Visual metamorphoses in insects and malacostracans: Transitions between an aquatic and terrestrial life

Arthropods operate in an outrageous diversity of environments. From the deep sea to dense tropical forests, to wide open arctic tundra, they have colonized almost every possible habitat. Within these environments, the presence of light is nearly ubiquitous, varying in intensity, wavelength, and polarization. Light provides critical information about the environment, such as time of day or where food sources may be located. Animals take advantage of this prevalent and informative cue to make behavioral choices. However, the types of choices animals face depend greatly on their environments and needs at any given time. In particular, animals that undergo metamorphosis, with arthropods being the prime example, experience dramatic changes in both behavior and ecology, which in turn may require altering the structure and function of sensory systems such as vision. Amphibiotic organisms maintain aquatic lifestyles as juveniles before transitioning to terrestrial lifestyles as adults. However, light behaves differently in water than in air, resulting in distinct aquatic and terrestrial optical environments. Visual changes in response to these optical differences can occur on multiple levels, from corneal structure down to neural organization. In this review, we summarize examples of alterations in the visual systems of amphibiotic larval and adult insects and malacostracan crustaceans, specifically those attributed to environmental differences between metamorphic phases.

Evolutionary constraints on flicker fusion frequency in Lepidoptera

Flying insects occupy both diurnal and nocturnal niches, and their visual systems encounter distinct challenges in both conditions. Visual adaptations, such as superposition eyes of moths, enhance sensitivity to low light levels but trade off with spatial and temporal resolution. Conversely, apposition eyes of butterflies enable high spatial resolution but are poorly sensitive in dim light. Although diel activity patterns of insects influence visual processing, their role in evolution of visual systems is relatively unexplored. Lepidopteran insects present an excellent system to study how diel activity patterns and phylogenetic position influence the visual transduction system. We addressed this question by comparing electroretinography measurements of temporal response profiles of diverse Lepidoptera to light stimuli that were flickering at different frequencies. Our data show that the eyes of diurnal butterflies are sensitive to visual stimuli of higher temporal frequencies than nocturnal moths. Hesperiid skippers, which are typically diurnal or crepuscular, exhibit intermediate phenotypes with peak sensitivity across broader frequency range. Across all groups, species within families exhibited similar phenotypes irrespective of diel activity. Thus, Lepidopteran photoreceptors may have diversified under phylogenetic constraints, and shifts in their sensitivity to higher temporal frequencies occurred concomitantly with the evolution of diurnal lifestyles.

Ultrastructural comparison of the compound eyes of the Asian corn borer Ostrinia furnacalis (Lepidoptera: Crambidae) under light/dark adaptation

The Asian corn borer Ostrinia furnacalis is one of the most destructive pests of maize throughout eastern Asia and the South Pacific. In the present study the fine structure of the compound eyes of adult O. furnacalis was investigated under light/dark adaptation using light and electron microscopy. The compound eyes of male and female O. furnacalis are superposition eyes with electron-lucent clear zones. The sexual differences of the compound eyes of O. furnacalis are mainly reflected in eye size rather than ommatidial ultrastructure. Each ommatidium of both sexes contains 12 retinula cells, one of which is the basal retinula cell. All the retinula cells form a centrally-fused, two-tiered rhabdom, whose distal layer passes through the clear zone and distally connects with the crystalline cone. The ultrastructural changes under light/dark conditions mainly involve the rhabdom occupation ratio to retinula cell volume in the proximal layer of the rhabdom as well as the dimensions of the subcorneal zone and the crystalline tract. Pigment movements occur within the retinula cells and primary pigment cells, but are undetectable within the secondary pigment cells. Regardless of light or dark adaptation, in other words, the pigments never migrate into the clear zone.

Evolution of Phototransduction Genes in Lepidoptera

Vision is underpinned by phototransduction, a signaling cascade that converts light energy into an electrical signal. Among insects, phototransduction is best understood in Drosophila melanogaster. Comparison of D. melanogaster against three insect species found several phototransduction gene gains and losses, however, lepidopterans were not examined. Diurnal butterflies and nocturnal moths occupy different light environments and have distinct eye morphologies, which might impact the expression of their phototransduction genes. Here we investigated: 1) how phototransduction genes vary in gene gain or loss between D. melanogaster and Lepidoptera, and 2) variations in phototransduction genes between moths and butterflies. To test our prediction of phototransduction differences due to distinct visual ecologies, we used insect reference genomes, phylogenetics, and moth and butterfly head RNA-Seq and transcriptome data. As expected, most phototransduction genes were conserved between D. melanogaster and Lepidoptera, with some exceptions. Notably, we found two lepidopteran opsins lacking a D. melanogaster ortholog. Using antibodies we found that one of these opsins, a candidate retinochrome, which we refer to as unclassified opsin (UnRh), is expressed in the crystalline cone cells and the pigment cells of the butterfly, Heliconius melpomene. Our results also show that butterflies express similar amounts of trp and trpl channel mRNAs, whereas moths express ∼50× less trp, a potential adaptation to darkness. Our findings suggest that while many single-copy D. melanogaster phototransduction genes are conserved in lepidopterans, phototransduction gene expression differences exist between moths and butterflies that may be linked to their visual light environment.

The giant butterfly-moth Paysandisia archon has spectrally rich apposition eyes with unique light-dependent photoreceptor dynamics

The palm borer moth Paysandisia archon (Burmeister, 1880) (fam. Castniidae) is a large, diurnally active palm pest. Its compound eyes consist of ~ 20,000 ommatidia and have apposition optics with interommatidial angles below 1°. The ommatidia contain nine photoreceptor cells and appear structurally similar to those in nymphalid butterflies. Two morphological ommatidial types were identified. Using the butterfly numbering scheme, in type I ommatidia, the distal rhabdom consists exclusively of the rhabdomeres of photoreceptors R1–2; the medial rhabdom has contributions from R1–8. The rhabdom in type II ommatidia is distally split into two sub-rhabdoms, with contributions from photoreceptors R2, R3, R5, R6 and R1, R4, R7, R8, respectively; medially, only R3–8 and not R1–2 contribute to the fused rhabdom. In both types, the pigmented bilobed photoreceptors R9 contribute to the rhabdom basally. Their nuclei reside in one of the lobes. Upon light adaptation, in both ommatidial types, the rhabdoms secede from the crystalline cones and pigment granules invade the gap. Intracellular recordings identified four photoreceptor classes with peak sensitivities in the ultraviolet, blue, green and orange wavelength regions (at 360, 465, 550, 580 nm, respectively). We discuss the eye morphology and optics, the photoreceptor spectral sensitivities, and the adaptation to daytime activity from a phylogenetic perspective.

Ultrastructure and Morphology of Compound Eyes of the Scorpionfly Panorpa dubia (Insecta: Mecoptera: Panorpidae)

Mecoptera are unique in holometabolous insects in that their larvae have compound eyes. In the present study the cellular organisation and morphology of the compound eyes of adult individuals of the scorpionfly Panorpa dubia in Mecoptera were investigated by light, scanning electron, and transmission electron microscopy. The results showed that the compound eyes of adult P. dubia are of the apposition type, each eye comprising more than 1200 ommatidia. The ommatidium consists of a cornea, a crystalline cone made up of four cone cells, eight photoreceptors, two primary pigment cells, and 18 secondary pigment cells. The adult ommatidium has a fused rhabdom with eight photoreceptors. Seven photoreceptors extend from the proximal end of the crystalline cone to the basal matrix, whereas the eighth photoreceptor is shorter, extending from the middle level of the photoreceptor cluster to the basal matrix. The fused rhabdom is composed of the rhabdomeres of different photoreceptors at different levels. The adult ommatidia have the same cellular components as the larval ommatidia, but the tiering scheme is different.

Compound eyes of insects and crustaceans: Some examples that show there is still a lot of work left to be done

Similarities and differences between the 2 main kinds of compound eye (apposition and superposition) are briefly explained before several promising topics for research on compound eyes are being introduced. Research on the embryology and molecular control of the development of the insect clear-zone eye with superposition optics is one of the suggestions, because almost all of the developmental work on insect eyes in the past has focused on eyes with apposition optics. Age- and habitat-related ultrastructural studies of the retinal organization are another suggestion and the deer cad Lipoptena cervi, which has an aerial phase during which it is winged followed by a several months long parasitic phase during which it is wingless, is mentioned as a candidate species. Sexual dimorphism expressing itself in many species as a difference in eye structure and function provides another promising field for compound eye researchers and so is a focus on compound eye miniaturization in very small insects, especially those that are aquatic and belong to species, in which clear-zone eyes are diagnostic or are tiny insects that are not aquatic, but belong to taxa like the Diptera for instance, in which open rather than closed rhabdoms are the rule. Structures like interommatidial hairs and glands as well as corneal microridges are yet another field that could yield interesting results and in the past has received insufficient consideration. Finally, the dearth of information on distance vision and depth perception is mentioned and a plea is made to examine the photic environment inside the foam shelters of spittle bugs, chrysales of pupae and other structures shielding insects and crustaceans.

Adaptations for Nocturnal Vision in Insect Apposition Eyes

Due to our own preference for bright light, we tend to forget that many insects are active in very dim light. Nocturnal insects possess in general superposition compound eyes. This eye design is truly optimized for dim light as photons can be gathered through large apertures comprised of hundreds of lenses. In apposition eyes, on the other hand, the aperture consists of a single lens resulting in a poor photon catch and unreliable vision in dim light. Apposition eyes are therefore typically found in day-active insects. Some nocturnal insects have nevertheless managed the transition to a strictly nocturnal lifestyle while retaining their highly unsuitable apposition eye design. Large lenses and wide photoreceptors enhance the sensitivity of nocturnal apposition eyes. However, as the gain of these optical adaptations is limited and not sufficient for vision in dim light, additional neural adaptations in the form of spatial and temporal summation are necessary.

Dimensional limits for arthropod eyes with superposition optics

An essential feature of the superposition type of compound eye is the presence of a wide zone, which is transparent and devoid of pigment and interposed between the distal array of dioptric elements and the proximally placed photoreceptive layer. Parallel rays, collected by many lenses, must (through reflection or refraction) cross this transparent clear-zone in such a way that they become focused on one receptor. Superposition depends mostly on diameter and curvature of the cornea, size and shape of the crystalline cone, lens cylinder properties of cornea and cone, dimensions of the receptor cells, and width of the clear-zone. We examined the role of the latter by geometrical, geometric-optical, and anatomical measurements and concluded that a minimal size exists, below which effective superposition can no longer occur. For an eye of a given size, it is not possible to increase the width of the clear-zone cz=dcz/R1 and decrease R2 (i.e., the radius of curvature of the distal retinal surface) and/or c=dc/R1 without reaching a limit. In the equations 'cz' is the width of the clear-zone dcz relative to the radius R1 of the eye and c is the length of the cornea-cone unit relative to R1. Our results provide one explanation as to why apposition eyes exist in very small scarabaeid beetles, when generally the taxon Scarabaeoidea is characterized by the presence of superposition eyes. The results may also provide the answer for the puzzle why juveniles or the young of species, in which the adults possess superposition (=clear-zone) eyes, frequently bear eyes that do not contain a clear zone, but resemble apposition eyes. The eyes of the young and immature specimens may simply be too small to permit superposition to occur.

Synaptic organization in the lamina of the superposition eye of a skipper butterfly, Parnara guttata

The first optic neuropil of the compound eye, the lamina, of the skipper butterfly Parnara guttata, was examined by light microscopy after Golgi-impregnation and by electron microscopy (EM) to clarify the cellular and synaptic organization. In the lamina, five different types of lamina neurons (L neurons) were characterized by using Golgi-impregnation. By EM, each cartridge was found to contain all nine receptor axons from an ommatidium, five L neurons, and a few putative centrifugal elements. Axons from photoreceptors (retinula cells) R2, R3, R4, R6, R7, and R8 terminate as short visual fibers (svfs) in the lamina cartridge. Those from R1, R5, and R9 penetrate the lamina and terminate in the medulla as long visual fibers (lvfs). In the cartridges, the synaptic contacts were formed from svfs onto L neurons, from the lvfs of R1 and/or R5 to the lvf of R9 and L neurons, and from the lvf of R9 to L neurons. The putative centrifugal fibers also make synapses to svfs and L neurons. At the most distal level of the cartridge, one of the centrifugal fibers containing dense-core vesicles makes presynaptic contacts to the putative long collaterals of the L neuron. A novel characteristic feature of this lamina is that svfs of R3 and R7 and the lvfs of R1 or R5 have long collaterals extending into neighboring cartridges. Presynaptic contacts were confirmed in such long collaterals from the svf. These results imply that receptor axons provide direct intercartridge connections as well as providing indirect connections to neighboring cartridges by way of their input upon L neurons.

Regionally different optical systems in the compound eye of the water-flea Polyphemus (Cladocera, Crustacea)

Polyphemus pediculus (L.) is a small (1 mm long) predatory crustacean that lives in bodies of standing freshwater. It has a single fused compound eye, which occupies most of its head. The eye comprises 130 ommatidia with five distinct types of crystalline cones. Four of these cone types were found to focus light by means of gradient index optics (lens cylinders). The edge ommatidia differ by having the focus displaced below the distal rhabdom tip. This was found to be correlated with their special type of rhabdom, which is characterized by its short, broad shape and the absence of a palisade. The central-type crystalline cone, contributing to a zone of acute vision, is functionally different from the other four cone types. The focusing on the rhabdom tip is in this case achieved by a prism, inside the cone, corrected for optical aberration with a complex refractive index gradient. The prism is interpreted as a way of compressing a long focal length into a short optical system, i. e. to enable high resolution in spite of the small size of the eye. Extreme regional differences in interommatidial angles were found to be the main reason for the different optical design between central and peripheral ommatidia.

The dorsal eye of the mayfly Atalophlebia (Ephemeroptera)

The dorsal eye of Atalophlebia has two unusual features, the sensitivity only to ultraviolet (u. v.) light, and the candelabra-shaped rhabdom. In addition, the crystalline cone is surrounded to its tip by a yellow pigment, and the tip tapers gradually as a dense fibre. These details, particularly the pigment distribution, indicate that a superposition image cannot be formed by u. v. light. Also, there is no refracting or reflecting structure that could form a sharp superposition image. Instead, it is suggested that u. v. rays are sharply focused on the cone tip and conducted by the retinula cell columns acting as light guides across the clear zone. Light of longer wavelength, on the other hand, is partially focused through the yellow pigment, and, although it is not seen by the insect, it is available to photoregenerate the visual pigment. This method of boosting sensitivity is appropriate for a pure u. v. eye and does not require a sharp focus of the regenerative rays, although the clear zone is an essential part of the mechanism. The rhabdom has an extraordinary shape like a flat 5-armed candelabra in cross section, with five posteriorly directed arms which are formed by six retinula cells. There is also a 7th retinula cell without a rhabdomere. This cell penetrates laterally the rhabdom of the other six, and also forms a sheath around half of its own ommatidium and half of the the adjacent ommatidium. The exceptional relations between this cell, and the other six, together with the orientated candelabra pattern of the rhabdom, and the large size of the 7th retinula axon, is interpreted as a way of enhancing the current flow down the 7th axon which runs direct to the medulla, bypassing the lamina.

Graded-index optics are matched to optical geometry in the superposition eyes of scarab beetles

Detailed measurements were made of the gradients of refractive index (g.r.i.) and relevant optical properties of the lens components in the ventral superposition eyes of three crepuscular species of the dung-beetle genus Onitis (Scarabaeinae). Each ommatidial lens has two components, a corneal facet and a crystalline cone; in both of these, the gradients provide a significant proportion of the refractive power. The spatial relationship between the lenses and the retina (optical geometry) was also determined. A computer ray-trace model based on these data was used to analyse the optical properties of the lenses and of the eye as a whole. Ray traces were done in two and three dimensions. The ommatidial lenses in all three species are afocal g.r.i. telescopes of low angular magnification. Parallel incident rays emerge approximately parallel for all angles of incidence up to the maximum. The superposition image of a distant point source is a small patch of light about the size of a rhabdom. There are obvious differences in the lens properties of the three species, most significantly in the shape of the refractive-index gradients in the crystalline cone, in the extent of the g.r.i. region in the two lens components and in the front-surface curvature of the corneal facet lens. These give rise to different angular magnifications M of the ommatidial lenses, the values for the three species being 1.7, 1.3, 1.0. This variation in M is matched by a variation in optical geometry, most evident in the different clear-zone widths. As a result, the level of the best superposition image lies close to the retina in the model eyes of all three species. The angular magnification also sets the maximum aperture or pupil of the eye and hence the brightness of the image on the retina. The smaller M , the larger the aperture and the brighter the image. By adopting a suitable value for M and the appropriate eye geometry, an eye can set image brightness and hence sensitivity within a certain range. Differences in the eye design are related to when the beetles fly at dusk. Flight experiments comparing two of the species show that the species with the higher value for M and corresponding lower sensitivity, initiates and terminates its flight earlier in the dusk than the other species with 2.8 times the sensitivity.

The eye of the soldier beetle Chauliognathus pulchellus (Cantharidae)

The soldier beetle eye is unusual in having large optically isotropic corneal cones which project inwards from a thick isotropic cornea. Refraction is mainly at the corneal surface. Calculation shows that the first focal plane is near the tip of the cone, from which the optical pathway continues as a crystalline tract. At the distal end of the crystalline tract, 3 micrometer in diameter, the four cone cells enclose the proximal tip of the corneal cone; at the proximal end they enclose the distal tip of a long fused rhabdom rod. The eye is remarkable in that there are two classes of retinula cells; four cells contribute to the long thin axial rhabdom, 2 micrometer in diameter and 120 micrometer long, and the other four cells form two rounded rhabdoms, 10 x 4 micrometer in cross-section and 20 micrometer deep, which lie to one side of the optical axis. The physiological properties of individual retinula cells were measured by intracellular recording. The retinula cells are of three spectral types with peaks near 360, 450 and 520--530 nm. Except by the criterion of spectral sensitivity, the retinula cells sampled could not be sorted into more than one class. The measured value of the acceptance angle, near 3 degrees in the dark-adapted state, is consistent with the hypothesis that all sampled cells were of the anatomical type that participate in the central rhabdom rod. A calculation of the theoretical field size of individual retinula cells from measurments of refractive index and lens dimensions predicts that cells which participate in the central rhabdom will have acceptance angles near 3 degrees. The conclusion, therefore, is that only one anatomical type of cell has so far been sampled.

A diurnal moth superposition eye with high resolution Phalaenoides tristifica (Agaristidae)

This common Australian moth flies slowly and only in bright sunlight, but the compound eye has a sharply focused superposition image, to which 120-150 facets contribute. A distant point source is focused to form a blur circle on the retina. The diameter of this blur circle at 50% intensity, calculated from the width of the angular distribution of eyeshine, is 1.5-2.0° subtended at the centre of the eye, matching the interommatidial angle. The compromise between acuity and sensitivity, which is dependent on rhabdom spacing in relation to the blur circle diameter, has been struck with approximately 50% of the converging light from a distant point source falling on a single rhabdom. The F value of the focusing arrangement, defined as (focal length/aperture) is approximately 0.9. The structure of the eye, which undergoes negligible change on adaptation to light, is described by electron microscopy. The optical system is examined. The cornea and crystalline cone taken together constitute an afocal combination of lenses. Although the retinula cell columns act as light guides, a negligible contribution passes down them because rays are not focused upon their distal ends. The rhabdom columns are isolated from each other by complete sleeves of specialized tracheae backed by screening pigment. The e. r. g. is dominated by a single visual pigment resembling a rhodopsin with peak near 530 nm but adaptation studies suggest a small amount of a second pigment peaking near 360 nm. There are 14-16 retinula cells per ommatidium. An explanation for this great number, together with the use of a superposition image in daylight, cannot be offered until the visual behaviour is better understood. Apparently the eye is adapted for extreme sensitivity compatible with good resolution, and possibly this moth can see small contrast differences in small objects. For reasons that are discussed, a similar sharp focus is not to be expected in the eyes of moths that fly in dim light.

The organization of perpendicular fibre pathways in the insect optic lobe

High resolution serial photomicrography has been used to plot the axonal projection patterns between retina, lamina and medulla in the optic lobes of various insects with differing ommatidial receptor arrangements. Observations are reported on the cabbage white and skipper butterflies, the bee, locust, fly, backswimmer and waterbug. The patterns of these fibre pathways have previously eluded non-rigorous analyses primarily because of their physical dimensions but are revealed in this study to have striking precision and uniformity between species when examined at the level of individually identifiable cells. Axon bundles of the tracts between retina and lamina or lamina and medulla project between a single ommatidium and its corresponding lamina cartridge or between corresponding lamina and medulla cartridges. Lateral interweaving of axons between adjacent bundles is absent. The bundles preserve the retinotopic order within their total array, so transferring the pattern of retinulae directly upon the lamina and thence after horizontal inversion in the chiasma upon the medulla. Within the lamina neuropile on the other hand the trajectories of the individual terminals from a bundle have patterns which are species-specific, sometimes involving lateral divergences. In species with open-rhabdomere ommatidia the terminals distribute to a group of lamina cartidges with a pattern which resembles the receptor pattern in the overlying ommatidium. In species with fused-rhabdome ommatidia the terminals of a single retinula behave less interestingly and all enter the same cartridge, within which, again, each occupies a position related to its cell body position within the retinula. Long visual fibres in both eye types penetrate the lamina and terminate in the particular medulla cartridge that connects with the lamina cartridge underlying their ommatidium. The perpendicular fibre pathways therefore project the visual field exactly upon the medulla in all species while the lack of interweaving between adjacent fibre bundles precludes their involvement in lateral interactions between pathways with differing visual axes. Uniformity of these projection patterns between cell layers and species differences in retinular terminal locations in the lamina can be correlated with different modes of axon growth between and within neuropile layers during optic lobe neurogenesis. Further discussion surrounds the question of which particular receptors give rise to which type of axon, for which no clear generalization has yet emerged.

The ommatidium of the lacewing Chrysopa (Neuroptera)

The eye is a clear zone eye with extensive movement of retinula cells on adaptation to light. The ommatidium has three types of rhabdomere, at different levels, so that the eye necessarily abstracts at least three kinds of information simultaneously from the incoming rays. In the lightadapted state light can enter each ommatidium only via a crystalline tract that is surrounded by dense pigment grains. A small distal rhabdomere (cell 7) always lies at the end of this tract. In the dark-adapted eye the retinula cell nuclei and distal rhabdomere move to the cone tip and the crystalline tract is drawn into the cone. There is then a region of the retinula cell column, between cone tip and proximal rhabdoms, across which there is no structure that could act as a light guide. A key question, therefore, is how the light is focused across this clear zone in the darkadapted state. As shown by the wide angular distribution of eyeshine when a parallel beam is incident on the dark-adapted eye, rays are poorly focused upon the columns of the large rhabdoms. The wide visual fields of receptors 1-6 in the dark-adapted eye, inferred from the observation of eyeshine, are seen as a way of narrowing the bandwidth of spatial frequencies, so that only the largest objects in the visual field contribute to motion-detection. This would improve the signal-to-noise ratio, not in the receptors themselves, but in the neural mechanism, by simplifying the incoming signal.

Two models of the partially focused clear zone compound eye

1. The theory of the unfocused clear zone eye is extended to cover cases where rays are partially focused upon the receptor layer. 2. Light is admitted through facets according to a Gaussian distribution of angle of incidence defined with respect to the axis of the facet. 3. The same light crosses the clear zone in an average direction related to its direction of origin outside the eye, so that it tends to be concentrated around that receptor on the radial axis pointing towards the source of light. This effect, defined as focusing in contrast to the unfocused eye, allows a simultaneous improvement in sensitivity and acuity. 4. An eye can be focused partially because rays diverge from each cone tip or because, on average, they converge above or below the receptor layer on the radial axis pointing towards their source. The two situations are analysed quantitatively. 5. In a partially focused eye neither the measured angular sensitivity nor the absolute sensitivity allow a prediction of the ray paths because many different distributions from the cone tips produce a given final result. Therefore the optics of partially focused clear zone eyes must be analysed by direct measurement of light distribution from the cone tips.

The dioptric system of the eye of Cybister (Dytiscidae: Coleoptera)

1. The compound eye of Cybister is anatomically similar to that of Dytiscus and Hydrophilus . 2. The cornea and crystalline cone in the compound eye of Cybister (Dytiscidae) are composed of layers of unequal refractive index. With the exception of the outer 10 µ m of the cornea (where they are horizontal) the layers are arranged concentrically around a region of highest refractive index on the axis. 3. The refractive index of the cornea decreases from the central layer (1.724) to the periphery (1.561). The corresponding values for the crystalline cone are 1.435 and 1.366. The refractive index of the area between cornea and cone is 1.343; that of the clear zone is 1.341, and that of the proximal rhabdom is 1.361. 4. Parallel rays entering a facet converge to a focal region which extends from the proximal part of the cornea to the distal part of the cone. Rays cross the clear zone in a direction which depends on the angle to the axis and position on the facet. Up to an angle of 32° to the axis the rays are mainly bent back into the quadrant of origin so that rays entering many facets converge to a second focal region beyond the clear zone. These findings are consistent with the report of a first inverted although fuzzy image in the cone and a second image (Exner 1891). 5. Ray diagrams were constructed for three different positions of the distal pigment. If the cone tip is completely exposed, the receptor acceptance angle is 46°. With the pigment in the typical dark-adapted position the field of view is 38° wide. A light-adapted ommatidium would have an admission function about 18° wide according to ray tracing, but this could be reduced by the properties of the crystalline tract down which the light must pass. 6. It is concluded from the ray tracing that acuity is poor, but summation across the clear zone could confer a high sensitivity for the dark-adapted eye.