The integument of Peripatopsis moseleyi has been examined by light and electron microscopy with particular reference to the structure and formation of the cuticle. The evidence supports the idea that Peripatus is a true arthropod but not that it has direct affinities with the annelids. The characteristics of arthropod cuticle are present in their simplest form and pore canals and dermal glands are lacking. The cuticle is 1 or 2µ thick except in the hardened claws and spines. Above the procuticle (chitinprotein) is a thin 4-layered epicuticle. It is possible that the innermost of the 4 layers (prosclerotin) may correspond to cuticulin of other arthropods. In the claws and spines tanning in this layer extends to the procuticle. Hydrofuge properties of the cuticle probably depend on the outer layers of epicuticle, and it is suggested that the lamina concerned might consist of oriented lipid associated with lipoprotein (Dr. J. W. L. Beament). Wax and cement are absent. Non-wettability of the cuticle is probably ensured by the contours of micropapillae which cover the surface. Similar structures arise in Collembola and other terrestrial arthropods by convergence.

The formation of new cuticle before ecdysis is described. After the epicuticular layers are complete, the bulk of the procuticle is laid down in a manner probably common to all arthropods. Secreted materials originate in small vesicles derived from rough endoplasmic reticulum and from scattered Golgi regions. The latter contribute to larger vacuoles which rise to the surface of the cell and liberate material in a fluid state. This later consolidates to form procuticle. Vesicles may also open to the surface directly, and ribosomes probably occur free in the cytoplasm. At this stage the cell surface is reticulate, especially under micropapillae.

The ordinary epidermis has only one kind of cell, attached to the cuticle by tonofibrils disposed like the ribs of a shuttlecock, and to the fibrous sheaths of underlying muscle-fibres by special fibres of connective tissue. These features and the presence of numerous sensory papillae are associated with the characteristic mobility of the body wall. The appearance of epidermal pigment granules, mitochondria, the nuclear membrane, and a centriole are noted. No other cells immediately concerned in the formation of cuticle have been found. By contrast myriapods, which do not have wax either, possess dermal glands secreting far more lipid than is found in the Onychophora. The wax layer found in insects and some arachnids constitutes an advance of high selective value which emphasizes the primitive condition of the Onychophora. It is noted that the thick layer of collagen separating the haemocoel from the epidermis probably restricts the transfer of materials. It is suggested that since some features of cuticular structure and formation appear to be common to all arthropods, it is possible that some of the endocrine mechanisms associated with ecdysis may also be similar throughout the phylum.

THE evolutionary position of the Onychophora has been much discussed. A specimen collected by Sloane early in the last century was identified by Shaw as Nereispedata (Bouvier, 1905). Guilding (1826) considered Peripatus to be a mollusc, but most of his contemporaries referred it to the annelids or to a group linking the annelids and myriapods. Subsequent studies supported the latter view. The classic work of Balfour (1883) and Sedgwick (1885-8) suggested that although Peripatus had affinities with the annelids, it was evidently an arthropod. Chief arthropod features were the presence of hardened claws and jaws, a cuticle shed by ecdysis, the tracheal system, the large haemocoel, and dorsal ostiate heart. By contrast the extensible body with its smooth muscle and soft cuticle, serial nephridia, the presence of cilia, and the simple eyes were more characteristic of annelids. These observations supported the idea put forward on other grounds that the annelids and arthropods constituted a super-phylum (e.g. Lankester, 1904). The notion is established in several textbooks of zoology, where the Arthropoda are usually defined as a group with an annelid-like central nervous system, and Peripatus is regarded as a primitive link between the 2 phyla. Although Peripatus undoubtedly has primitive features, it has been pointed out that many of its apparently annelid characters have an adaptive value which must affect their validity as phylogenetic markers (Manton, 1959). It is possible to relate many of them to the cryptic habits of the Onychophora rather than to the legacy of ancestral annelids. The group is still regarded by some as a distinct phylum (Snodgrass, 1938, 1952; Butt, 1959, 1960), but it seems difficult not to accept the Onychophora as a class of the Arthropoda (Moseley, 1874; Sedgwick, 1885-8; Manton, 1950, 1953, 1958, 1961; Tiegs and Manton, 1958). Within this phylum they have most in common with the myriapods and insects.

Recent embryological studies have failed to disclose any further annelid resemblances (Manton, 1949, 1960; Butt, 1959, 1960). On the other hand Tiegs (1940, 1945, 1947) described several features in the development of myriapods which suggest common ground with the Onychophora. It is beyond doubt that the Onychophora are an ancient group (e.g. Manton, 1949), and it has even been thought that their distribution indicates that of former land masses (Clark, 1915). The Cambrian fossils of Aysheaia represent marine Protonychophora (Walcott, 1911, 1931; Hutchinson, 1930), but unfortunately they throw little light on phylogenetic problems. Living Onychophora hold great interest as primitive members of a special kind of forest fauna (Lawrence, 1953) and it is proposed to consider them in this light rather than to regard them as an inter-phyletic link.

Previous work on the cuticle

The integument consists of cuticle, epidermis, and dermal connective tissue (Balfour, 1883). Early authors refer to the cuticle as empirically ‘chitinous’ (Bouvier, 1905), but Kunike (1925) showed that the cuticle of Peripatus broelmanni gave a positive Schultze reaction. Manton (1938) recorded chitin and protein in the exuvium of Peripatopsis. The presence of chitin was confirmed by Richards (1951), while X-ray analysis has indicated that it occurs as a-chitin as in other arthropods (Lotmar and Picken, 1949; Rudall, 1955, and see 1963).

Ecdysis was recorded by Hutton (1876) and has since been described by Steel (1896), Manton (1938), and Holliday (1942, 1944). A mid-dorsal split in the old cuticle starts behind the antennae and eventually extends the whole length of the animal. The cuticle is worked off from the front backwards, one appendage being freed at a time as from a multidigit glove. The process is assisted by contractions of the body and takes something over an hour (Lawrence, 1953). The lining of the anterior part of the gut is also shed (Manton, 1938). The crumpled pellet of cuticle is usually eaten (Steel, 1896). Accounts referring to Peripatopsis moseleyi, the species studied here, are those of Holliday (1942, 1944) and Lawrence (1953). Manton’s careful observations of 4 species of Peripatopsis established that moulting occurs every 2 weeks. The cuticle is shed at birth, and ecdysis continues throughout life. The endocrine mechanism responsible for moulting has not been studied in Onychophora. There may prove to be some parallel with apterygote insects, where ecdysis also continues throughout life (Boelitz, 1933; Paclt, 1956; Watson, 1962), or perhaps more especially with myriapods (CloudsleyThompson, 1958; Juberthie-Jupeau, 1963; Scheffel, 1961, 1963). Gabe (1954) has found neurosecretory cells in 3 onychophoran genera, but their functions are not yet known. Sanchez (1958) re-examined Peripatopsis moseleyi and suggested that the infracerebral organs may store neurosecretory products ; but a possible homology with corpora allata of insects is still un- substantiated (von Kennel, 1885-8; Dakin, 1922). In captivity Onychophora may continue to grow and to moult for a number of years (Manton, 1938), but in natural conditions their activity fluctuates with seasonal climatic changes and it is not known to what extent feeding, growth, and other processes are reduced when winter temperatures are low.

The shed cuticle is unhardened except for the claws, the jaws and their apodemes, and, to a lesser extent, sensory spines. The proximal surface is hydrophil, and the most crumpled specimen can be extended if it is floated on water (Steel, 1896; Holliday, 1944). The cast skin then exceeds the normal length of the animal by one-third (Manton, 1938). As it is not noticeably elastic the slack must normally be implicated in surface furrows. When cast the cuticle is white, but in position it is transparent and has no colour apart from the brown claws and jaws. The characteristic orange, black, or green pigments occur in the periphery of epidermal cells beneath (Balfour, 1883).

In view of the cryptic habits of the Onychophora (Lawrence, 1953; Manton, 1958) it is not surprising that the sensory functions of the integument are well developed. The whole body is covered with rounded or conical papillae which usually end in a sensory spine, and whose arrangement is a specific character (Bouvier, 1905-7). Corresponding to cells of the epidermis there are secondary papillae. Papillae, spines, and pigmentation are accentuated on the dorsal surface. The spines are sensitive to touch. Similar ones occur in large numbers on the antennae, where they are less robust, and on the sensory pads of the feet. These sense organs have been described by Schneider (1902), Bouvier (1905), and Manton (1937), and their development was studied by Duboscq (1920). Manton has also described open-ended conical spines from the lips and tongue which are undoubtedly chemoreceptors. The smooth corneal covering of the eye is produced by modified epithelial cells (Dakin, 1920).

Onychophora live in damp habitats and the cuticle is better adapted to repel water than to resist desiccation. Neither drops of water nor ejected slime spread on the cuticle (Hutton, 1876), as the general surface is hydrofuge. If held under water an animal is covered by a film of air (Willey, 1898). Although in such conditions specimens always drown, some species have been recorded from very wet places (e.g. Peripatopsis balfouri under a waterfall (Brink, 1957); P. albida in damp caves (Lawrence, 1931)). Peripatopsis moseleyi prefers a relative humidity of 95% (Bursell and Ewer, 1950). In dry conditions, on the other hand, water loss is rapid ; P. sedgwicki in a wind tunnel at 30°C loses water 80 times faster than a cockroach and twice as fast as an earthworm (Manton and Ramsay, 1937; and see Dodds and Ewer, 1952). At 24°C, Epiperipatus and Oroperipatus over calcium chloride lost water at half the rate of an earthworm, and at 20 times that of a millipede (Morrison, 1946). Water is lost through the many tracheal openings, for these cannot be closed (Manton and Ramsay, 1937; see also Mendes and Sawaya, 1958). The behavioural responses studied by Bursell and Ewer thus probably protect the animal from desiccation much better than its cuticle. On the other hand, water does not seem to be absorbed through the cuticle. Fluid may be taken in by drinking, or in Opisthopatus by eversible sacs on the legs (Alexander and Ewer, 1955). It may be mentioned that the problem of desiccation does not arise during development, as nearly all the Onychophora are viviparous. They are all uricotelic (Manton, 1937).

The structure of the cuticle, with which this paper is concerned, has been recorded several times. Balfour (1883) and Schneider (1902) illustrate it as a thin layer 1 or 2 µ thick which follows every wrinkle of the epidermis. Besides the scale-like papilla above each cell, the surface shows minute regular corrugations (here termed micropapillae). Manton (1937) has confirmed this and notes two layers within the chitinous cuticle: the outer stains red after Mallory and takes up iron haematoxylin, while the inner stains with Mallory only and is then blue. In addition there is a refractile layer at the surface (Manton, 1959). Manton and Ramsay (1937) state that the integument is formed of chitin covered by a very thin cuticle, referring no doubt to the differentially-staining layers and to the refractile layers respectively. They attribute the non-wettability of the cuticle to the presence of the numerous papillae.

The epidermal cells are columnar and have large oval nuclei (Balfour, 1883). Their structure in relation to the cuticle at any stage in the moulting cycle has not yet been examined. Manton (1937), however, has studied the cells of the alimentary canal in relation to the renewal of the periptrophic membrane.

Although circumstances have not permitted the study of a complete moulting series, a general description of the structure of the integument is attempted.

Observations have been made on Peripatopsis moseleyi Wood-Mason. Histological sections were prepared from specimens treated with ether, chloroform, or carbon dioxide before fixing. Fixatives used were Susa for staining with Mallory or Masson (Pantin, 1948), formaldehyde-calcium for frozen sections (Baker, 1944), and osmium ethyl gallate (Wigglesworth, 1957a, 1959a).

The most successful material for electron microscopy was fixed in 1% OsO4 in veronal-acetate buffer at pH 7-2, usually for an hour at 0°C(Sjöstrand, 1956). It was taken through acetone and propylene oxide (Luft, 1961) into araldite (Glauert, 1962) and embedded in a specially hard mixture (Luft, 1961) containing

Sections were cut on a Huxley ultramicrotome and examined after staining with uranyl acetate (Gibbons and Grimstone, 1960) or lead hydroxide (Watson, 1958; Peachey, 1959). Electron micrographs were taken on 35 mm film with a Phillips E.M. 200 at 60 kV. An objective aperture of 25 µ was used.

Replicas of the cuticle were prepared by shadowing with carbon and then with chromium, backing with a layer of |% formvar or collodion, and dissolving away the specimen with 20% chromic acid or 10% sodium hypochlorite. Replicas were washed with distilled water, and cut into pieces which were picked up on grids for examination.

The external surface

The general appearance of Peripatopsis moseleyi is seen in fig. 1, A; at this magnification the larger sensory papillae are just visible. The surface of the animal may conveniently be studied in cast cuticles, which provide material for whole mounts and for chemical tests. Cell outlines, tracheal openings, and papillae can be seen (fig. 3, A to c). Replicas examined with the electron microscope confirm the details of light microscope observations. It has long been known that the cuticle is covered with blunt prickles (fig. 3, c). These now appear as micropapillae with characteristically rounded contours (fig. 1, B). Over the apex of an epidermal cell they easily exceed 1 μ, but towards the periphery they are only about 0 ·1 μ high and they are absent from the furrows corresponding to cell boundaries. Fig. 2, from another replica, shows a small sensory papilla (see also fig. 3, A). Neither replicas nor electron microscope sections reveal any granules or filaments on the outer surface.

FIG. 1.

(plate), A, Peripatopsis moseleyi ?, approx. ×2. B, electron micrograph of a replica of the cuticle of P. moseleyi, showing micropapillae. The print covers about a quarter of the surface area of a flat epidermal cell (compare figs. 2, 3). Collodion shadowed with Cr at 6o° (p. 285).

FIG. 1.

(plate), A, Peripatopsis moseleyi ?, approx. ×2. B, electron micrograph of a replica of the cuticle of P. moseleyi, showing micropapillae. The print covers about a quarter of the surface area of a flat epidermal cell (compare figs. 2, 3). Collodion shadowed with Cr at 6o° (p. 285).

FIG. 2.

(plate). A replica of the cuticle, showing a small sensory papilla in surface view. The spine in the centre of the rosette of cells has collapsed and cannot be seen. Formvar shadowed with Cr at 40°.

FIG. 2.

(plate). A replica of the cuticle, showing a small sensory papilla in surface view. The spine in the centre of the rosette of cells has collapsed and cannot be seen. Formvar shadowed with Cr at 40°.

FIG. 3.

(plate). Peripatopsis moseleyi, whole mounts of cast cuticle.

A, surface view, showing a medium-sized sensory papilla (right), 2 minor papillae, and a tracheal opening (lower centre). Micropapillae can be seen within some of the cell outlines. Phase contrast.

B, another preparation, focused to show furrows corresponding to cell boundaries.

c, tracheal opening in surface view. The ragged oval outline is probably the inner border of the tracheal pit, within which would open the tracheoles. Note contours and micropapillae of surrounding cells.

D, large sensory papilla, showing argentaffin reaction of spine (after 90% methanol).

E, jaws from shed cuticle boiled in 10% potash. Phase contrast.

F, right jaws of a large specimen: whole mount. Note apodeme (lower left). Phase contrast.

FIG. 3.

(plate). Peripatopsis moseleyi, whole mounts of cast cuticle.

A, surface view, showing a medium-sized sensory papilla (right), 2 minor papillae, and a tracheal opening (lower centre). Micropapillae can be seen within some of the cell outlines. Phase contrast.

B, another preparation, focused to show furrows corresponding to cell boundaries.

c, tracheal opening in surface view. The ragged oval outline is probably the inner border of the tracheal pit, within which would open the tracheoles. Note contours and micropapillae of surrounding cells.

D, large sensory papilla, showing argentaffin reaction of spine (after 90% methanol).

E, jaws from shed cuticle boiled in 10% potash. Phase contrast.

F, right jaws of a large specimen: whole mount. Note apodeme (lower left). Phase contrast.

This surface is hydrofuge, and drops of water and slime are easily shaken off. The proximal surface, on the other hand, is so hydrophil that if a recently-shed exuvium is placed on water it may flick out spontaneously to its full extent. The lower surface of a floating preparation takes up dilute dyes such as aniline blue, but stains do not penetrate from the outer surface.

Alcohols and lipid solvents destroy the hydrofuge properties of the outer cuticular surface. Whole animals narcotized for fixation, for instance, retain an enveloping air film when placed in formalin unless first dipped in ethanol. Shed cuticles become permeable to water after treatment with even 50% ethanol, or with methanol, chloroform, or ether: this may be seen with a cuticle floating on water, as any area on to which solvent has been dropped is then readily penetrated by drops of water. There is little apparent change when the water on which a cuticle is floating is heated to 90° C, and detergents such as teepol and C.09993 (provided by Dr. J. W. L. Beament), which disperses wax in insect cuticles (Beament, 1945), have little visible effect.

These observations do not suggest that the hydrofuge properties of the surface depend on the presence of free lipid, and nothing which might reasonably be interpreted as such a layer has been seen in electron microscope sections. The permeability of the cuticle has not yet been measured. It may be found that water molecules pass more readily out of the cuticle than inwards, as in insects (Beament, 1960, 1961), but the animal has so many spiracles that tracheal evaporation must account for much of the ordinary water loss (Manton and Ramsay, 1937). Tracheal openings are up to 10 µ in diameter (fig. 4, A, C) and an adult P. moseleyi has at least 50 in each segment. The cuticle must normally be hydrated as dried exuvia are brittle.

The composition of the cuticle

In sections the ordinary cuticle is 1 or thick (fig. 3). Two layers are usually demonstrated by staining (Manton, 1937). Since the outer layer takes up acid fuchsin and iron haematoxylin and predominates in hardened regions such as the sensory spines, claws, jaws, and cornea, Manton suggested that it is sclerotized. In osmium / ethyl gallate sections it usually shows as a dark line. In this account it is called prosclerotin (Blower’s term is used for convenience, see p. 295) and regarded as the innermost layer of the epicuticle. The inner layer is blue after Mallory staining and green after Masson, and will be referred to as procuticle (Richards, 1951).

The outer epicuticle covering the surface is very thin and has been studied only in electron microscope sections.

The cuticle gives positive chitin and protein tests before and after ecdysis. Exuvia may be regarded as chemically intact, for the shed cuticle is normally eaten and it is doubtful whether any material is reabsorbed from the inner zone before ecdysis (p. 291). The whole cuticle is so thin that crude solubility tests are not likely to reveal differences between the different layers. In the course of preparing replicas for electron microscopy it was established that the cuticle remains whole in concentrated solutions of urea, lithium chloride and thiocyanate, and acidified potassium permanganate. It may partly dissolve in 72% sulphuric acid or diaphanol. It is dissolved slowly by 20% chromic acid, and more rapidly by 10% sodium hypochlorite or a warm solution of potassium nitrate in nitric acid. This suggests possibly that chitin and protein molecules may be bound together in equivalent molecular proportions in most of the cuticle.

Pryor (1940) records that in paraffin sections of Peripatopsis capensis the outermost cuticle (‘a well-developed epicuticle’) gives a positive argentaffin reaction throughout. In unfixed material, however, a positive reaction is given only by the claws, jaws, sensory spines, and chemoreceptors (fig. 3, D). Previous treatment with ethanol or methanol or with C.09993 makes little difference, and this suggests that some intrinsic property of tanned regions may be involved. It is not clear why only the middle region of spines reacts ; Wolfe’s suggestion (1955) that argentaffin-positive areas in the cuticle of Calliphora larvae reflect regions of greater permeability does not seem applicable, and further tests are needed.

Weak fluorescence in the cuticle appears to coincide with the distribution of sclerotization. Frozen sections of fresh and fixed material (formaldehydecalcium) were examined with a Reichert microscope (for the use of which I am indebted to Dr. J. G. Cruickshank of the Department of Pathology, Cambridge). A pale blue-green fluorescence appears to be confined to the outer zone of the cuticle. Sensory spines are bright and somewhat yellower, especially in fresh sections. Claws and jaws do not fluoresce strikingly in dry mounts. No attempt has been made to extract fluorescent materials. Fluorescence in other arthropod cuticles is well known (Willis and Roth, 1956). Andersen (1963) concludes that in the locust, bluish-white fluorescence of the protein resilin is due to 2 amino-acids which may form cross-links between peptide chains.

The epidermis

The cuticle is secreted by the epidermis, whose cells have prominent oval nuclei (fig. 4, D, F). Between the epidermis and the circular muscle of the body wall is a layer of connective tissue up to 30 µ thick, consisting mainly of collagen fibres (fig. 8, D). In addition the epidermal cells appear to be anchored individually to fibrous sheaths surrounding the circular muscle-fibres (Owen, 1959). As shown in fig. 4, F the base of the cell, which may be drawn out into several lobes, is connected by specific fibres to the underlying muscle. Owen has found that the fibres stain neither as collagen nor as reticulin. They might ensure cohesion of the dermis, for it is important that the epidermis should be closely knit to the tissues beneath : the body wall of Peripatus gives rapid local responses besides showing wide variations in tone (Manton, 1958).

Tonofibrils usually extend from the base of the cell to the cuticle (fig. 4, c, D). They stain differently from connective tissue fibres outside the cell and are not obviously related to them. Little is known of how they are attached to the cuticle, or indeed of their precise function, especially as here they are not continuous with myofilaments. When many tonofibrils are present, other cytoplasmic components appear to be aligned between them (Balfour, 1883). Mitochondria are generally abundant, and pigment granules are concentrated distally to the nucleus. It is not known whether pigment granules can migrate or increase in number in response to changes in the environment. P. moseleyi occurs in green and brown varieties, both of which develop tints of grey or black (probably melanin). As in other Peripatopsidae, the pigments are fast in alcohol and are probably bound to protein, but little more than this is yet known. Dr. G. Y. Kennedy (personal communication) has suggested that dark green P. moseleyi may possibly contain a y-carotenoid.

Cytoplasmic elements corresponding to micropapillae are sometimes seen in histological sections as peripheral corrugations of the cell (figs. 4, A, B, F). Large papillae arise by cell division: as the rosette of cells enlarges, a sense organ and spine may develop (Duboscq, 1920). The development of a sense organ can be traced at least to the previous moult. The appearance of the integument is a taxonomic character (Bouvier, 1905-7), and it would be interesting to study some of the genetic factors involved, as Lees and Waddington (1942) have done for Drosophila.

Ecdysis

Changes in the epidermis during the moulting cycle have not yet been examined histochemically, but it should not be difficult to do so. Old and new cuticles are very similar histologically, although the new cuticle is somewhat thinner and material is probably added to it after ecdysis (p. 292). The newly-exposed cuticle of a moulting animal is moist (Holliday, 1944), which indicates that some kind of ‘moulting fluid’ may be present. The animals are particularly sensitive to humidity changes at ecdysis and easily die of suffocation in a saturated atmosphere. This might suggest that the new cuticle was for a short time more permeable than the old : on the other hand, the surface even of an animal damp with ‘moulting fluid’ is hydrofuge, and forms an air film under water. Critical changes in the tracheal system may occur at this time. It is possible that ‘moulting fluid’ is reabsorbed through the surface, as the new cuticle soon appears dry. It is likely that the development of a hydrofuge surface may be achieved before moulting, and also that sclerotized parts may be hardened as they form since the animal’s activities seem unrestricted as soon as the old cuticle has been shed. When the latter is eaten the limbs are in full use and the jaws and claws are at least hard enough to be functional. Evidence from electron microscopy so far supports these suggestions.

Structure of the cuticle

Electron microscope sections of the cuticle show layers corresponding to those in histological sections, but their relation to those recognized in the cuticles of other arthropods is not immediately clear. The 2 layers differentiated by histological stains are provisionally referred to as inner epicuticle and the procuticle (p. 295). The ‘réfringent layer’ covering these (Manton, 1959) is then outer epicuticle, and is composed of 3 thin layers.

The epicuticle thus consists of 4 layers in all (numbered 1 to 4 in fig. 5, D). The outermost is a uniform lamina between 10 and 15 mµ thick (1, figs. 5, A to E; 6, A). It tends to lift away from its substrate, although it is more firmly attached over micropapillae and other projections. A similar but thinner layer beneath (3, perhaps 5 mp thick) is separated from the outer layer by a dark line, which is interpreted as layer 2. Layer 2 is not strikingly osmiophil in unstained sections. It is not homogeneous, for it may include fine striations or wisps of material which appear to connect layer 1 to layer 3 (fig. 5, E). Unlike these 2 layers, which hardly stain with heavy metals, the lowest layer of the epicuticle is distinctly osmiophil and stains with lead (4, fig. 5, D). It is perhaps o-i p. thick and lacks resolvable texture. While it is not impossible that it may correspond to the cuticulin of insects (Locke, 1961) it is preferred for the moment to regard it simply as tanned prosclerotin. The 4 layers invest the ordinary surface and spines (fig. 6, A). They are laid down at an early stage and appear to be complete well before ecdysis (p. 291).

There is no direct evidence as to the composition of the epicuticle. Electron microscope observations make the presence of wax or cement unlikely. In an attempt to detect free lipid, specimens anaesthetized with carbon dioxide were treated with dripping cold chloroform for 1 min and then fixed, embedded, and sectioned. This procedure immediately renders the surface permeable to water, but none of the epicuticular layers seems visibly affected by it. It is possible that the thin second layer stains less darkly, but the fine striations or filaments at this level seem to persist. Unless they occur in layer 2, free lipids thus seem to be absent. Bound lipids are not excluded and they might for instance account for the osmiophil nature of layer 4.

It has been suggested that hydrofuge surfaces from which free lipid is absent may be expected to include tanned protein components (Davies and Edney, 1952; Beament, 1960, 1961). The compact appearance of the epicuticular laminae supports the idea that they contain chemically bound and dehydrated protein. This has not yet been proved but the fact that the cuticle is permeated by 50% ethanol despite remaining hydrofuge at 90°C is consistent with the suggestion.

The underlying procuticle is basically composed of chitin with untanned protein. In electron microscope sections it is at least 0-2 p, thick and usually more. Recently-added material has a loose fibrillar texture and stains darkly with uranyl acetate (fig. 5, A, B; p. 292), but the bulk of the procuticle has a closer structure and stains less intensely. In spines, the outer zone of the procuticle becomes indistinguishable from the inner epicuticle (fig. 6, A). It is presumably hardened and it fails to stain with uranyl acetate. Islands or lamellae of similar material (fig. 5, E) extend through the zone regarded as ‘sclerotized’ on histological grounds. Unhardened procuticle persists as the main core not only of the micropapillae but of the ridges on spines, where it stains more darkly than elsewhere (fig. 6, A). Further examination of hardened structures is required, particularly in sections stained with lead (Locke, 1961 ; Noble-Nesbitt, 1963 a, bY

The term ‘exocuticle’ has not been used because in the few sections which have been examined, hardened procuticle seems to be continuous with the inner epicuticle. If further study confirms that there is in principle only one sclerotized layer in the Onychophora (here termed prosclerotin), homologies with the cuticles of other arthropods may be established with greater certainty (see p. 295).

Secretion of the cuticle and moulting

Peripatus seems to have neither dermal glands nor pore canals. Before ecdysis, however, the peripheral cytoplasm is thrown into a reticulum, especially towards the apex of the cell. There are usually nodes corresponding to the positions of micropapillae (fig. 6, B). After moulting most of the villi are withdrawn and the cell surface is minimal except for blunt lobes under cuticular projections. In the next cuticle many micropapillae thus arise directly beneath existing ones.

At ecdysis the old cuticle is still almost intact. The space between old and new cuticles is so narrow that little ‘moulting fluid’ can be present (figs. 5, A, c; 6, B). All the layers of the new epicuticle have been laid down before any space can be seen, and at an early stage the new cuticle is almost continuous with the old. The only feature which recalls the usually obvious activity of arthropod moulting fluid is a preliminary change in the old procuticle. The loosely-packed layer bordering the cytoplasm becomes less marked, and the region above this, particularly where it dips between micropapillae, stains more darkly with uranyl acetate than before (figs. 5, A; 7, B). AS it looks coarser in texture and is not osmiophil, this is perhaps a zone from which dissolved material has been removed.

The evidence available suggests that the arrangement of cuticular layers indicates the order in which they are formed. The outer layer of epicuticle (1) appears suddenly in the viscous material between the cell membrane and the altered procuticle. It condenses out as 2 lines initially less than 10 mp. apart, between which the substance of this layer becomes concentrated. The new lamina increases in thickness and it soon looks fully formed (fig. 5, A, c). Layer 3 originates as 1 or perhaps 2 lines at the same stage (fig. 5, A, c) and is later continuous with layer 4 beneath. As the outer layers of epicuticle become defined the material surrounding them becomes attenuated and persists only as filaments. Strands between layers 1 and 3 remain as the striations previously noted within layer 2 (p. 290; fig. 5, D, E), but filamentous connexions between the outer lamina and the old cuticle soon disappear.

It may be speculated that the lines which initiate the formation of layers i and 3 (fig. 5, c) might represent films of newly-secreted polar lipid. The attenuation of the matrix as these layers develop suggests that orientation of materials already present may then occur. If the materials concerned are proteins it is possible that they might have been salvaged from the old cuticle in the first place. The sections suggest that early delineation of epicuticlc follows gradients of surface activity and spreads to the periphery of the cell. All processes are well marked at the micropapillae, where the outer layers of epicuticle adhere firmly and at early stages look thicker than elsewhere.

Layer 4 arises beneath these laminae. Material is secreted which at first resembles in texture the altered regions of old procuticle (fig. 5, A). The cytoplasm stains fairly intensely and granules resembling ribosomes are seen near the cell membrane. The inner layer of epicuticle (4) is soon distinct, so that ‘sclerotization’ and possibly the secretion of an osmiophil lipid may be proceeding at this stage. Following this the procuticle is laid down, with further sclerotization in regions such as spines.

Once the secretion of procuticle is under way, it is laid down from material in large vacuoles which liberate their contents at the surface of the cell. The procuticle increases in thickness after ecdysis, and a certain number of vacuoles is always present. They are often seen beneath micropapillae during the intermoult (fig. 8, A). During the active secretion before ecdysis, mitochondria are abundant (figs. 6, B; 8, c; 9), rough endoplasmic reticulum is well developed, and a number of Golgi regions can usually be seen in each cell (fig. 6, c). The systems of cytoplasmic membranes give rise to numerous small dense vesicles, which coalesce to form first the large vacuoles and later their contents. Figs. 6, c; 7; 8, A, B; and 9 illustrate this process. It is certain that Golgi regions proliferate some of the vesicles (figs. 6, c; 7, A; 9, c), which then aggregate into vacuoles (figs. 7, A; 8, B; 9, A). At this stage the vesicles show no internal structure, are bounded by a well-defined unit membrane, and are about 50 m/z in diameter. The vesicles disperse, liberating filamentous material which eventually fills the vacuole (figs. 7, A; 8, A, B; 9, A). The fate of the vesicular membranes is not clear, although it is possible that their substance is transferred to the lining membrane of the vacuole as it expands (compare Whaley and Mollenhauer, 1963). The large vacuoles pass to the surface of the cell, which at this stage is highly convoluted, and release their contained material (fig. 7, B). The filaments of newly-secreted material are much less than 10 m/z in diameter. It seems not unreasonable to suggest that the filaments may contain chitin, for the Golgi system is known to play a part in the formation of some other polysaccharides, for instance in the growth of cell walls in plants (Mollenhauer, Whaley, and Leech, 1961; Whaley and Mollenhauer, 1963).

Electron micrographs suggest that the much ramified granular reticulum also contributes material to the procuticle. Cisternae appear not to open at the cell surface but to give rise also to small vesicles (figs. 6, c ; 7, B ; 9). These might pass directly to the surface or else contribute to the vacuoles described above. The granular membranes themselves arise from the outer membrane of the nucleus, as seen in fig. 9, A, B. Free ribosomes probably occur in the cytoplasm as well, as at an earlier stage.

The procuticle is thus secreted at least in part by a froth of vacuoles which rise to the surface of the cell. The peripheral reticulum associated with this bubbling is gradually withdrawn and the cytoplasm later forms simple lobes under the micropapillae. Newly-secreted fluid material becomes more compact and stains less intensely as it solidifies, a change which may reflect the bonding of chitin and protein molecules (see Picken, 1960).

A few other features of the epidermal cells may be noted. The nuclear envelope is usually lined by dense material corresponding to Feulgen-positive chromatin (fig. 9). Nuclear pores are few but clearly defined (fig. 9), and the apparent continuity between nuclear contents and the cytoplasm may be emphasized by indentations of the nuclear membrane (fig. 9, c). Nuclear changes have not been examined. The adjacent centriole appears to have a typical structure (fig. 9, D; see Gall (1961)).

In the distal cytoplasm pigment granules are identified as intensely osmiophil spheres bounded by a membrane (figs. 7, A; 8, c, E). They appear to develop from whorls of membranes (fig. 9, c) as in comparable pigment bodies (e.g. Dowling and Gibbons, 1962). Bundles of tonofibrils are common, and in tangential sections they may surround the nucleus to the exclusion of ‘undifferentiated’ cytoplasm. The fibrils are of considerable length and fairly uniform in section, perhaps 15 mp. in diameter (figs. 7, A; 8, c). They are associated with less dense, possibly fibrillar material and are unstriated. They look not unlike broad myofilaments, but their function is another matter. It has been noted that bundles of these fibrils are disposed like shuttlecock ribs within the cell, and they are thus seen on each side of intercellular junctions (figs. 8, E; 4, D). Tonofibrils could maintain cohesion between epidermal cells and the cuticle since they seem to be attached to it. It has not yet been seen whether they also reach the basal connective tissue. Should the cells change in height during the moulting cycle as in other arthropods (Wigglesworth, 1961 ; Sewell, 1955a) it is possible that the tonofibrils may somehow be concerned. At any rate the intercellular junctions are characteristically folded at the periphery and possess desmosomes (not clear in fig. 8, E) which would allow of such changes.

Collagen fibres of the underlying connective tissue differ altogether from tonofibrils. They are about 0-2 µ broad and show a periodicity of about 64 mp.. Fig. 8, D suggests that bands may sometimes recur at less than 60 mp.; it is possible, however, that the section was compressed during cutting.

The cuticle of Peripatus is renewed at ecdysis and contains chitin and protein. Except for spines, claws, and jaws it is unhardened. The surface is hydrofuge and has numerous papillae.

The observations reported here confirm that a procuticle and epicuticle are present. The epicuticle has 4 layers. The innermost is thought to be tanned and may contain bound lipid. The other peripheral layers are altogether less than 30 mp. thick. Consistently with the fact that wax and cement appear to be absent, the cuticle is penetrated by 50% ethanol although it retains hydrofuge properties at 90° C.

There are no dermal glands or pore canals, and present observations suggest that the arrangement of cuticular layers represents the order in which they are laid down. There is little ‘moulting fluid’ and the old cuticle does not change much before new epicuticle appears. By the time procuticle is secreted, the epicuticle looks complete. Materials for the procuticle are derived from the Golgi system and almost certainly from rough endoplasmic reticulum as well. Small vesicles contribute to larger vacuoles or else reach the surface independently. During active secretion the cell surface becomes reticulate, forming microvilli which are later withdrawn, but the procuticle probably continues to grow slowly until the next moult. Change of texture visible in electron microscope sections is interpreted as solidification of the chitin-protein secretion. Sclerotized regions in spines look yet more condensed in that they have no visible structure. Further details of sclerotization and of epicuticular structure would be valuable.

The evidence which has been described does not suggest that Peripatus has direct affinities with the annelids. The cuticle of the earthworm (few other annelids have been examined) differs radically from that of an onychophoran. It consists largely of thick orientated collagen fibres, penetrated by numerous processes from the epithelial cells which reach the surface and may be continuous with a pile of microvilli there (Randall, 1957 ; Ruska and Ruska, 1961 ; Laverack, 1963). The earthworm cuticle can be regenerated but the animal does not moult. It is true that the setae contain /Lchitin, but chitin is absent from the non-collagenous residue of the cuticle as the polysaccharides found contain little or no nitrogen (Watson and Smith, 1956; Maser and Rice, 1962, 1963). While it would be desirable to confirm this evidence with polychaetes before taking it to indicate a general annelid condition, it is clear that the integument of Peripatus bears few resemblances to that of the earthworm.

It would seem rather that Peripatus is an arthropod whose cuticle shows few specialized features. In insects and Crustacea, for instance, a compound epicuticle covers an endocuticle and both layers are modified by phenolic tanning (and in the Crustacea often by calcification). In the terms appropriate to these groups (Wigglesworth, 1961 ; Locke, 1961 ; Beament, 1961 ; Dennell, 1960) Peripatus also has an endocuticle (here called procuticle), the outer part of which can form exocuticle by sclerotization. In this account hardened prosclerotin has been regarded as inner epicuticle, implying that it might perhaps foreshadow a cuticulin layer. Exocuticle would be developed only in the spines, claws, and jaws.

On general grounds the Onychophora are regarded as most closely related to myriapods (Tiegs, 1940, 1945, 1947; Tiegs and Manton, 1958). Blower (1951) interprets the myriapod cuticle in terms of sclerotization of an exocuticle present even in intersegmental membranes. This layer would corre- spond to what has been termed prosclerotin in Peripatus, a word in fact coined by Blower to describe it. It is very thick in the scutes and the outer- most region may become so hardened that it is resistant to staining and may be termed sclerotin. Tanning may extend into the endocuticle below, which corresponds to the procuticle in Peripatus. In Blower’s terms, Peripatus has a prosclerotin exocuticle and an endocuticle which becomes tanned in claws and spines. Blower considers that the myriapod exocuticle is not homo- logous with the epicuticle of insects because it contains chitin. Apart from this feature the exocuticle is impregnated with lipid and might perhaps have some of the properties of cuticulin.

Pore canals seem to be a general arthropod feature associated with seg- mental sclerotization. Apart from the Onychophora, which in this respect appear unspecialized, it is possible that they may be lacking in centipedes (Blower, 1951), and they are secondarily absent in the larvae of some higher pterygotes (Dennell, 1946). The microvilli of Peripatus are, however, not un- like the cytoplasmic filaments of pore canals (Locke, 1961; Noble-Nesbitt, 1963 a, b; Kawaguti and Ikemoto, 1962) except that they withdraw from the procuticle as it hardens. As regards tanning in Peripatus there is as yet no information but there seems no reason to suspect that a phenolic process does not occur as in other arthropods (Wigglesworth, 1957b; Dennell, 1958, 1960; Hackman, 1959). It is worth noting that the minimal activity of the ‘moulting fluid’ in Peripatus is unusual, since in general, unmodified endocuticle is dis- solved by enzymes before ecdysis and reabsorbed.

In most arthropods the outer layers of cuticle contain lipid, which is con- tributed by tegumental glands or specialized haemocytes (Kramer and Wigglesworth, 1950; Blower, 1951; Sewell, 1955a; Skinner, 1963). The cuticle of Peripatus contains little lipid. Cells corresponding to oenocytes have not been seen, and substances from the haemocoel must diffuse through a thick layer of connective tissue before reaching the epidermis. Insects and some spiders and ticks (Sewell, 1955a; Lees, 1947) achieve a hydrofuge sur- face (and reduced water loss) by means of a peripheral layer of wax. The presence of wax does not indicate evolutionary affinities, but its absence in Onychophora may be regarded as a non-advanced condition. The hydrofuge properties of terrestrial arthropods without wax probably depends on tanned proteins in the epicuticle (Davies and Edney, 1952; Edney, 1957) associated with highly oriented lipid (Beament, 1961).

The myriapod cuticle has not yet been studied with the electron microscope, but there is a thin sudanophil epicuticle above the exocuticle (Blower, 1951) which may correspond to the outer epicuticle of other forms. The composi- tion of epicuticular layers in Peripatus is not yet known, although the first appearance of the outer lamina as a double line is perhaps suggestive of lipid monolayers (p. 291 ; fig. 5, A, c). It is not inconceivable that if the substance of this lamina consists of lipoprotein, the hydrofuge properties of the surface may perhaps depend on a monolayer of polar molecules oriented with respect to such a substrate (Beament, 1961 and personal communication). A system of this kind would be likely to lose its hydrofuge properties on contact with 50% ethanol (or with any agent disorganizing the underlying protein) but would be little affected by heat (p. 286).

Morphological details of the insect epicuticle show some likeness to those of Peripatus. Locke (1961) finds a darkly-staining line 8 mp. above the cuti- culin layer in Galleria larvae, which would also be consistent with the pre- sence of oriented lipid, and Noble-Nesbitt’s electron micrographs of Podura (1963 A, b) suggest a similar arrangement. But by comparison even to aptery- gote insects the Onychophora appear primitive. Noble-Nesbitt (1963 a, c) has shown that the hydrofuge properties of Podura cuticle depend on the wax capping a large number of small surface tubercles about 0 · 4 μ. apart. The intervening surface is wettable. So long as the wax on the tubercles is intact, only fluids whose contact angle is less than critical will penetrate. The micro- papillae of Peripatus present a similar situation in that the air film enveloping a submerged animal is evidently trapped between them. The tips of the micropapillae, however, seem little differentiated except in the specially firm attachment of the outer epicuticle. It may be that a particular contact angle combined with the characteristic shape of the papillae (fig. i, B) suffices to account for non-wettability. It is tentatively suggested that 50% ethanol penetrates because it has a lower contact angle than water and that it also alters an oriented structure so that the hydrofuge properties of the cuticle are destroyed. There are instances of micropapillae in the presumably hydrofuge arthrodial cuticle of myriapods (Lagner, 1937) and spiders (Sewell, 1955), so that structures of this kind are liable to arise by convergence.

Peripatus emerges from this study as a real arthropod. Features of the integument which appear primitive have lent themselves to the elaboration of flexibility in the body wall in relation to the cryptic habits of the Onycho- phora. The numerous sense organs, the tonofibrils, and the connective tissue strands anchoring epidermal cells to the muscle-fibres beneath would seem to be of particular adaptive value (see Manton, 1961). Besides being essential for the animal’s movements, the thick layer of subdermal connective tissue is probably a primitive feature associated with non-sclerotization.

The secretion of the procuticle resembles the process by which chitinous cuticles or membranes appear to be produced in most other arthropods : that is to say, chitin and protein produced in cytoplasmic vesicles is liberated in a fluid state at the surface and the new cuticle is solidified as bonds form between chitin and protein molecules (Wigglesworth, 1930; Picken, 1960; Rudall, 1963). If the formation of the cuticle can be shown to involve basically similar processes in all arthropods, some of the endocrine changes controlling ecdysis may also prove to be alike. The notable advances which have been made in this field concern mainly the insects and Crustacea (see Wiggles- worth, 19596, 1963; Passano, 1960). There is much less information about arachnids, myriapods, or the Onychophora, and these groups provide a wealth of material for future investigation.

I wish to thank Professor J. H. Day and members of the Zoology Depart- ment, University of Cape Town, for the opportunity to begin this work; Professor S. F. Bush and colleagues in the Zoology Department, University of Natal, for generous help with facilities and material; and Professor B. I. Balinsky and members of the Zoology Department, University of the Wit- watersrand, for further assistance. I am indebted to Dr. A. V. Grimstone for help in learning the methods of electron microscopy, and to Dr. J. W. L. Beament for suggestions and very helpful criticism. A grant from the Royal Society towards the cost of the plates is gratefully acknowledged.

Alexander
,
A.
, and
Ewer
,
D. W.
,
1955
.
Ann. Natal Mus
.,
13
,
217
.
Andersen
,
S. O.
,
1963
.
Biochim. Biophys. Acta
,
69
,
249
.
Baker
,
J. R.
,
1944
.
Quart. J. micr. Sci
.,
85
,
1
.
Balfour
,
F. M.
,
1883
.
Quart. J. micr. Sci
.,
23
,
213
.
Beament
,
J. W. L.
,
1945
.
J. exp. Biol
.,
21
,
115
.
Beament
,
J. W. L.
,
1960
.
Nature
,
186
,
408
.
Beament
,
J. W. L.
,
1961
.
Biol. Rev
.,
36
,
281
.
Blower
,
G.
,
1951
.
Quart. J. micr. Sci
.,
92
,
141
.
Boelitz
,
E.
,
1933
.
Zool. Jb., Abt. Anat
.,
57
,
375
.
Bouvier
,
E-L.
,
1905-7
.
Ann. Sei. nat. (Zool
.),
2
,
1
, and 5, 6r.
Brink
,
P.
,
1957
.
Onychophora. In South African Animal Life
, Vol.
4
, p.
7
.
Stockholm
(
Almqvist and Wicksell
).
Bursell
,
E.
, and
Ewer
,
D. W.
,
1950
.
J. exp. Biol
.,
26
,
335
.
Butt
,
F. H.
,
1959
.
Smithson, mise. Coll
.,
137
,
43
.
Butt
,
F. H.
,
1960
.
Biol. Rev
.,
35
,
43
.
Clark
,
A. H.
,
1915
.
Smithson, mise. Coll
.,
65
,
1
.
Cloudsley-Thompson
,
J. L.
,
1958
.
Spiders, scorpions, centipedes and mites
.
London
(
Pergamon
).
Dakin
,
W. J.
,
1920
.
Quart. J. micr. Sei
.,
65
,
163
.
Dakin
,
W. J.
,
1922
.
Ibid
.,
66
,
409
.
Davies
,
M. E.
, and
Edney
,
E. B.
,
1952
.
J. exp. Biol
.,
29
,
571
.
Dennell
,
R.
,
1946
.
Proc. Roy. Soc. B
,
134
,
79
.
Dennell
,
R.
,
1958
.
Biol. Rev
.,
33
,
178
.
Dennell
,
R.
,
1960
.
Integument and exoskeleton. In The physiology of Crustacea
, Vol.
1
, edited by
T. H.
Waterman
, p.
449
.
New York
(
Academic Press
).
Dodds
,
S. E.
, and
Ewer
,
D. W.
,
1952
.
Ann. Natal Mus
.,
12
,
275
.
Dowling
,
J. E.
, and
Gibbons
,
I. R.
,
1962
.
J. Cell Biol
.,
14
,
459
.
Duboscq
,
O.
,
1920
.
Arch. Zool. exp. gén
.,
59
,
21
.
Edney
,
E. B.
,
1957
.
The water relations of terrestrial arthropods
.
Cambridge
(
University Press
).
Gabe
,
M.
,
1954
.
C.R. Acad. Sci
.,
238
,
272
.
Gall
,
J. G.
,
1961
.
J. biophys. biochem. Cytol
.,
10
,
163
.
Gibbons
,
I. R.
, and
Grimstone
,
A. V.
,
1960
.
Ibid
.,
7
,
697
.
Glauert
,
A. M.
,
1962
.
J. roy. micr. Soc
.,
80
,
269
.
Guilding
,
L.
,
1826
.
Zool. Journ
.,
2
,
443
.
Hackman
,
R. H.
,
1959
.
Biochemistry of the insect cuticle. In Biochemistry of insects
, edited by
L.
Levenbook
, p.
48
.
London
(
Pergamon
).
Holliday
,
R. A.
,
1942
.
Ann. Natal. Mus
.,
10
,
237
.
Holliday
,
R. A.
,
1944
.
Ibid
.,
10
,
433
.
Hutchinson
,
G. E.
,
1930
.
Proc. U.S. nat. Mus
.,
78
, part. 11,
1
.
Hutton
,
F. W.
,
1876
.
Ann. Mag. nat. Hist
.,
18
,
361
.
Juberthie-Jupeau
,
L.
,
1963
.
Arch. Zool. exp. gén
,
102
,
1
.
Kawaguti
,
S.
, and
Ikemoto
,
N.
,
1962
.
Biol. J. Okayama Univ
.,
8
,
43
.
Kennel
,
J. von
,
1885-8
.
Arb. zool. Univ. Würzburg
,
7
,
95
; 8, 1.
Kramer
,
S.
, and
Wigglesworth
,
V. B.
,
1950
.
Quart. J. micr. Sci
.,
91
,
63
.
Kunike
,
G.
,
1925
.
Z. vergl. Physiol
.,
2
,
233
.
Lagner
,
E.
,
1937
.
Zool. Jb. (Abt. Anat
.),
63
,
483
.
Lankester
,
R.
,
1904
.
Quart. J. micr. Sci
.,
47
,
523
.
La verack
,
M. S.
,
1963
.
The physiology of earthworms
.
London
(
Pergamon
).
Lawrence
,
R. F.
,
1931
.
Ann. S. Afr. Mus
.,
30
,
101
.
Lawrence
,
R. F.
,
1953
.
The biology of the cryptic fauna of forests
.
Amsterdam
(
Balkema
).
Lees
,
A. D.
,
1947
.
J. exp. Biol
.,
23
,
379
.
Lees
,
A. D.
, and
Waddington
,
C. H.
,
1942
.
Proc. Roy. Soc. B
,
131
,
87
.
Locke
,
M.
,
1961
.
J. biophys. biochem. Cytol
.,
10
,
589
.
Lotmar
,
W.
, and
Picken
,
L. E. R.
,
1949
.
Experientia
,
6
,
58
.
Luft
,
J.
,
1961
.
J. biophys. biochem. Cytol
.,
9
,
409
Manton
,
S. M.
,
1937
.
Phil. Trans
.,
227
,
411
.
Manton
,
S. M.
,
1938
.
Ann. Mag. nat. Hist
.,
1
,
515
.
Manton
,
S. M.
,
1949
.
Phil. Trans
.,
233
,
483
.
Manton
,
S. M.
,
1950
.
J. Linn. Soc. (Zool
.),
41
,
529
.
Manton
,
S. M.
,
1953
.
Soc. exp. Biol. Symp
.,
7
,
339
.
Manton
,
S. M.
,
1958
.
J. Linn. Soc. (Zool
.),
44
,
58
.
Manton
,
S. M.
,
1959
.
System. Assoc. Publ. no
.
3
,
23
.
Manton
,
S. M.
,
1960
.
Biol. Rev
.,
35
,
265
.
Manton
,
S. M.
,
1961
.
Experimental zoology and problems of arthropod evolution. In The cell and the organism
, edited by
J. A.
Ramsay
and
V. B.
Wigglesworth
, p.
234
.
Cambridge
(
University Press
).
Manton
,
S. M.
, and
Ramsay
,
J. A.
,
1937
.
J. exp. Biol
.,
14
,
470
.
Maser
,
M. D.
, and
Rice
,
R. V.
,
1962
.
Biochim. biophys. Acta
,
63
,
255
.
Maser
,
M. D.
, and
Rice
,
R. V.
,
1963
.
J. Cell Biol
.,
18
,
569
.
Mendes
,
E. G.
, and
Sawaya
,
D.
,
1958
.
Rev. Brasil. Biol
.,
18
,
129
.
Mollenhauer
,
H. H.
,
Whaley
,
W. G.
, and
Leech
,
J. G.
,
1961
.
J. Ult. Res
.,
5
,
193
.
Morrison
,
P. R.
,
1946
.
Biol. Bull
.,
91
,
181
.
Moseley
,
H. N.
,
1874
.
Phil. Trans
.,
164
,
757
.
Noble-Nesbitt
,
J.
,
1963a
.
Quart. J. micr. Sci
.,
104
,
253
.
Noble-Nesbitt
,
J.
,
1963b
.
Ibid
.,
104
,
369
.
Noble-Nesbitt
,
J.
,
1963c
.
J. exp. Biol
.,
40
,
681
.
Owen
,
G.
,
1959
.
Quart. J. micr. Sci
.,
100
,
421
.
Paclt
,
J.
,
1956
.
Biologie der primar fliigellosen Insecten
.
Jena
(
Fischer
).
Pantin
,
C. F. A.
,
1948
.
Notes on microscopical technique for zoologists
. 2nd imp.
Cambridge
(
University Press
).
Passano
,
L. M.
,
1960
.
Molting and its control. In The physiology of Crustacea
, Vol.
I
, edited by
T. H.
Waterman
, p.
473
.
New York
(
Academic Press
).
Peachey
,
L. D.
,
1959
.
J. biophys. biochem. Cytol
.
5
,
511
.
Picken
,
L. E. R.
,
1960
.
The organization of cells and other organisms
.
Oxford
(
Clarendon Press
).
Pryor
,
M. G. M.
,
1940
.
Proc. Roy. Soc. B
,
128
,
293
.
Randall
,
J. T.
,
1957
.
J. cell. comp. Physiol
.,
49
,
113
27
.
Richards
,
A. G.
,
1951
.
The integument of arthropods
.
Minneapolis
(
Univ. Minnesota Press
).
Richards
,
A. G.
,
1958
.
Ergebn. Biol
.,
20
,
1
.
Rudall
,
K. M.
,
1955
.
Soc. exp. Biol. Symp
.,
9
,
49
.
Rudall
,
K. M.
,
1963
.
Adv. Insect Physiol
.,
1
,
257
.
Ruska
,
C.
, and
Ruska
,
H.
,
1961
.
Z. Zellforsch
.,
53
,
75g
.
Sanchez
,
S.
,
1958
.
Arch. Zool. exp. gén
.,
96
,
57
.
Scheffel
,
H.
,
1961
.
Zool. Jb. (Anat
.),
79
,
529
.
Scheffel
,
H.
,
1963
.
Zool. Jb. (Physiol
.),
70
,
284
.
Schneider
,
K. C.
,
1902
.
Lehrbuch der vergleichender Histologie der Tiere
.
Jena
(
Fischer
).
Sedgwick
,
A.
,
1885-8
.
Quart. J. micr. Sci
.,
25
,
449
; 26, 175; 27, 467; 28, 373.
Sewell
,
M. T.
,
1955a
.
Ann. ent. Soc. Amer
.,
48
,
107
.
Sewell
,
M. T.
,
1955b
.
Quart. J. micr. Sci
.,
96
,
73
.
Sjostrand
,
F. S.
,
1956
.
Electron microscopy of cells and tissues. In Physical techniques in biological research. Vol. 3. Cells and tissues
, edited by
G.
Oster
and
A. W.
Pollister
, p.
241
.
New York
(
Academic Press
).
Skinner
,
D. M.
,
1963
.
Biol. Bull
.,
123
,
635
.
Snodgrass
,
R. E.
,
1938
.
Smithson, mise. Coll
.,
97
,
1
.
Snodgrass
,
R. E.
,
1952
.
A textbook of arthropod anatomy
.
Ithaca
(
Comstock
).
Steel
,
T.
,
1896
.
Proc. Linn. Soc. N.S.W
.,
21
,
94
.
Tiegs
,
O. W.
,
1940
.
Quart. J. micr. Sci
.,
82
,
1
.
Tiegs
,
O. W.
,
1945
.
Ibid
.,
85
,
191
.
Tiegs
,
O. W.
,
1947
.
Ibid
.,
88
,
165
.
Tiegs
,
O. W.
, and
Manton
,
S. M.
,
1958
.
Biol. Rev
.,
33
,
255
.
Walcott
,
C. D.
,
1911
.
Smithson, mise. Coll
.,
57
,
109
.
Walcott
,
C. D.
,
1931
.
Ibid
.,
85
, no.
3
,
i
.
Watson
,
J. A. L.
,
1962
.
Dissertation for Ph.D. Degree, Cambridge
.
Watson
,
M. L.
,
1958
.
J. biophys. biochem. Cytol
.,
4
,
475
.
Watson
,
M. R.
, and
Smith
,
R. H.
,
1956
.
Biochem. J
.,
64
,
10P
.
Whaley
,
W. G.
, and
Mollenhauer
,
H. H.
,
1963
.
J. Cell. Biol
.,
17
,
216
.
Wigglesworth
,
V. B.
,
1930
.
Quart. J. micr. Sci
.,
73
,
593
.
Wigglesworth
,
V. B.
,
1957a
.
Proc. Roy. Soc. B
,
147
,
185
.
Wigglesworth
,
V. B.
,
1957b
.
Ann. Rev. Entom
.,
2
,
37
.
Wigglesworth
,
V. B.
,
1959a
.
Quart. J. micr. Sci
.,
100
,
315
.
Wigglesworth
,
V. B.
,
1959b
.
The control of growth and form
.
Ithaca
(
Cornell University Press
).
Wigglesworth
,
V. B.
,
1961
.
The epidermal cell. In The cell and the organism
, edited by
J. A.
Ramsay
and
V. B.
Wigglesworth
, p.
127
.
Cambridge
(
University Press
).
Wigglesworth
,
V. B.
,
1963
.
J. exp. Biol
.,
40
,
231
.
Willey
,
A.
,
1898
.
Zoological Results
, Vol.
I
, p.
1
.
Willis
,
E. R-
, and
Roth
,
L. M.
,
1956
.
Ann. ent. Soc. Amer
.,
49
,
495
.
Wolfe
,
L. S.
,
1955
.
Quart. J. micr. Sci
.,
96
,
181
.