ALBERT

Notes: Summary 1. Azimuth orientation in the halfbeak fishDermogenys was studied in the laboratory to find out whether its spontaneous heading directions in a vertical beam of linearly polarized light involve perception of thee-vectorper se or merely of concomitant light intensity patterns. Responses were tested with polarized and unpolarized light as well as with either a uniform white screen horizontally surrounding the experimental vessel or with one divided into black and white alternating quadrants. 2. Measured as counts within 10 °, 45 ° or 90 ° sectors through 180 ° the fish's azimuth orientation was random with unpolarized light and the white surround (Fig. 2). 3. In contrast significant preferential orientation was shown in the presence of linearly polarized light and the white surround (Fig. 3). The 10 ° sector centered on the plane of vibration had the most counts (Fig. 3A). Combining the data into four sectors each 45 ° in extent makes clear a significant predominance of orientation parallel to thee-vector (Fig. 3B) as do the total counts for parallel and perpendicular quadrants (Fig. 3D). 4. With the black and white quadrants combined with unpolarized light preferential orientation was clearly shown toward the light sectors (Fig. 4A, B). Since maximum differential scattering from a linearly polarized light beam is perpendicular to the plane of vibration the positive sign of this phototactic response ofDermogenys is evidence that the observed orientation parallel to thee-vector cannot also be a similar response to intensity pattern. Hence the plane of vibration must be perceived through a distinct information channel and polarotaxis is different from phototaxis. 5. Tests with linear polarized light combined with the black and white surround proved that phototaxis predominated over polarotaxis under our experimental conditions and that the interaction between the two types of behavior in this case was not a simple additive one (Fig. 4C-F).

Notes: Summary 1. Because behavioral evidence indicates that fishes can perceivee-vector direction in plane polarized light, intracellular recordings were made on bipolar cells, ganglion cells, amacrine cells and horizontal cells in the goldfish retina. In flattened retinal fragments stimulated with polarized flashes no evidence for significante-vector sensitivities was found. 2. Dichroism could not be demonstrated in the cornea, lens or other dioptric elements of this fish eye. 3. Finally extracellular spike recordings of single units in the goldfish optic tectum were made to determine whethere-vector discrimination could be measured in the output of the intact eye in the living fish. 500 msec test flashes were presented to the retina with narrow band spectral red, green and blue light (quantum equalized) as well as with white light over a 4–5 log unit range of intensities. 4. Practically all tectal cells of the 47 successfully recorded showed some sensitivity toe-vector direction. The response function was roughly sinusoidal with maxima and minima 90° apart. Maximal responses to polarization plane orientation were found in all sixe-vector directions tested. Consequently an analyzer without discrete channels for particular polarization planes must be present quite different from that present in many rhabdom bearing eyes. 5. E-vector direction influenced various parameters of the tectal responses (on, off, sustained, etc.) in some cases in the same direction and in other cases in the opposite direction. Both excitatory and inhibitory components, as well as color coded ones, were affected. 6. Intensity response curves indicate polarized light sensitivity ratios ranging from 1.4 to 31.7, with a mean of 8.2 for 13 cases.

Notes: Summary 1. Spike responses of single optic tectal units to plane polarized light have been recorded with extracellular electrodes in the goldfish. Responses to a series of eight 500 msec flashes were summed and the quantitative effects ofe-vector direction studied over 4–5 log units of intensity with white light and narrow spectral bands of equal quantal content at 460, 540 and 620 nm. 2. As previously reported all or nearly all tectal cells tested (115) were sensitive to the direction of the stimulus' polarization plane. Control experiments with a thin high order birefringent retarder next to the cornea and acting as a pseudo depolarizer provided further proof that thee-vector discrimination found in the optic tectum (as well as the previously reported oriented behavior to polarized light in fish) depends on an intraocular analyzer. 3. Systematic study of tectal cells has been made at various depths and in all directly accessible areas. No localized or differential effects of recording depth one-vector discrimination have been found. However, both directions of maximum response and degree of polarized light sensitivity (PLS) show distinctive patterns over the tectal area. 4. On the basis of established tectal projection maps these data show that preferred retinale-vector directions are tangentially arranged around the eye axis when the stimulus is axial. Distribution of this angular sensitivity seems continuous without discrete channels favoring particular polarization planes. 5. Sensitivity (determined from intensity response curves) is minimal in the center of the retina for an axial stimulus and increases peripherally out to 50–60° or more off axis. Sensitivity ratios of 4–6 are common and much larger ones have been occasionally recorded. 6. Shifting the direction of stimulus from axial to 45° and 60° upward from the axis proved that retinal patterns of preferrede-vector directions and sensitivities are symmetrical relative to the beam axis rather than the anatomical eye axis. Therefore individual tectal cells and before them their retinal receptor elements show different directions ofe-vector preference and different PLS ratios depending on the stimulus configuration. 7. Intensity response curves determined with the red, green and blue narrow spectral bands were essentially superimposable. Therefore, we have no evidence that there is any special interaction of λ and PLS. 8. The effects of stimulus intensity and retinal adaptation on the IR curves do not indicate any particular selective effects. Substantial polarized light sensitivity was present in both the light adapted and dark adapted state. PLS ratios in various cases were (1) about the same at different intensity levels or (2) greater at low than high intensities or (3) greater at high than low intensities. 9. We conclude that linearly polarized light evokes a large entoptic image in the goldfish eye. Two opposite light sectors perpendicular to thee-vector alternate with two dark sectors parallel to thee-vector. Both Haidinger's brushes and Boehm's brushes in human vision show some similarities but also some important differences. The goldfish PL image is achromatic, weak or absent axially and strong peripherally. Also the contrast between the sectors does not depend on movement of the stimulus on the retina to prevent fading as it does in both of the other phenomena. 10. The most likely mechanism for the observed PLS in the goldfish eye is either differential scattering intraocularly or oblique entry of light into the receptor outer segments. Yet present data prevent a firm choice between them.

Notes: [Auszug] Nevertheless, there are actually three previous reports of polarized light responses by teleost fishes3-5, but none of these was adequately followed up to provide needed additional data or reasonable explanations for certain ambiguous or anomalous experimental results. We now report consistent ...

Notes: Summary 1. Receptor potentials have been recorded intracellularly from single retinular cells in the anterior and dorsal quadrants of the compound eye of the crayfish Procambarus (Fig. 1) stimulated with equal quantum flashes of linearly polarized monochromatic light. Comparisons between two orthogonal stimulus e-vectors respectively parallel and perpendicular to the microvilli of each receptor cell's rhabdomere were made sequentially in about one minute's running time at 20 nm intervals between 400 and 740 nm (Fig. 2). 2. Of the 91 cells studied 17 responded maximally in the violet (av. λ max= 440 nm) whereas the other 74 cells were most responsive in the yellow-orange (av.λ max=594 nm) (Table 1). For the latter group the λ max of individual cells ranged widely from 538 to 634 nm (Fig. 4). Violet sensitive cells were found only in the anterior quadrant of the eye. 3. For 29 cells spectral sensitivity curves were plotted from the spectral efficiency curves using response-energy functions determined at λ max or spectral efficiency curves taken at two or more stimulus energy levels (Figs. 5B, 6B, 7). When the sensitivity curves are normalized the vertical and horizontal e-vector responses are closely similar indicating that dichroism of the visual pigment is undoubtedly responsible for the observed differential sensitivity (Figs. 5C, 6C). 4. For 51 yellow-orange cells where e-vector comparisons can be made more than half (57%) were more responsive to vertical e-vector (Table 2) corresponding very closely with the estimated percentage of retinular cells with microvilli parallel to the body's dorso-ventral axis (57.2%). In contrast five of the seven violet cells available for this comparison gave stronger responses to horizontal e-vector suggesting they may predominantly be the one asymmetrical cell in each ommatidium. Nevertheless both color discriminating types were found to be present in both e-vector channels. 5. For the 29 cells for which spectral sensitivity curves can be plotted the average sensitivity ratio for the two polarization planes is 3.1 with a range from 1.2 to 11.9 at λ max. Since dichroic absorption ratios directly measured in crayfish have previously been shown to be about 2, the origin of greater spectral sensitivity ratios in individual retinular cells most likely must depend on other functions than photon absorption by a single rhabdomere.

Notes: Summary The eighth retinular cell (R 8) of Grapsus lacks cytoplasmic pigment granules and basically resembles those previously known in the ghost crab Ocypode and the mysid Praunus. Distally located, R 8 comprises four lobes inserted between the outer ends of the seven regular retinular cells (R 1–R 7). A thin cytoplasmic bridge connects these lobes. One lobe adjacent to R 1 contains the nucleus of R 8 and gives rise proximally to the cell's axon. The short distal eighth rhabdomere consists of microvilli (mvl) protruding axially from all four lobes. Similar R 8's were found also in two other crab families and in two other genera of mysids. In Grapsus the eighth rhabdomere is extraordinary in possessing mvl oriented in two orthogonal directions parallel to the mvl of R 1–R 7. The distal 20% of the rhabdom consists of mvl originating exclusively from R 8. These appear in somewhat irregular bands and are alternately oriented parallel to the animal's vertical or horizontal axis. More proximally the retinula contains eleven sectors but the rhabdom still comprises bands of alternating mvl with those from R 8 joined respectively by the rhabdomeres of R 1, 4, and 5 (horizontal) and R 2, 3, 6 and 7 (vertical). The rest of the rhabdom shows typical decapod organization with seven interdigitating rhabdomeres.