ABSTRACT
This paper describes an investigation of the physiological characteristics of the light receptors situated in the skin of the ammocoete larva of the brook lamprey, Lampetra planeri (Bloch). The interest of these organs lies in the fact that they are the only type of simple photoreceptor among the Vertebrates of which we possess any precise information. They may well be considered a kind of functional link between the simpler types of photoreceptor systems, which are found in many Invertebrates, in Pholas, Mya and Ciona, for example, and the fully developed image-forming eye of Vertebrates. We are at present almost completely ignorant of the photosensitive processes of animals at levels of organization below that of the image-forming eye, and an analysis of the properties of these simpler types of receptors offers one of the most promising approaches to an understanding of the common pattern of photoreceptor processes in general.
The eyes of the ammocoete are rudimentary organs buried beneath the skin of the head, and do not appear to act as light receptors. The creature normally remains in its tube in the mud or sand of the stream-bed during the daytime, but may swim freely in the water at night. They were sometimes seen by day swimming in the water, usually under the shade of a bridge or overhanging tree ; but when disturbed in the course of collecting operations, most ammocoetes promptly buried themselves in the mud once more.
Few observations have been published on the reactions to light of the lamprey or its ammocoete larva. Parker (1905) found that a swimming reaction was started by shining a beam of light on the tails of quiescent lampreys. Young (1935) demonstrated that the whole skin is more or less sensitive to light, but that the region of greatest sensitivity is in the dorsal fluke of the tail at about its widest point. He showed also that the reaction is a simple photokinesis without directional properties, and that the sensory pathway from the receptors to the central nervous system is by the lateral line nerves and not by the spinal cord. Francis & Horton (1936) published an account of the reactions of ammocoetes to various stimuli, including some experiments on their relative sensitivity to light of different colours. They concluded that the animals are insensitive to light of wave-length longer than 550 m/x, and most sensitive in the blue-green region of the spectrum. Their data, however, were not recorded in a form from which the spectral sensitivity of the reaction can be calculated.
The experiments described below were designed to measure the spectral sensitivity, intensity discrimination, dark adaptation and the effect on the reaction time of varying the intensity and duration of the stimulus ; and to present the information in a form which would enable comparison to be made with similar data from other animals.
Material and methods
All the animals used in the following experiments were ammocoetes of Lampetra planeri (Bloch), collected from the River Tyne, East Lothian. They were kept for use as required in the laboratory in large glass aquarium jars filled with river water over a bottom of the fine sandy mud, containing rotting leaves and other debris, from which they were collected. They can be kept in this way for an indefinite period, and require no feeding or other attention.
The optical arrangement of the apparatus used for light stimulation is shown diagrammatically in Fig. 1. A 150 c.p. a.c. Pointolite was arranged to project horizontally an accurately paralleled beam about 2 in. in diameter. In the path of the beam were placed a glass cell containing water to absorb heat and ultra-violet radiation, a camera shutter with adjustable diaphragm, neutral wedges and colour filters as required. The light intensity was varied by a neutral wedge of graded density 0-6, which could be slid horizontally along a pair of grooves attached to the mounting of the camera shutter. A counter-wedge of the same density gradient was mounted immediately in front of the first to provide a uniformly illuminated test field. The filter holders were also attached to the same mounting. Immediately above the dish containing the animal the beam was reflected vertically downwards by a 90° prism with silvered back, and brought to focus by a convergent lens. The size of the test patch was varied simply by raising or lowering the mounting which held the dish containing the ammocoete.
The following general procedure was adopted in all experiments. The ammocoete to be tested was placed in water in a shallow rectangular glass vessel, which was a little longer than the animal and wide enough to permit its natural swimming movements. The depth of water was just sufficient to allow it to maintain its normal posture when at rest. Despite the small volume of water, a few millilitres only, small ammocoetes showed no signs of respiratory distress during experiments which sometimes lasted for several hours. The surface area of the vessel was relatively large, and presumably the animal’s periods of activity disturbed the surface of the water sufficiently to maintain the necessary interchange of gases.
The test vessel was mounted on a platform attached to a movable stage, which permitted a few centimetres of movement in both horizontal planes. The ammocoetes were not usually disturbed, provided the adjustments of position were carried out slowly and without jerking. For making tests the apparatus was adjusted so that the test patch illuminated the whole of the animal’s tail, which lay diametrically across it with the tip on the far circumference, as shown in Fig. 2. This arrangement could be set up simply and repeatedly, so that the same area of skin was stimulated at every test. The size of the test patch was adjusted so that the centre of the field illuminated the dorsal lobe of the tail at its widest point. For an ammocoete of 50-60 mm. length, the test patch was about 5 mm. in diameter. Adjustments of position were made in deep red light, which was obtained by placing an Ilford no. 609 filter in the path of the beam. Nearly all the radiation transmitted by this filter is of wave-lengths longer than 660 mµ, and did not disturb the ammocoete. If the animal settled in a different position after a burst of swimming, the vessel was adjusted by means of the controls of the movable stage until the tail was brought back to its correct position relative to the test patch.
Swimming reactions were observed in dim red light from a 2 V. lamp mounted above the apparatus. This also was fitted with a piece of Ilford no. 609 filter, and its Properties of the photoreceptors of brook lamprey 353 brightness could be adjusted by means of a rheostat to the minimum required by the observer. A piece of black cardboard was placed over the anterior end of the ammocoete to screen it from stray light, although the head is actually less sensitive than the tail. Before starting a series of experiments the animals were left.in darkness for at least 20 min., and were again dark adapted for about 10 min. after each group of measurements.
Analysis of reaction time
To obtain measurements of dark adaptation, spectral sensitivity and other ‘visual’ functions in a form suitable for quantitative comparison with similar data from other animals, it is necessary to measure the light energy required to produce a constant physiological effect. When, as in the present case, we are concerned with a photokinetic reaction of the whole animal, we have in practice to measure the intensity of stimulus needed to elicit a constant response. The definition of a constant response is therefore of critical importance, and for this reason it was necessary first to investigate the relation between the intensity and duration of the stimulus in initiating a swimming reaction. It was clear from Young’s (1935) experiments that the mechanism of the reaction of the ammocoete is similar to those of Mya, Pholas and Ciona, which were analysed by Hecht (1918, 1921 a, b, 1928). The time interval between the onset of the stimulus and the start of the swimming reaction is made up of an initial sensitization period, followed by a latent period, during which it is believed that the primary products of the photochemical reaction are engaged in initiating nervous impulses, and during which the illumination need not be continued. At high intensities of stimulus the sensitization period is much shorter than the latent period, but at levels close to the threshold of response it may be considerably longer. In the following experiments the minimum threshold of sensitivity of the dark adapted ammocoete was adopted as the index of a constant response, and the intensity/time relation had therefore to be examined for periods of stimulation up to several seconds duration.
The latent period was estimated by exposing dark adapted ammocoetes to flashes of light of high intensity and short duration, and measuring the time interval between the flash and the start of the swimming reaction. Using flashes of 0·2 and 0:1 sec., the latent period for different animals varied between 1·0 and 1·6 sec., measured to the nearest 0·2 sec. Young’s estimate was 1·8 sec:, and his minimum sensitization period o·4−0·6 sec. These discrepancies are not important, however, since the sensitization period depends upon the intensity of light available, while the latent period is affected by a number of environmental factors, such as the temperature, and probably varies a little with the size of the animal also. Young worked principally with large ammocoetes 100−150 mm. in length, while mine were all small ones of about 50 mm.
The intensity/time relation of each ammocoete used for other experiments was measured over a range of 2·5 to 3·0 logarithmic units of illumination, which gave reaction times up to nearly 10 sec. at the lowest intensities. Five measurements of reaction time were made at each level of illumination, and the average calculated. All the animals yielded similar intensity/time curves, though the actual reaction times showed some variation. Typical curves obtained from two ammocoetes are shown in Fig. 3. These show clearly that as the intensity of illumination was increased, the reaction time decreased to a limiting value, which in most cases was about 1·2 sec. At the other end of the range, the reaction times became more variable as the intensity of the stimulating light approached the animal’s absolute threshold of sensitivity.
But in Mya, as also in the ammocoete, the total reaction time, r, is made up of a sensitization period, t, folowed by a latent period, p ; from which it can be shown that p
Equation (3) is of a straight line type, from which it follows that if the reaction is adequately described by equation (2), a graph of the total reaction time plotted against the reciprocal of the logarithm of the intensity should yield a straight line, cutting the time axis at p units above zero. Hecht found this to be the case for Mya. In Fig. 4 the data for the same two ammocoetes as in Fig. 3 are presented in this form. It is clear that the relation is adequately expressed by a straight line for reaction times longer than about 2 sec., and for practical purposes we may conclude that the intensity/time relation of the ammocoete conforms to the Bunsen-Roscoe Law in the form expressed by equation (2) for relatively long reaction times and low intensities of illumination.
It follows from.this analysis that the stimulus required to elicit a constant response can be defined so as to take into account variations in the reaction time in successive measurements. The criterion adopted was the intensity of illumination required to initiate a swimming reaction in 3·0 ± 1·0 sec., expressed in terms of a reaction time of 3·0 sec. In practice the procedure was as follows:
The camera shutter was opened and a stop-watch started simultaneously, the latter being stopped as soon as a reaction was observed. If there was no response, the shutter was closed at 5 sec., the wedge controlling the intensity was moved to a new position and the ammocoete left for about 2 min. in darkness before repeating the stimulus. Reactions which occurred less than 2·0 or more than 4·0 sec. from the onset of the stimulus were disregarded, while those which took place between these limits were adjusted to the intensity value corresponding with a reaction time of 3·0 sec. Reaction times were measured to the nearest 0·2 sec., and the correction was made empirically from the intensity/time relation curve for each ammocoete.
Dark adaptation
The increase in sensitivity of the photoreceptors in darkness, following a preliminary period of light adaptation, was measured for three ammocoetes. The animals were light adapted by illuminating the whole test cell from above with an unshaded 15 or 100 W. lamp mounted at a distance of 12 in. They were restless, and frequently swam continuously during the period of light adaptation, but usually settled down within a minute or two in darkness. Measurements of the threshold of sensitivity were begun as soon as the ammocoete ceased swimming, and were continued at intervals as near to 2 min. as possible until dark adaptation was judged to be complete. To make a measurement, the wedge controlling the light intensity was set initially at a level of illumination estimated to be below the threshold, and the shutter was opened for 5 sec. If no response was observed, the level of illumination was increased by about 0·15 log unit and the test repeated after an interval of about half a minute. This procedure was repeated until a swimming response was obtained, when the position of the wedge and the time elapsed since the beginning of dark adaptation were recorded. After each response the animal was left in darkness for at least a minute before starting the next measurement.
This procedure is cumbersome compared with the methods developed in recent years for measuring the dark adaptation of the human eye, since relatively few readings could be completed during the period when the sensitivity of the tail was increasing most rapidly. Preliminary trials, however, enabled the observer to estimate the threshold fairly accurately, and thus effect considerable economy in the number of subthreshold stimuli required before obtaining a response.
Two typical dark adaptation curves from the same ammocoete, shown in Fig. 5, serve to illustrate the principal features of the process. In all cases adaptation followed a simple course, and was complete in 20-30 min. It was more rapid, and the final steady state attained more quickly following short periods and low intensities of preliminary light adaptation than after longer periods and higher intensities. The increase in sensitivity was small compared with the range of dark adaptation of most Vertebrate eyes, being only about 1·0-1·5 log units of intensity.
Although the measurements throughout this work were recorded in arbitrary units of illumination, the brightness of the test patch was measured with a Sei Visual Photometer. The values so obtained for the final dark-adapted threshold ranged from 0·25 to 0·95 millilambert, i.e. approximately 0·25-1 equiv. f.c.
Intensity discrimination
Intensity discrimination was measured in the following manner. The ammocoete was first adapted to a certain level of illumination for 15 min. The illumination was then increased in successive steps of 0·15 log unit for 5 sec. each until a response was obtained. The same source of light was used both for adapting and stimulating, the shutter being kept open throughout the experiment. The test field was larger than that used in other experiments, being adjusted to illuminate the whole tail end of the animal, whether it was swimming or lying still. As in the dark adaptation experiments, the ammocoetes were restless at higher levels of adapting illumination, but usually remained quiet in dim light. Readings were obtained for two ammocoetes in steps of 0·3 log unit over a range of 3·6 log units. Fig. 6 illustrates the result of one such experiment, in which the differential threshold, or increase in stimulus required to elicit a response, log dI/I, is plotted against the level of adapting illumination, log I. At high intensities the relation conformed closely to the Weber-Fechner Law, log dI/I= a constant, but at lower levels the increment of illumination required to obtain a response increased progressively. There was no sign of a similar increase in log dl/I at the highest intensities.
Spectral sensitivity
Spectral sensitivity curves were obtained by measuring the threshold stimulus required to elicit a response from dark adapted ammocoetes with monochromatic light of different wave-lengths. Spectral bands were isolated by means of combinations of filters, based on the Ilford ‘monochromatic’ series. The optical density and range of spectral transmission of the combinations used were measured with a Hilger Constant Deviation Wave-length Spectrophotometer, and their characteristics are listed in Table I. The relative spectral sensitivity of the ammocoetes was calculated from the average of several observations at each wave-length. The averaged value was then corrected for the relative transmission of the filter combination, the energy distribution and quantum effectiveness of the light source. The energy emission of the 150 c.p. Pointolite used in these experiments approximated closely to that of a gas filled tungsten filament lamp at 2600-2700° K.
The spectral sensitivity curve shown in Fig. 7 was constructed from the combined observations on three ammocoetes, and has been adjusted to equal 2·0 on a logarithmic scale at the maximum. The same data are expressed in Fig. 8 as a percentage of the sensitivity at the maximum. The maximum lies in the green, close to 530 mµ, and the sensitivity falls away steeply towards both longer and shorter wave-lengths, so that at 600 mµ it is only about 2 % of the maximum. There is, however, some indication that the sensitivity rises again in the extreme violet, at wave-lengths shorter than 450 mµ.
Discussion
The measurements described in this paper provide information on which to base deductions concerning the nature and mode of action of the photoreceptors. The shape of the curves obtained for dark adaptation and intensity discrimination suggest that the system contains a single type of receptor and photopigment only, since both functions vary in a continuous manner with respect to time and intensity of adapting illumination respectively. Neither of them shows any sign of a break, or sudden change in slope, such as is commonly found in similar experiments on eyes which contain both rods and cones in the retina, and distinct scotopic and photopic visual mechanisms. The range of dark adaptation of the ammocoete is strikingly small, amounting to an increase in sensitivity after light adaptation of some 10 to 30 times. For comparison, human dark adaptation can, under suitable experimental conditions, show about a million-fold increase in sensitivity, and with relatively simple apparatus can be measured over a range of 10,000 times (4 log units). The final stable threshold of sensitivity of the fully dark adapted ammocoete is also surprisingly high. Francis & Horton (1936) estimated it at about 3 equiv. f.c. The values in my experiments are rather lower, ranging from about 0·25 to 1 equiv. f.c., which is still some 100,000 times less sensitive than the dark adapted threshold of medium-sized retinal fields in the parafoveal region of the human eye. These findings suggest that the photoreceptor system of the ammocoete is not concerned with detecting or discriminating between low levels of illumination, but simply with activating the animal in the presence of bright light.
The general conclusion which may be drawn from the measurements of dark adaptation is that there are relatively few receptor cells compared with the number of retinal elements of an image-forming eye, and that the individual receptors probably lack the capacity to accumulate large amounts of the photosensitive substance. In these respects the physiological findings agree with the morphological, since there are certainly no obvious concentrations of cells in the tail of the ammocoete of any type likely to be photoreceptors, nor has any photosensitive pigment yet been identified or extracted from them.
The spectral sensitivity curve is of special interest. The maximum sensitivity of the ammocoete at about 530 mµ falls between the maxima of scotopic and photopic vision of the human and most Vertebrate eyes. The photopigment concerned cannot be ordinary visual purple (rhodopsin), the absorption maximum of which is close to 500 mµ, nor does it correspond with any known photopic visibility curve, since all such curves possess maxima at wave-lengths longer than 530 mµ and are sensitive to radiation between 600 and 700 mµ in the red end of the spectrum. The position of the maximum at 530 mµ does, however, strongly suggest a photopigment of the visual violet (porphyropsin) type, in which vitamin A2 replaces vitamin A1 in the chromophore. The absorption maximum of porphyropsin lies between 520 and 530 mµ, at 522 ± 2 mµ according to Wald (1939), which is reasonably close to the maximum sensitivity of the ammocoete. Porphyropsin visual systems are characteristic of fresh-water fish, and although direct evidence is not available for the brook lamprey or its ammocoete larva, one would expect them to contain vitamin A2. Wald (1942) found that the eyes of the anadromous sea lamprey, Lampetra marinus, contain predominantly porphyropsin and vitamin A2, though the liver and other tissues contain principally vitamin A1 Spectral sensitivity curves, however, have been obtained for the eyes of some species of fresh-water fish, which possess the porphyropsin visual system. The scotopic sensitivity of the carp, Cyprinus species, and the tench, Tinea vulgaris, were measured by Granit (1941) by means of the micro-electrode technique on the excised opened eye. The curve for the tench corresponds closely with the absorption spectrum of porphyropsin, prepared from the same species by Bayliss, Lythgoe & Tansley (1936), and resembles the spectral sensitivity of the ammocoete in certain important respects. The two curves are compared in Fig. 8, which shows also (filled circles), data by Bayliss et al. for tench porphyropsin plotted as a percentage of the maximum absorption. The maximum sensitivity of both the tench and ammocoete is close to 530 mµ, and both curves tend to flatten about 450 mµ or even to rise again in the violet. The principal difference between them is that the curve for the ammocoete falls away more steeply on either side of the maximum, so that its spectral range of sensitivity appears to be more restricted than that of the tench, or than the absorption spectrum of porphyropsin.
This difficulty may be resolved by taking into account the effect of differences in the concentration or in the thickness of the absorbing layer of the photosensitive substance. Hecht, Shlaer & Pirenne (1942) pointed out that the shape and width of the percentage absorption curve of a substance varies with its concentration, and the spectral sensitivity must represent the percentage absorption curve of a particular concentration of the photopigment. This fact is obscured when the absorption spectra are presented in the form of extinction or optical density curves, as is the usual practice. They computed percentage absorption curves for various concentrations of visual purple, which show clearly that when the maxima are adjusted to the same value (100 %) irrespective of the actual fraction of the light absorbed, the width of the curves increases as the concentration is increased. Similar data are not available for porphyropsin, but the hypothesis that this is the photopigment concerned in the light reaction of the ammocoete can be tested empirically by adjusting the percentage absorption curve of a solution of porphyropsin for various concentrations or thicknesses until the best fit is obtained. The most comprehensive series of absorption spectra for porphyropsin are probably those of Wald (1939), who investigated five species of North American fresh-water fish. The curves all agree closely with one another for wave-lengths longer than about 500 mµ, and the maxima all lie at 522 ± 2 mµ. They diverge to some extent at shorter wave-lengths, all five showing a rise in absorption in the violet, which is probably due in part to impurities. The rise in the violet is greatest in the pickerel, Esox reticulatus, and least in the calico bass, Pomoxis sparoides, and the white perch, Morone americana. The last two probably most nearly represent the absorption spectrum of pure porphyropsin, and I have therefore selected the curve for the calico bass for comparison with the spectral sensitivity of the ammocoete. The data were first converted from extinction to a percentage absorption basis (Table 2), and the resulting curve is shown as the interrupted line in Fig. 9. From this curve a family of percentage absorption curves for different concentrations or thicknesses of porphyropsin can be derived in the same way as for visual purple ; since each layer of equal thickness absorbs an equal fraction of the light, precentage absorption increases by the square as the thickness is doubled, and so on. Fig. 9 illustrates such a transformation, in which the spectral sensitivity of the ammocoete is compared with the percentage absorption of a porphyropsin solution of thickness one-sixth that of Wald’s curve for the calico bass. It is clear that the correspondence between the two is very close over the whole spectral range.
This analysis demonstrates that the properties of porphyropsin provide an adequate basis for the light reaction of the ammocoete. No special significance can be attached to the fact that the animal’s sensitivity curve fits the sixth power of the percentage absorption spectrum, but we may conclude that the conditions in the photoreceptors of the tail are such that the concentration and effective thickness of the absorbing layer combine to yield an action spectrum which corresponds with a percentage absorption curve for porphyropsin, whose maximum absorption is several times lower than that of preparations from the eyes of fresh-water fish.
It is instructive to pursue this line of inquiry by comparing the spectral sensitivity of the ammocoete with the curves for various Invertebrates. We find that the simple light reactions of the Molluscs Mya arenaria and Pholas dactylus, investigated by Hecht (1921 a, 1928), also exhibit sharply defined maxima and relatively narrow spectral limits. These reactions resemble that of the ammocoete in being mediated by diffusely scattered photoreceptors in the skin, and not by image-forming eyes. That of Pholas in particular closely resembles the ammocoete, with the maximum close to 530 mµ, and with a tendency to rise again in the violet (Fig. 10). The curves are indeed so similar that one is tempted to suggest porphyropsin as the photopigment in Pholas also, at a slightly higher concentration than in the ammocoete. This would be surprising, since the bulk of the evidence available indicates that porphyropsin and vitamin A2 are associated exclusively with fresh water or anadromous Vertebrates, and the matter clearly merits further investigation. In Mya, on the other hand, the position of the maximum close to 500 mµ suggests a system based on rhodopsin (visual purple). Although the sensitivity curve was constructed from measurements at six wave-lengths only, it corresponds not unreasonably with the percentage absorption curve for a dilute solution of visual purple.
Sensitivity measurements of the eyes of various Invertebrates show that the latter tend to resemble Vertebrate eyes in possessing broad spectral bands, which do not fall away so steeply on either side of the maximum. Thus the spectral sensitivity of the eye of Limulus is very similar to the human scotopic visibility curve, (Graham & Hartline, 1935), and that of Dytiscus is abroad band with the maximum about 540 m/z, (Granit, 1947). This suggests that broad spectral sensitivities are in general characteristic of image-forming eyes, whether Vertebrate or Invertebrate, which contain large numbers of retinal elements and relatively high concentrations of photosensitive pigments ; while narrow bands with sharply peaked maxima are found in the simpler systems, in which the receptor cells are few and scattered, and the concentration of photosensitive pigment low. So far as it is possible to analyse them at present, the properties of these simple photoreceptor systems appear to be compatible with the known photosensitive pigments of image-forming eyes, and it seems unnecessary to invent special substances to account for their physiological differences.
Summary
A method is described for measuring the principal characteristics of the light reaction of the ammocoete, and data are presented on the relation between the intensity and duration of the stimulus, dark adaptation, intensity discrimination and spectral sensitivity.
The intensity/time relation approximates to the Bunsen-Roscoe Law at low intensities of illumination. Use is made of this fact to define a criterion for a constant response to a given stimulus, for measuring the other physiological properties of the system.
Dark adaptation and intensity discrimination at different levels of adapting illumination both yield simple curves, which suggest that the system contains a single photosensitive pigment. The range of dark adaptation following a period of light adaptation is from 10 to 30 times. The threshold of excitation of a dark adapted ammocoete is from 0·25 to 0·95 millilambert.
The spectral sensitivity possesses a sharply peaked maximum about 530 mμ. Reasons are presented for suggesting porphyropsin (visual violet) as the photosensitive pigment concerned.
ACKNOWLEDGEMENTS
I wish to thank Prof. James Ritchie and Dr T. D. M. Roberts for much helpful advice and criticism in the preparation of this paper.