The Australasian Mycetophilid Glowworms

Transactions and Proceedings of the Royal Society of New Zealand, Volume 88, 1960-61, Page 577

The Australasian Mycetophilid Glowworms By J. Bronte Gatenby (From Trinity College, Dublin, and Sydney University, N.S.W.) [Received by the Editor, February 19, 1960.]. Abstract An account of Australasian glowworm larvae, and of the Tasmanina and New Zealand male adults. The metamorphosis of the reflector and the tracheoles and nerves of the light organ are discussed. In various writings on the New Zealand glowworm, it is stated that similar larvae exist in Australia and Tasmania. While passing through Australia, the present author had the opportunity of examining some glowworms from the Blue Mountains region, it would have been desirable to have spent more time in different localities on the study of the live glowworms, in which the colour of the various parts could be noted Fixed specimens are useless for this purpose. It will be interesting to compare the imagines of the N.Z. and N.S.W. forms, when the Australian glowworm has been bred through to the adult phase. For helping in various ways, the author thanks Professor P. D. F. Murray and Dr. A. R. Woodhill, of the Entomology Department, Sydney University; Dr. A. J. Nicholson, Chief of the Division of Entomology of the Commonwealth Scientific and Industrial Research Organisation, Canberra; Dr. J. W. Evans (Director) and Mr. David McAlpine (Assistant) of the Australian Museum, Sydney; and Mr. David Lee, of the School of Hygiene and Tropical Medicine of Sydney University Dr. H. E. Hinton, of the Entomology Department of Bristol University, kindly read the first draft of this paper and made suggestions Professor V. Wigglesworth, F.R.S, was good enough to examine the photomicrographs on Pl 42, figs. 14 and 16, and to give his opinion on the question of the tracheoles, a doubtful and difficult matter needing special techniques for its solution. The micrographs from which Text-fig. 2 was drawn were used by permission of Mr. Ward, Director of the Dominion Physical Laboratory, Wellington, N.Z. List OF Material (1) B. tasmaniensis. 1 male adult, tarsus of each second leg missing. Wings tattered Several larvae, somewhat macerated This material had been collected at the Ida Bay Caves in 1936 by Dr. J. W. Evans, of Sydney, and stored in alcohol. (2) B. luminosa (NZ). 1 male (Waitomo Caves). 3 females (2 Waitomo Caves, 1 bred by Dr. Simon Cotton and writer). 3 pupae, one partly crushed. Numerous larvae, whole mounts and sections. (3) Australian Form (N.S.W.). Several larvae (Dr. A. R. Woodhill and Mr. D. McAlpine). (4) Sections of the end of abdomen of N.Z. male and female adult (see Gatenby, 1959). All this material except the sections suffered somewhat during the long sea route from New Zealand and Australia. As will be realized, this material is inadequate. No cave larvae of the N.S.W. species have been studied; though the N.S.W. larvae seen lived in situations somewhat like those where N.Z. larvae were collected. The relative positions of the organs drawn freehand in Pl. 39, fig. 3, depend somewhat on the contraction or

stretching of the body. In the case of Fig. 3, the larva was beginning to turn to the right and was stretched in front. Comparison between the cuticular parts of larvae from N.Z., N.S.W., and Tasmania, proved disappointing as taxonomic evidence. No really good hot hydroxide or lactic acid preparations of larvae were made, but Faure mounts of unstained larvae were adequate. The N.Z. female from which Fig. 12 was drawn is unusually small (10 mm body and head length), instead of 17 mm. It had been well fed, but chose to pupate sooner than expected, in early autumn. The room in which it was kept was heated only occasionally, and it must have experienced cold nights. The outer genitalia of the N.Z. male could not be studied in toto, because the hind segments had been removed for sectioning (see Gatenby, 1959, Plate 27, fig. 11). Except for Figs. 3, 12B and 12C, all the drawings were made with a camera lucida. It was not desirable to pull off individual legs to get the lengths of parts in all cases exactly right, but every care was taken to do this from the mounted specimen. Part of the missing region of the middle leg in Pl. 41, fig. 11, has been dotted in. This paper, and previous papers (Gatenby, 1959; and Cotton, 1960) on the Genus Bolitophila have been written from the aspect of comparative anatomy. The taxonomist of the Mycetophilidae may find that the writer unwittingly has left out data which would appear important in taxonomy. It is hoped that the drawings are adequate and accurate enough to enable the taxonomist to find the information he may desire. It is unfortunate that the legs of the only N.Z. male were broken in transit. There is, however, a good photograph of a N.Z. male in the Waitomo pamphlet, the original negative being in the archives of the N.Z. National Publicity Studios. On Pl. 41, fig. 11B the coxa and femur of the metathoracic leg of the N.Z. male is drawn at the same magnification as the Tasmanian male for comparison. Unfortunately it has not been possible to compare the prothoracic legs of the two males, in order to confirm Ferguson's statement that the ratios of the five metatarsi and tibiae of the two species are different. The N.Z. male was so much larger than the Tasmanian form that it was not possible to draw it at the same magnification within the space used for Plate 41, fig. 11. In his description of the new species B. tasmaniensis, Ferguson does not mention the fundamental fact of the general difference in build, of the N.Z. and Tasmanian species. In fact, so far as the writer is able to understand Ferguson's description of his new species, this would apply to the N.Z. male as well, save for the length of the antenna and difference of the relative lengths of the fore metatarsi and tibiae which the writer was unable to test. Ferguson states that there are two basal joints and fifteen flagellar joints on the antenna of A. tasmaniensis— that is, seventeen altogether. In both the Tasmanian and the N.Z. male examined the present writer is quite certain that there are only sixteen altogether. Technique For the benefit of those who may desire to begin work on mycetophilids, the following notes may be useful. The brightest whole mounts were made with Gilson & Bouin fixation, followed by Borax Carmine. There is no doubt that Mayer's acid haemalum would be better, but it was not available. Paracarmine is usually superior to Borax Carmine, but Gower's Carmine is worth trying. The most penetrating fluid is Carnoy, but as soon as the larvae or pupae are fixed, which only takes about one hour, they should be got into 90% alcohol. For externals it is best to drop the larvae into 70% alcohol, and some will die stretched out. Mounted straight from water in Faure's medium, these show the combed areas well. The usual silver and gold nerve methods failed, and these will need adjusting for the purpose. Observations on colour must be made with live larvae. For cytology of the cytoplasm, Sjövall's method (neutral formalin followed by controlled osmication) has given good preparations. Mann-Kopsch is recommended as well. Dried preparations

of the snare stained well in Leishmann's blood stain, and were mounted straight in balsam after drying. Recently Wigglesworth (1959) has successfully described insect nerve histology by using his new osmium/ethyl gallate method, which can be recommended for those who intend to study the nervous system of B. luminosa, and which certainly could not be as capricious as the gold and silver methods. The Snare In a previous paper (Gatenby, 1959) the relations of the mucus droplets and the vertical fishing lines was described. The snare of the N.S.W. larva has been examined with reference to this point. The snare was pulled from under the rocky shelf and immersed in 70% alcohol, and subsequently stained for one hour in Mann's methyl blue eosin. The silken threads vary from very fine ones known to from the vertical pendent threads, and others of a coarser nature, which probably take part in the formation or suspension of the horizontal runway. In some cases silk threads are embedded in flat elongate bodies which appear to be mucus, but there is no evidence that the larva could make a definite mixture of mucus and silk. The nature of these elongate bodies, Pl. 40, fig. 5, has been the subject of doubt At first they were thought to be pieces of rotten wood with spiral vessels. When, however, they were isolated and pulled out with forceps, the coils were found to be silk. How the sometimes quite close spring-like threads are spun and become embedded in mucus is not known. One could understand how silk of different thicknesses could be spun, but the flat bodies of mucus in which coiled threads of silk are embedded, would need especially thick mucus for this purpose. The watery mucus droplets on the vertical lines do not fix as such in alcohol—to show them, one must dry the droplets on a cover slip, and then the dry mucus content will stain well. Though there is no evidence, it is likely that the larva can control the water content, that is the viscosity, of the mucus it secretes. Further work by those who can get the material is necessary. Leishman's blood stain gives good preparations of dried mucus and silk (Gatenby, 1960). There are four regions of the snare to be considered (1) the runway; (2) the vertical lines; (3) the tunnel region; (4) the supporting lines of the runway. The vertical lines have already been figured from Leishman preparations (Gatenby, 1960). It is not known whether the hiding place is lined with silk and/or mucus. Preparations of the runway of the N.Z. snare were lost, but from cursory examination, the bodies shown in Pl. 40, fig. 5, were not noticed at the time. The Segmentation AND External Morphology OF The Larva In the previous paper (Gatenby, 1959) the anatomy of the larva was described from specimens which had been fixed when slightly pressed under a coverslip. This was necessary in order to keep them straight, and to see throught the final preparation. This type of mount does not show the externals properly and some useful facts were overlooked in consequence. In Pl. 39, fig. 1, from an alcohol specimen (Tasmanian) the segmentation is shown viewed from the dorsal surface, scale on the left hand side. There are the three thoracic segments, then three smooth abdominal segments which would appear nearly circular in cross section, then three with distinctly flattened surface, raised to ridges laterally. These three segments have deep cross folds. Following these segments is one (the seventh) without very distinct cross folds or ridges, and finally there is the eight or light organ segment, which is smooth.

The Hooked or Combed Areas of the Larva A diagram of these combed pads or areas has been given in the previous paper (Gatenby, 1959). The subject has been reviewed because Tasmanian and N.S.W. larvae became available, and it was hoped that specific differences between the three types of larvae might be discovered. In all larvae, the large mainly ventral combed area between the ultimate and penultimate segments is present, but it differs within members of the N.Z. species in its pigmentation and emphasis. Elucidation of the arrangement of the areas in front of segment seven is difficult because the segmentation of the larvae is usually not easy to make out, and the hooked areas may appear irregularly spaced owing to contraction, and are often difficult to see at all. The best specimens possessed by the writer were kindly examined by Dr. F. B. O'Connor, of this Department, but he could not establish that the arrangement of the bands was quite the same in each N.Z. larva. The combed areas stretch up from the ventral surface towards the lateral regions, leaving the dorsal surface of the larva without combs. In Pl. 40, fig. 7, is a plan of the segmentation of a N.Z. larva Beginning at segment eight (LO) there is the large hooked area (I) which is single, and which lies within and at the anterior edge of the eight segment. Passing forward one finds four paired areas (II-V), then two single areas (VI–VII). In front of this there are cross ridges in appropriate places, but hooks appear absent. There can be no doubt at all that the large area at (I) varies very considerably in different specimens of the N.Z. larva. This variation appears to be due not so much to large differences in the number of rows of hooks but to their size and pigmentation. In Pl. 40, fig. 10A, are two rows of hooks from two N.Z. larvae of the same stage of development. In the upper row the hooks are barely twice the size, and they are darker. The last double area is apparently in the fourth and abdominal segment and the two areas in front are single bands in nearly every case examined. There are usually from ten to twelve rows of hooks on each combed ridge. As shown in Pl. 42, Fig. 17, the hooks end forwards, and obviously are used to hold the weight of the animal when it reaches down to examine parts of the snare below the runway. In addition the hooks would assist the larva when it lies ventral surface down and is retreating backwards. By anchoring itself at the eighth segment it could rapidly pull its body into a contracted condition. Presumably, however, the hooks would impede the larva if it were retreating forwards ventral side down. On Pl. 40, fig. 9 is a plan from a Faure preparation (unstained) of the hooked areas in nearly full grown N.Z. larva. In this larva there were altogether six hooked regions. The roman numerals on the right correspond to the positions given in Pl. 40, fig. 7, save that in Fig. 7 there were seven hooked areas. In fig. 9, there were about thirteen rows of hooks on I, nine and eleven respectively on the two areas in II, the same but reversed in III and IV, and only eight in the sixth area (VI). The double areas in IV and V closely resembled their neighbours, but are not put in this figure. The groups of hooks varied from about 8–27 when not joined, and were sometimes staggered. The numbers (arabic) refer to the number of groups in the middle of each area; at the sides reaching up laterally the rows become reduced, so that each pad is spindle-shaped. It is interesting to note that in this specimen the group in the eight segment (I) is smaller than usual, and less pigmented. One Tasmanian larva mounted in Faure's medium has been examined for hooked areas. These are present, but smaller than in any N.Z. or N.S.W. larva.

The Sensory Setae of the Chordotonal Organs The setae have remained intact in only one of twenty-four whole stained mounts of the N.Z. larva. The photograph in Pl. 42, fig. 16S, of the curved seta, was made from a N.Z. larva which had been kept in alcohol. In fact, these setae are present in Dr. Evans' Tasmanian larvae kept in alcohol for nearly a quarter of a century. The setae apparently will become detached during the late stages of the staining and mounting routine, and must be examined in the fresh, or with flattened formalin and alcohol specimens. Owing to the lack of sufficient larvae of different ages of the three types, it is not possible to say for certain whether the Australian larva has longer setae than the N.Z. form, but it is believed to be so. The position of the short (Y) and longer (X) curved setae are shown in Pl. 40, fig. 8A. The longer seta curves upwards in this specimen, but in the opposite direction in Pl. 42, fig. 16, where the palp has been photographed ventral surface upwards. In all dead specimens, no matter how fixed, these papillae slope downwards as shown in the profile drawing of a Tasmanian lava in Pl. 40, figs. 8 and 8A. In other words, if the larva normally rests ventral side upwards, the papillae do not touch the runway, but are protruded upwards into the mucus in which the animal lies. The special advantage of this position is not known, but it may be suggested that the larva could notice disturbances in the snare over a wider area in a liquid medium dispersed over the runway. The runway is always wetted with mucus, so presumably this medium which extends out to the upper ends of the vertical lines acts as the carrier of vibrations. Position of the Larva on the Snare In a previous paper (Gatenby, 1959), it was stated that a captive larva reposed on its horizontal runway ventral surface upwards. This was ascertained by examining with a binocular dissecting microscope the position of the head, which bends down somewhat during life. The position occupied by wild larvae in the field is not known for certain. The reflector is ventral to the light organs, and the latter lie in a hollow deep saucer-shaped shell made by the reflector. This reflector is not transparent, and can be seen as a somewhat opaque white body in the living larva. It does not seem that light could shine through this maze of tracheoles, and for the larva to make the best of the light coming from the luminescent cells, the position necessary would be ventral surface upwards. By lying in this position, light could be reflected down on the beads of mucus, strung on the vertical fishing lines; thus increasing the area of luminescence. It was believed by Norris (see Hudson, 1950), that the larva really lived in a tube (of mucus). It is true that in the newly made snare of captive larvae the latter are covered by mucus, which could be regarded as a tube (see Gatenby, 1959), but it is not known whether larvae in the field are completely covered with a mucus layer, or whether they lie free on the runway. The writer did not manage to ascertain the facts in this matter, but from the behaviour of the larvae watched with the naked eye, it appears unlikely that the finished snare has a tubular runway with strong walls. It seems certain that the silk threads of the runway do not form a tube, but the runway is always damp with mucus, and if there were really a silk tube as exists in some spiders' snares, the larva would not be free to bend over anywhere along the runway as it certainly can, and reach down to recover prey. The conclusion is that during the night the larva assumes the most favourable position to reflect its light down on to the beads of mucus, and to do this it must have its ventral surface upwards or sideways. There are, however, other points. Presumably to hold on to the runway, it must turn partly or wholly ventral surface down, because the hooked bands lie mainly on the ventral surface.

That the larva normally lives in a mucus bed ventral surface upwards has, therefore, been concluded for the following reasons: (1) In this position the light can be directed downwards on the the snare, since the reflector is ventral. (2) The hooklets on the ventral surface are directed forwards and would tend to interfere with free movements of the larva except backwards. (3) In cases where a binocular microscope could be used on captive larvae lying on the newly formed snare, the position was ventral surface upwards. (4) Some N.S.W. larvae examined alive appeared to move freely under a loose coverslip either dorsal or ventral surface upwards. (5) The dorsal surface of the larva is flattened posteriorly so as to bend into the horizontal runway. The Transformation of the Reflector During Metamorphosis In a previous paper (Gatenby, 1959) it was noted that the reflector of the adult insect seemed different from that of the larva. New sections of a pupa have become available. In Pl. 40, fig. 10, of the pupa in a cut across the four light organs (LO 1-LO 4), there is a reflector consisting of apparently empty spaces. This appearance has been produced by a hollowing out, and some increase in the size of the normal larval reflector shown in Pl. 40, fig. 6. The phase photograph in Pl. 42, fig. 14, is from a fortunate section across the adult reflector. It answers two important questions, first, is the newly organized reflector connected to the tracheal system of the adult, and second, are the cells of the light organs in contact with nerves? The hollow reflector (R. upper) is seen to pass forwards into a longitudinal tracheal channel (TT), while at (AN) there is a fine nerve trunk passing between lobes of the reflector. The larval reflector consists of a tangle of extremely fine tracheoles forming a boat, the central part of its wall being occupied by a row of nuclei (Pl. 40, fig. 6N). In the pupal reflector in Pl. 40, fig. 10, the cells belonging to these nuclei are not seen inside the cavity of the reflector, and must have migrated to the wall of these spaces. The wall of the pupal reflector spaces is not well organized, but in that of the imago, there is a covering of cells which form a distinct cortex. The metamorphosis of the larval reflector into the hollow pupal and imaginal reflector is probably brought about by the activity of the central cells of the larval organ. This change must take place within not more than 24 hours. Not all the very fine tracheoles degenerate, because some are seen adherent to the light organ cells, as in Pl. 42, fig. 15 (T). The function of the reflector is supposed to be two-fold. Firstly, to bring oxygen to the light producing cells, and secondly to reflect light outwards. A familiar object which resembles the reflector of B. luminosa is a tracheal sac of Apis. The reflector of the glowworm becomes detached as a shiny light aluminium-grey thick walled boat, when dissected, and it floats to the top of the water in the dish. Nerve Endings and Tracheoles The luminescent organ is supplied with nerves passing towards or originating in the 7th abdominal segment. Nerves from this ganglion also are connected to the hind genital region (adult) and chordotonal organs (larval). The malpighian tubes emerge from the gut at the upper third of the abdominal segments, and it is curious that nerves from the penultimate segment should innervate the distal ends of the malpighian tubules which constitute the light ogran. Perusal of some of the relevant literature on insect nerve endings did not help the writer in regard to the possible relation of nerve fibres and luminescent cells in this larva. The material used had been fixed in Gilson or formalin, and the finer ramifications of the fibres could best be seen under phase contrast. The apparent coagulum (R) in the reflector is due to the use of phase contrast. By normal lighting there is seen to be a true cavity at (R).

On Pl. 42, figs. 14 and 15, are two phase contrast photomicrographs of the light organ region to show a supposed nerve ending and probable tracheoles. The difficulty is to distinguish between connective tissue, the smaller nerve fibres, and the tracheoles. For example, in Fig. 14, from an adult male, it is certain that the fibre (AN) is a nerve, because it has been followed back as it thickens to join the ganglion in the seventh segment. But the smaller fibres (TX) which appear only possibly to originate from the nerve (AN) could be, and probably are, tracheoles. In fig. 15, from an adult female, the nerve (AN) can be followed branching to give rise to the object (NE) which has been tentatively interpreted as a nerve ending. Likewise, the spindle shaped body to the left of the numeral 15 is also traceable to the nerve (AN). In the same photograph the tubes (T) are interpreted as tracheoles. From these photographs it can be concluded that fibres (AN) from the seventh ganglion, pass down and branch in close contact with the cells of the light organ, but the nature of the nerve endings is still in doubt. Text-fig. 1. — Plan of association of reservoir cavity (R), tracheoles (T), and supposed nerve endings (NE), with luminescent cell (LO). On the right the projections of the cell membrane (CL) into the cell (LO) are closed, on the left open (OP). (MIT), mitochondria; (AN), nerve fibre. The relative sizes of the diverticula from the cell membrane into the cell, and the other parts are considerably out of proportion. This hypothesis of light control is based on electron micrographs taken at the Dominion Physical Laboratory by Mr. W. Bertaud. . In Text-fig. 1 is a diagrammatic interpretation of the relations of nerves and tracheoles to the light organ cells. From electron microscopy, it is known that the cell membrane has deeply penetrating tubes and infolds, which, in larvae fixed during the day appear closed, as on the right of Fig. 1 at (CL), but have been drawn open as at (OP) on the left. It is suggested that the larva controls luminescence by cutting off the oxygen supply by closing the apertures (OP) under stimulus from the nerve ending (NE). It is possible that this might be tested by killing larvae at night whilst they are luminescent. Since, however, these infoldings of the cell membrane cannot be resolved by optical microscopy, this is a future task for electron microscopists.

Imaginal Discs In Pl. 39, fig. 2, and Pl. 40, fig. 7, a sketch is given of the main imaginal discs. Those inside the head capsule cannot be seen well in whole mounts but appear in sections In fig. 2, the six imaginal discs corresponding to the adult legs have been formed at L 1–L 3. By focusing the preparation, higher up are seen the discs belonging, in segment two, to the wings (W), and in segment three, presumably to the halteres. At this stage the two are nearly the same size. In Pl. 40, fig. 7, the discs relating to the hind abdominal region are shown at (HA). This region is photographed in a previous paper (Gatenby, 1959, Pl 26, fig. 12). The testes are found in the sixth abdominal segment (TS, in fig. 7). In Pl. 40, fig. 10, the male genital ducts are forming at (MG), and the intestine, still intact, has apparently shed its intima (INT), though the writer is not quite sure of the identity of this material. The Australian and Tasmanian Glowworms In the Australian region, the adult of the glowworm was described from the Ida Bay Caves of Tasmania by E. W. Ferguson in 1925. He considered this form was closely similar to the New Zealand species and named it Arachnocampa tasmaniensis. Ferguson had larvae and adults of both sexes. A male and three larvae of the Tasmanian form were given to the present author by Dr. J. W. Evans. These alcohol specimens collected many years ago were not in good order, but the larvae closely resemble those of the New Zealand form and are the same approximate size. Two small batches of the glowworms of New South Wales were examined alive. Dr. A. R. Woodhill's specimens came from Lithgow, in the foothills of the Blue Mountains. Those supplied by Mr. David McAlpine were found on steep damp rocky banks around a waterfall which it was said tended to dry up in summer. These glowworms were scarce. This situation was at Hazelbrook, also in the foothills of the Blue Mountains. Two snares examined by the writer seemed identical with those seen in New Zealand. Dr. Nicholson kindly supplied the present writer with a synopsis of knowledge of the N.S.W. form. He wrote: “Many years ago I became interested in glowworms when I saw them in a deep narrow ravine near Wentworth Falls. I made simple observations on their feeding habits but failed in my not very serious attempts to breed them. I am not sure whether the local species has been described, but the biology of the genus Ceroplatus (Mycetophilidae) was discussed by Skuse in a long taxonomic paper published in 1888 (Proc. Linn. Soc. N.S.W. 3: 1151), and he described Ceroplatus mastersi in this paper. He suggested that it had a luminous larva. Two years later (Proc. Linn. Soc. N.S.W. 5: 601) he mentioned it again, saying he had bred it from “luminous larvae” in decaying logs. This situation strongly contrasts with rock cracks in which I found “Glowworms”. Perhaps there is more than one species here. Edwards (Proc. Linn. Soc. N.S.W. 54: 174) in 1929 described C. mangalorensis from Tasmania, but gave no information on its biology. Skuse in 1890 (Proc. Linn. Soc. N.S.W. 5: 677) described the New Zealand species as Bolitophila luminosa. Tonnoir was very interested in the Mycetophilidae, and I seem to remember his saying that he considered the New Zealand and the Blue Mountain glowworms to be identical. This opinion is such a surprising one that I feel fairly confident about my recollection. Tonnoir's collection may throw some light upon this. It now belongs to the School of Tropical Medicine.” Dr. Nicholson's remark about Tonnoir's view as to the identity of the N.Z. and N.S.W. forms would explain why no special effort has been made in the past to breed out the adult of the N.S.W. form, as Skuse already had a New Zealand specimen from Hudson. Tonnior and Edwards (1926) mention that imagines of

Bolitophila luminosa are rare and found only in a few collections. This statement still applies today. Hudson (1955) and Gatenby and Cotton (1959) have been able to breed out adults in captivity. Directions for doing this have been given in previous papers by Gatenby (1959), Gatenby and Cotton (1959) and in a recent number of “Tuatara” (1959) by Gatenby. At Sydney University, Dr. Woodhill had, to begin with, shown the writer some lantern slides of photographs taken by Dr. A. J. Nicholson. One of these photographs of the larva was also reproduced in K. C. McKeown's “Insect Wonders of Australia,” 1935, and after examination of these photographs, the present author knew what to expect when he looked at living New South Wales glowworms. There are remarkable differences between the New Zealand form and the New South Wales specimens. The latter are more sharply and darkly pigmented and colourful, and the less pigmented segmental combed bands or areas appear to protrude more, so that especially in the lower region, the segments are very clearly marked when the larva moves. In the New Zealand species the comb band at the eight abdominal segment is the largest, whereas in the New South Wales form the combed areas in the 7in segment in the live specimen appeared to protrude more. Since the glowworm seems to live at times ventral surface upwards, the flattened dorsal side would be applied to the runway. The other differences are in colour and arrangement of pigmented areas. In the New Zealand form the fat bodies are greyish yellow, in the N.S.W. form they range from bright amethyst green to greyish green. The Lithgow forms are bright green, those from Hazelbrook are distinctly greenish. In the N.S.W. species from both localities, the pigmented areas beneath the integument are more regularly and sharply arranged and darker. The areas of pigment extend well into the 7th segment, where, in the N.Z. form they tend to become less dark. The mucus glands of the Lithgow specimens were chocolate colour, each cell being marked by a yellow central disc which, together with the ground colour, produced a café au lait effect. The specimens from Hazelbrook had light chocolate mucus glands, those from New Zealand were dark chocolate. In the N.S.W. specimens the malpighian tubes were greyish pink or brownish pink, in the N.Z. species they were pure pink. In the N.S.W. form the oesophageal valve was not as yellow as it is in the N.Z. form. These differences in colour might be partly due to pigments absorbed from the food; possibly the colour of the fat body may be due to the eating of greenish midges, but nothing is known about this. The extension of very dark coffee or raw sienna brown pigment areas into the 7th segment is unlike anything seen in the N.Z. form. On the basis of colour the two types of glowworms can be distinguished from each other by naked eye examination. The colours, as worked out above, were observed by a direct artificial spotlight on living glowworms, and it must be pointed out carefully that these differences are only seen properly in living larvae. Formalin specimens retain their colour rather badly for only a few days, and alcohol and Carnoy material is useless. Worms extended under coverslips do not show the shape of the body, and specimens dropped into 70% alcohol usually become contracted, and so far as the contour of the dorsal surface is concerned, are not always reliable material. Remarks E. W. Ferguson, in 1925, mentions that a Mr. Arthur M. Lea made the surprising statement that the Ida Bay Cave glowworms had two predators, a “spiderlike creature” (unidentified) and a carabid, Idacarabus. The present author found several species of true spiders in the Waitomo Caves, but these were near the opening of the cave, and it was concluded then, and is still believed, that such

spiders could not eat the larval glowworms. No beetles were found at Waitomo, and the backing of flood waters, cannabilism, and starvation are presumably the main factors in keeping down the numbers of glowworms in such situations in N.Z. At Arapuni, judging from remains on webs, no spiders were found attacking the glowworms on banks outside. It is probable from what we know of the voracity of carabid beetles, that Lea could be correct in stating that they could pull glowworms from their snare runways and eat them. The “spider-like creature” of Lea may be a harvestman (Phalangida), small specimens of which were found in the Waitomo Caves, but were not regarded as predaceous by the present writer; further observations may show that Lea was correct so far as concerns true spiders. The Tasmanian glowworms were collected a quarter of a mile within the Caves. Ferguson states that his species differs from B. luminosa chiefly by the relative lengths of the fore metatarsi and tibiae. Keith C. McKeown (1935) has something to say about the early history of our knowledge of the N.S.W. glowworm. He states that it was seen certainly as far back as 1804 by Caley (the elder) when he visited the Grose Valley (Blue Mountain area). It was eighty years later that Meyrick (see Hudson, 1950) became interested in the N.Z. form at Auckland. The position with regard to the glowworm of N.S.W. is that the adult is unknown, so far as the present writer can ascertain. It seems very likely that had Tonnoir or Skuse bred out adults, this somewhat difficult feat would have been described in the literature. The breeding out of adults of the N.S.W. larva will be awaited with interest. It will be noted that the N.Z. female in Pl. 41, fig. 12, has eight abdominal segments, whereas the Tasmanian male in fig. 11 has nine. Mr. D. L. Lee (personal communication) has the following remarks to make on this: “The abdomen of the Ida Bay male has 8 visible segments, plus the genitalia. The genitalia are your 9th, but comprise the modified 9th and 10th segments. The anal cerci (usually called lamellae in papers on Mycetophilidae) are accessory to the ovipositor, and hence female structures are not to be found in the male. In this they differ from the characteristic cerci of more primitive orders such as Orthoptera, where cerci are found in both sexes as appendages of the 10th segment. In the Mycetophilidae the number of abdominal segments, exclusive of the genitalia may be either 7 or 8.” Mr. Lee gives the following leg measurements of an Ida Bay male:— Coxa. Femur. Tibia Tarsus I. II. III. IV. V. Front leg 24 50 64 115 47 28 17 9 Mid leg 25 56 85 88 40 23 13 9 Third leg 24 80 100 82 (missing) . Measurements in micrometer divisions; 21 divisions = 1 mm.

Comparison Between the N.Z. and Tasmanian Males The Tasmanian male is drawn on Pl. 41, fig. 11, and parts of the N.Z. male are given on Pl. 41, fig. 11B and Pl. 41, Figs. 12A and 12B and 13B. Having regard to the state of the material, three parts for comparison were chosen: (1) The lengths of the 3rd and 4th segments of the antenna. (2) The shape of the haltere. (3) The lengths and shape of the coxa and femur of the hind leg. In the first place the N.Z. male is a more heavily built insect (Text-fig. 2) and deeply pigmented. Dr. Evans' specimen of the Tasmanian male had been twentythree years in alcohol, had probably bleached somewhat but was throughout more yellowish than the N.Z. form. It is well to remark that in such cases colour is not always of certain value: for the N.Z. male was found near enough the opening of the cave to get some daylight, whereas the Tasmanian specimen was found in a darker place. In Pl. 41, fig. 11A, are five segments of the antenna of the Tasmanian male, and in Pl. 41, fig. 12A, four segments of the N.Z. male. Reference to segments 3 and 4 in each will show that the N.Z. male has thicker, heavier, shorter segments, the Tasmanian form longer and slenderer parts. In Pl. 41, fig. 12D, are camera lucida drawings of the lengths of the antennae of first, the Tasmanian male, and below the N.Z. male and female Both of the latter specimens were presumably from full grown larvae, and were captured in the Waitomo caves. There are sixteen segments in the antennae of all these specimens In fig. 12D, the natural contour of the antennae has been drawn. In Figs. 13A and B, are firstly the haltere of the Tasmanian male (A), and (B), that of either sex in the N.Z. form: here again the N.Z. form is the thicker and less graceful In fig. 11B of Pl. 41, the coxa and femur of the metathoracic leg are given for comparison with the same region in fig. 11 of the Tasmanian form. Ferguson does not state where he saw a N.Z. male. There is still doubt as to whether only one species of Bolitophila exists in N.Z., and the matter of comparisons between the N.Z. and the Tasmanian forms is not by any means settled. It is curious that Ferguson did not stress the fact of the remarkably delicate build of the Tasmanian male. Discussion None of the mycetophilid larvae described by Madwar (1937) closely resembles that of B. luminosa. Perhaps the nearest is Brachypeza radiata (Jenkinson), which however, has tracheal stigmata, and a body nearly cylindrical. Madwar describes the hooked areas (his locomotory pads) as lying across the segmental line, whereas in B. luminosa, the pad always appears to the present writer to lie in front of each segment. Madwar states that on that part of the pad in front of the segmental line the hooks point towards the head, and those on the pad behind the line point backwards. This is different from B. luminosa. According to Madwar the gastric coeca (mucus glands of the present author's papers) probably produce digestive enzymes. In Brachypeza these glands are poorly developed compared with B. luminosa. If Brachypeza could be regarded as more generalized, it would appear that the larva of B. luminosa has become specialized in the following: (1) The pad hooks point forwards. (2) The mucus glands are highly developed for the production of mucus, not apparently for secreting enzymes. (3) The shape of the body is cylindrical only in front, the hinder portion being more specialized. (4) There are no lateral chordotonal organs as in Brachypeza, so far as Ganguly and the present writer could ascertain. (5) Respiration is apneustic. (6) The anal papillae have elaborate chordotonal organs, which if present in other Mycetophilidae have not been found by authors previous to Ganguly. (7) The reflector and luminescent part of the milpighian tubes are absent in other mycetophilid larvae, though the

malpighian tubes of these larvae are carried down posteriorly. Dr. Gouri Ganguly kindly lent the writer her sections of the British Ceroplatus larva. In it the mucus (?) glands are poorly developed in comparison with the same organs in the N.Z. and N.S.W. larvae of Bolitophila; there are some doubtful cells in the anal papillae, which Dr. Ganguly believes may be sensory in nature. The section of the body is cylindrical, without dorsal folds as in B. luminosa (Pl. 39, fig. 1). The silk glands are very well developed in Ceroplatus. The malpighian tubes reach back as in other mycetophilids like Brachypeza. As is well known, this larva of Ceroplatus has luminescent fatty tissue cells, and leads the same sort of predaceous existence as B. luminosa. Future study on the genus Bolitophila in the Australian region should include the breeding out or capture of the imagines of the Blue Mountain localities. A further investigation of the ecology of the Ida Bay and other Tasmanian localities, especially with reference to the supposed predators, Idacarabus and cave spiders, would be rewarding. Captive N.S.W. larvae could be fed on house flies in order to see whether food alters the bright colour of the larvae How far these bright colours and dark pigment are due to the Australian sunlight is not known, as cave specimens from N.S.W. have not been examined by the writer for comparison. The chordotonal organs of the N.S.W. form appear to have longer protruding setae than the N.Z. form, but this observation needs confirmation. The sensory setae had become broken off on all the mounted N.S.W. larvae brought back to Dublin. The writer has had the privilege of discussing the relation between larval and adult forms in the Nematocera, with Mr. David Lee. He states that small larval differences will not necessarily appear in the adults. The modifications of the larva refer to that part of an insect's modus vivendi, and usually the imagines of different species, as would be the case with the glowworms, live similar lives and would not be prone to anatomical differences Mr. Lee showed the writer the large collection of Mycetophilidae at the N.S.W. School of Tropical Medicine. There are few adult mycetophilids from this region as large as the imago of glowworm. The Ceroplatidae seen were all smaller than B. luminosa. It is quite certain that the Tasmanian and the N.Z. insects are different species as based on observations of the male. The antennae in both N.Z. and Tasmanian males are larger than those of the female of the N.Z. insect Since it now seems to be true that the N.Z. female usually does not luminesce, the male no doubt finds his mate by the use of sensillae in these comparatively very large antennae. It will be remembered from previous reports (Gatenby, 1959; Gatenby and Cotton, 1960) that three live female N.Z. adults were never seen to light up, two having mated, one unmated. The only report of luminescence in the N.Z. female comes from Hudson (1955), whose observation cannot be doubted. The scarcity of glowworms in N.S.W. as compared with N.Z. is due to the fact that the climate in Australia is not so suitable. The drawing in Pl. 39, fig. 3, was made freehand from a living specimen, and was as accurate as the author could make it under these conditions. The last segment appears particularly swollen as compared with the previous drawing of the N.Z. specimen (Gatenby, 1959; Pl. 23, fig. 4), but in Fig. 7 in the same paper, something of the appearance of the present Pl. 39, fig. 3, is shown. The comparative size of this segment depends on the amount of haemocoelomic fluid directed into it, as the larva struggles under the coverslip. But this segment appears whiter and more translucent in the N.S.W. form because of the deeper pigmentation of the segments in front. The author got the impression that the mesenteron in the N.S.W. form was shorter than in the N.Z. form, but this will need more careful observation of freshly anaesthetised larvae of different sizes, but it is a point for future

Fig. 1—Tasmanian larva (about 23½ mm in length), from dorsal surface. Fig. 2.—Front end of N.Z. larva showing imaginal discs of legs, wings and halteres. Fig. 3.—Free hand sketch of N.S.W. larva, to show zones of colouration. Fig. 4.—Head of N.Z. larva, slightly from one side to show eyes and antenna base. Fig. 4A.-Part of section of head of larva showing the two types of eyes.

Fig. 5.—Part of snare of N.S.W. larva scraped from rock and fixed in 70% ethanol. Fig. 6.—Larval light organ and reflector in transverse section. Fig. 7.—Plan of externals of N.Z. larva, showing imaginal discs, testes, and hooked areas. Fig. 8.—Last three abdominal segments of Tasmanian larva from the side. Fig. 8A.-Position of short straight (Y) and long curved (X) chordotonal sensory setae (see Pl. 42, Fig. 16). Fig. 9.—Plan of hooked areas or pads of N.Z. larva. Fig. 10.—Transverse section of newly metamorphosed pupa showing hollow reflector and male ducts forming. Fig. 10A.-Row of hooks of two individuals from same N.Z. locality to show variation in size and pigmentation.

fig. 11.—The Tasmanian male. Fig. 11A.-First five segments of the antenna for comparison with those of the N.Z. male in Fig. 12A. Fig. 11B.-Coxa and femur of N.Z. male, hind leg Fig. 12-N.Z. female. Small specimen Fig. 12A.-Part of the antenna of N.Z. male. Figs. 12B and 12C.-External genital segment of N.Z. female. (12B from ventral surface.) Fig. 12D.-The contour and lengths of the Tasmanian and N.Z. adults antennae. fig. 13.—Halteres of the N.Z. and Tasmanian (A) forms.

fig. 14.—Course of nerve fibre (AN) towards the light origin (LO). Male Fig.-15.-Ditto in female. Note.-These photographs taken under phase-contrast, appear to show a background of congulum at (R) where the space is however, really quite empty under ordinary microscopic observation. fig. 16.—The curved external sensory seta (S) of the chordotonal organs. The base is seen where the seta emerges from the cuticle. The muscle of the palp and one of the overlying chordotonal organs can be seen sweeping across from upper right to lower left. fig. 17.—Photomicrograph of the hooks on the eight segment. These point forwards.

observers. In the same way the arrangement of the sense organs in the anal papillae depends somewhat on the amount of tension in the papillae, but the same four types of sensillae were present in both N.S.W. and N.Z. forms. Good fixed stained preparations of the internal sensory organs in the anal papillae are hard to make, only a small percentage of whole mounts being useful. Regarding the nervous system of the larva and imago of the N.Z. form, the following may be stated: Pyramidal prolongations of the nerve fibres (Pl. 42, fig. 15., NE) have been found abutting against the cells of the light organs, and for the time being these have been assumed to be the nerve endings. These, however, have a nucleus near where they adhere to the light organ cell. The actual nerve fibres have been traced to the ganglion in the seventh segment. It was not known what type of nerve ending to expect in connection with the light organ. There are least three branches emerging from the seventh segment ganglion, one to muscles, one to the light organ, and a large branch undoubtedly continuing down to the scolophores in the anal papilla on each side. In the adult as well, nerve ramifications from this ganglion, go to the hind gut and receptacula seminales. No muscles have been found in connection with the larval or adult light organ. It is significant, however, that the cell membrane of the light organ cells has been shown by electron microscopy, to be thrown into folds and tube-like indentations often of a succular nature (Text-fig. 2), and it is suggested that the light is dowsed by closure of the openings of these tubules, thus cutting off oxygen. In some of the preparations, it would appear that tracheoles penetrate the light organ cells, but the writer is very doubtful about this. There is now much information, derived from electron microscopy, on the ultra-structure of the light organ of the American coleopterous firefly (Photinus pyralis) and a certain amount is now known about the same organ in B. luminosa. The work on Photinus was carried out by Beams and Anderson (1955), and recently W. Bertaud, of the Dominion Physical Laboratory, has made a preliminary study of B. luminosa. The light organs of the coleopterous insects has been much studied by optical microscopy. The important paper by Beams and Anderson still leaves us in the dark as to how this insect controls its light. In Text-fig. 3 the present writer has attempted to interpret Beams and Anderson's excellent electron micrographs. It will be noted that the tracheal end cell (TEC) has concentric bodies in it, shown at a higher power in fig. 4. These bodies (M) were at one time thought to be contractile elements, and in some way were believed to compress the tracheal trunk (T) which passed through both the tracheal end cell, and the end bulb (EB), and thus to cut off the light which in Photinus flashes brightly at intervals as the insect flies. Beams and Anderson have now shown that the supposed contractile elements are really elongated mitochondria, each possibly in a saccule (F), and that all tracheoles are provided with taenidia (the internal supporting spiral), which presumably causes them to be incompressible. Taenidia in such tracheoles cannot be resolved by light microscopy and were not seen by previous observers of Photinus. In this insect, the reflector is a separate layer (R) containing urate bodies, and actually abuts against the columns of luminescent cells (LC), the space between the two in Text-fig. 3, being left for purposes of simplicity. Air enters at (TT) and presumably is used by the battery of mitochondria in (TEC) and (EB), but two tracheoles (T) continue past (EB) going on to lie between the various luminescent cells (LC, of which only one is shown). Thus it is evident from the work of Beams and Anderson, that not only do the tracheal end cells (fig. 4) partake of this air, but the luminescent cells also have their own supply (T) and there is no confirmation that this supply is cut off by controlled compression within either (TEC) or (EB). Beams and Anderson could not find nerves going to (TEC) or

(EB). We are forced to consider the bags (F) enclosing the mitochondria, certainly present in (TEC) and probably also in (EB), as a possible mechanism of control. But the relationship of the mouths of the bags, and the cell membrane abutting against the central tracheole is obscure. If there are infolds from the centre of the cell forming the bags, closure of their mouths might provide some sort of control. There are no infolds in the cell membranes of the photogenic cells of Photinus. Only in (TEC) and (EB) can such folds be identified as possible. Turning now to B. luminosa, we find a much simpler arrangement. In the first place, the luminescent cells in this mycetophilid do not appear to contain those photogenic granules, which are so conspicuous in Photinus (PG in Text-fig. 3). In the parts of the luminescent cells examined by Bertaud with the electron microscope, no such granules were seen. In Text-fig. 5, the reflector of B. luminosa is the tracheal supply as well, but swarms of mitochondria lie near the cell wall as in Photinus (Text-fig. 5, M). The infoldings of the cell membrane (MF in Text-fig. 5) have been referred to above. It should be noted that in both Text-figs. 3 and 5, the proportions of tracheal end cells, and luminous cells, and the infolds in the latter figure are quite out of proportion. The arrows point to the direction of light output, (EC) being the external cuticle of the beetle. It is well to notice that in the American firefly two factors have to be considered. First the turning on of the light, and secondly the mechanism of intermittent flashing. The latter does not occurr in B. luminosa. Only the question of the method by which the light is turned on and doused needs an answer. In Photinus the repeated bright flashing would need a considerable storage of material, which would become exhausted after a number of flashes had been emitted. But some workers might prefer to believe that the extraordinary richness

of mitochondria in the tracheal end cells may have some connection with this. Since the American authors have found no nerves going to the tracheal end cells, the mechanism of intermittent flashing is quite unknown, but the bag-like arrangement of the mitochondria might account for this if the tracheal end cell contracted rhythmically, allowing the escape of stored energy or some sort of stimulus sent across the cell membrane of the luminescent cells. Since the luminescent cells of Photinus are crammed with photogenic granules, the necessity for additional reinforcement from the tracheal end cells would seem unnecessary. It should be remembered that mitochondria observed in vivo are known to be able to elongate and contract quite quickly, but the hypothesis of compression of the tracheal leads would seem to be disproved definitely by Beams and Anderson. The same would apply to B. luminosa, where Bertaud's micrographs of the tracheal reflector of the larva show that the smallest tracheoles have taenidia, and the reflector contains no elements which could be contractile. In Bolitophila, the mechanism of control must be in the luminescent cell itself, and the cell membrane folds (MF in Text-fig. 5) are the only modification indicating a possible mechanism of control—a mechanism one may point out in this case, which is activated from the 7th abdominal ganglion. While the present writer hesitates to disagree with a worker of Beam's calibre, it seems surprising that no nerves have been found going to the tracheal end cells. It is still possible that the flashes in Photinus are triggered off in some way by nerves which have been overlooked by Beams and Anderson, and which presumably go to the tracheal end cells. As has been remarked, G. V. Hudson alone has seen luminescence in an adult—a female. But recent re-examination of the sectioned material (e. g., that in Pl. 42, Fig. 14) provides no reason for believing that anatomically, luminescence would be impossible. On the contrary, the tracheal system of the adult appears more efficient than that of the larva (Pl. 42, fig. 14, TT, for the large air channel which eventually connects to the tracheal stigmata). The capacity to luminesce presumably does not apply to the male where collapse and degeneration of the light organelles is evident, in one case at least. Past observers, as well as the writer, have noted that light shining on the beads of mucus droplets increases the area of luminescence, and to get this result the larva must throw its light downwards. For other reasons also it has been concluded that the glowworm rests ventral surface upwards. Proof of this in wild larvae can only be advanced by previously setting up a microscope in the field, arranged so as to give sufficient magnification to examine the head, and slope of the anal papillae, and then turning on a light source suddenly so as to observe the larva before it could escape. The use of coloured screens for the light source would be worth investigation—the reaction of the larvae to different colours not being known. The possiblilty that the size and arrangement of the segmental hooked areas might be useful taxonomically is at present doubtful owing to the fact that a number of larvae got from one situation in New Zealand showed considerable variation. A further examination of a larger number of larvae from N.S.W., N.Z., and Tasmania may, however, be helpful. The following problems need further research:— (1) The position of wild larvae on the snare. (2) The proportion of forward or backward retreats into the hiding place. (3) The study of parts of the snare wiped on cover-slips, dried and stained. (4) The manner of origin of the suspensory thread of the pupa. (5) Repetition of previous lighting up experiments with larvae. (6) The study of possible predators, especially in caves. (7) The main reason for mortality in large communities of larvae.

(8) The physical nature and attractiveness of the glowworm light. (9) The reaction of glowworms to lights of different colours. (10) A study of nerve endings with recent techniques. Summary (1) Larvae of the Tasmanian species up to 23 mm in length have been examined. The hooked areas or locomotory pads on one were very small. These areas on the N.S.W. species resembled more those found in the North Island of New Zealand. The taxonomic value of these hooks is still doubtful. The hooks always point forwards. (2) The N.S.W. larva is more highly coloured and pigmented than the N.Z. form, from comparable localities. (3) The three types of larvae examined all had the last few abdommal segments flattened on the dorsal suface. (4) The antenna of the male Tasmanian species is relatively much longer than that of a comparable N.Z. male. There are sixteen joints counting the two basal ones in both the Tasmanian and the N.Z. males. (5) The Tasmanian male is more delicately built than the N.Z. male, and is thereby unmistakable. (6) Nerves passing from the last abdominal ganglion in both larva and imago, ramify over the light organ cells, and appear to have special pyramidal nerve endings on the cell membrane. (7) There is no evidence that tracheoles pass from the reflector into or onto the light organ cells of the larva. During pupation the tracheolar reflector basket becomes hollowed out, and tracheoles originating from the walls ramify over the light cells. The hollow reflector retains its connection with the main tracheal trunk on each side. (8) The imago of the N.S.W. species is unknown. There is a certain amount of evidence that the larvae of the N.S.W. form tend to be smaller than the N.Z. larvae, and never have such large and dark combed areas. Professor J. Bronte Gatenby, B.A., D. Phil. (Oxon.), D.Sc. (Lond.), Trinity College, Dublin. Literature Cited Beams, H. W. and Anderson, E., 1955. Light and Electron Microscope Studies on the Light Organ of the Firefly (Photinus pyralis). Biol. Bull. 109, 375. Cook, O. F., 1913. Web Spinning Fly Larvae in Guatemalan Caves. Jour. Wash. Acad. Sci. III. 190. Ferguson, E. W., 1925. Description of a New Species of Mycetophilidae with a Luminous Larva. Proc. Linn. Soc. N.S.W. 50, 4. Ganguly, G., 1960. The Histology of the Larva of Bolitophila luminosa. Journ. Roy. Micros. Soc. London. 79 (2). Gatenby, J. Bronte, 1959. Notes on the New Zealand Glowworm, Bolitophila luminosa. Trans. Roy. Soc. N.Z..

Gatenby, J. Bronte, and Cotton, S., 1960. Snare Building and Pupation in Bolitophila luminosa. Trans. Roy. Soc. N.Z.. Hudson, G. V., 1950. “Fragments of N.Z. Entomology.” Ferguson & Osborn, Wellington, N.Z. Lee, D. J., 1959. (Personal Communication.) Madwar, S., 1937. Biology and Morphology of Mycetophilidae. Phil Trans. Roy. Soc. (B) 227, 1. McKeown, K. C., 1935. “Insect Wonders of Australia.” Angus & Robertson, Ltd. Nicholson, A. J., 1959. (Personal Communication.) Osten-Sagken, C. R., 1886. A Luminous Insect Larva in New Zealand. Ent. Month Mag. Wigglesworth, V., 1959. The Histology of the Nervous System of Rhodnius prolixus. Quart Jour. Micr. Sci., Vol 100, Part 2, 285. Lettering A=antenna. AN=nerve fibre. BP=median functional part of malpighian tubes, brownish pink (but clear pink in the N.Z. form). CO and CO6=combed bands (locomotory pads). COA=coagulum. D=mucus gland diverticulum, dark with café au lart centres to each cell in some examples. E=eye. FB=fat body. GA (G)=ganglion. GF=green fat bodies (grey in the N.Z. form). HA=imaginal disc of hind abdomen. HS=stetae on prothorax. INT, IN=intestine (pale to hyaline) L=inaginal disc of leg LE=leg of pupa LO=light organ lens ME=mesenteron, cells yellow grey, food brown. MG=male genital canal. MP=coiled exit part of malpighian tubes, yellow. N=nucleus. NE=supposed nerve ending. O=oesophagus (pale yellow) OV=oesophageal valve. OR=communicating fibre between sensillae cells. P=anal papillae. PM=peritrophic membrane R=reflector or reflector cavity. S=curved seta of chordotonal organ. SI=silk glands (hyaline). T=tracheole. TS=testis. TT=main tracheal trunk. TX=supposed tracheole. V=oesophageal valve (pale yellow). W=wing or halture imaginal disc. X=one of the chordotonal sense organs. Zones of Colour BZ=jet black with bluish edges. GZ=chocolate to coffee colour. GRZ=green and pink, green predominating. Note Bands (CO) lie mainly ventrally.