zellen als die entsptrechende Zone einer Puppe.
Diese Ergebnisse wurden so interpretiert, dass sie die ‘Status gwo’-Theorie der Wirkungsweise des Juvenilhormons unterstüzen.
Acknowledgements
This work has been supported by grants CF-8605 and HD-03163 of the U.S. Public Health Service. Part of the work was carried out in the Department of Zoology, Oxford University, and the hospitality of Dr Peter Brunet is gratefully acknowledged. Dr Ellis MacLeod pro-vided useful criticism of an early draft of the manuscript. My special thanks are due to Professor Carroll M. Williams. This study was begun under his direction at Harvard Univer-sity, and he has continued to provide stimulating suggestions throughout the course of the work and the preparation of the manuscript.