cells at a higher magnification. The general structure of the retina is very like 15 PHYSIO. CXXx

Transcription

1 225 J. Physiol. (955) 3, THE SPECTRAL SENSTVTY OF THE PURE-CONE RETNA OF THE SOUSLK (CTELLUS CTELLUS) BY G. B. ARDEN AND KATHARNE TANSLEY From the nstitute of Ophthalmology, London (Received 29 April 1955) n an earlier paper (Arden & Tansley, 1955) we described the spectral sensitivity curve of the apparently pure-cone retina of the grey squirrel obtained by means of the electroretinogram. This curve was unlike those hitherto reported for any whole eye, whether rod, cone or mixed, in that it was narrow and had a maximum at about 525 my. n view of this unexpected finding we thought it would be of interest to examine another member of the family Sciurus, the European ground squirrel or souslik, Citellus citellus. This animal is also diurnal and has a retina which, on histological grounds, is generally agreed to contain nothing but cones. Some preliminary observations on its electroretinogram have recently been published by Bornschein (1954). METHODS The general arrangement of the experiments was similar to that already described. All the results reported here were obtained on two animals under light Nembutal (sodium pentobarbitone) anaesthesia. Two more animals were killed, one for histological examination of the retinae, while a lens from the other was handed over to Dr R. A. Weale who measured its spectral absorption for us. All the electroretinograms were recorded with the Karpe machine, and a specially made corneal contact glass incorporating the active electrode. As before the indifferent electrode was a cottonwool pad soaked in.9% NaCl solution (saline) applied to a shaved area on the forehead by means of electrode jelly and held in position with a rubber band. The duration of the stimulus flashes was 1 sec and the same coloured and neutral tint filters were used as before. These were all recalibrated and their transmissions were found not to have changed. The conditions for the 'light-adapted' and 'dark-adapted' experiments were those already described. RESULTS The structure of the retina. Two photomicrographs of the souslik retina are shown in Pl. 1 (p. 232). One (fig. 1) gives a general low-power view of the whole thickness of the retina, while the other (fig. 2) shows more detail of the visual cells at a higher magnification. The general structure of the retina is very like 15 PHYSO. CXXx

2 226 G. B. ARDEN AND KATHARNE TANSLEY that of the squirrel, with almost as many ganglion cells as there are visual cell nuclei. The proportion seems to be about two to one over most of the retina, but more nearly one to one around the optic nerve. Vilter (1954) gives a figure of 2, cone nuclei to 9, ganglion cells for the whole retina. The optic nerve is, again, very big for a rodent and the nerve head is even more elongated than in the squirrel. The main difference between the squirrel and souslik retinae is in the visual cell layer. Both animals possess two types of cone, an inner and an outer, but whereas there are about equal numbers of each in the squirrel, the souslik has very few inner cones. These inner cones have long striated outer limbs visible between the inner limbs of the outer cones. The outer limbs of the outer cones are shorter and are buried in the pigment of the pigment epithelium. They are only visible in sections where the retina has come away from the pigment epithelium. The outer cones appear to have an ellipsoid between the two limbs; they also have long myoids which can sometimes be seen making contact with the outermost nuclei of the outer nuclear layer. As in the squirrel the pigment of the pigment epithelium lies in long processes outside the epithelial cells, between them and the cones. The electroretinogram. The normal light-adapted souslik electroretinogram is shown in Text-fig. 1. t has a very marked a-wave followed by a b-wave and an off-effect. There is no c-wave. The record is essentially the same as that from the squirrel, but the a-wave and off-effect are relatively bigger. There is no change in shape nor in the relative sizes of the waves when either the state of adaptation or the colour of the stimulus light is altered. Text-fig. 2 shows the relationship between stimulus intensity and the size of the response for two filters, the green (521 m,u) and the yellow (578 m,u), both for the b-wave and for the off-effect. n all cases the b-wave height was measured from the trough of the a-wave. The values for the yellow filter were corrected by a figure representing its stimulating power in comparison with that of the green filter. The yellow light was always a less effective stimulus than the green, the difference being about equivalent to a neutral filter of -3. n the experiment illustrated in Text-fig. 2 we never got large responses to the yellow stimulus and some of those to the low intensities were too small to be measurable, especially in the case of the off-effect. This is the reason for the small number of points on this stimulus-response curve. However, after the correction was applied the agreement between the yellow and green points is fairly good. The stimulus-reponse curves for white were also very similar and, since we had already found the responses to different wavelengths to resemble one another in the closely related grey squirrel, we felt justified in treating all the results together in the way described below. However, we later discovered that this assumption is probably not always correct for blue stimuli. We shall return to this point in the discussion.

3 SOUSLK ELECTRORETNOGRAM 227 A sensitivity curve was plotted separately for each experiment on the basis of the results given in Text-fig. 3. n this a stimulus-response curve to white light is given for the b-wave (upper left) and the off-effect (lower left) while the Text-fig. 1. The souslik electroretinogram. Note the marked a-wave, pointed b-wave and large off-effect. Stimulus duration: 1 sec. Calibration: 5FV. Taken with Karpe's machine and a contact glass electrode ,, , 2 -~~~~~~ ~~~~ ~~~~~~~~~~ Density of neutral filter Text-fig. 2. ntensity-response curves. Left: the curve for the off-effect. Right: the curve for the b-wave. *, green (521 m,);, yellow (578 m4). corresponding responses to each wavelength are shown on the right. From these curves we reduced all the responses to a common arbitrary level, the height of the response to white light with a neutral density filter of 1.. Using the filter calibrations and the correction for the deep yellow souslik lens taken 15-2

4 228 G. B. ARDEN AND KATHARNE TANSLEY from Text-fig. 4 we calculated the intensity necessary to produce a constant response at each wavelength. For the 'light-adapted' experiments there was a constant illumination of about 3 e.f.c. at the eye and the various stimuli were superimposed upon this. The 'dark-adapted' experiments were done in the darkened room after about ,, 5 *= 4 D3 -o ' w Density of neutral filter Wavelength (mpu) Text-fig. 3. Derivation of the sensitivity curves. Left: intensity-response curves for white light for the b-wave (above) and the off-effect (below). Right: mean size of response at each wavelength from the same experiment. 1 min initial dark adaptation. n each case a series of records was taken using white light of different intensities followed by a series with the coloured lights. Each coloured run was made from short to long wavelength and back again and was followed by another white run. Often several white and coloured runs were made interspersed with one another. n this way we could detect any instability in the preparation. f there were the least indication of a consistent change, such as a steady drop in the response compared with earlier records, all the affected results were discarded.

5 SOUSLK ELECTRORETNOGRAM 229 Text-fig. 5 shows the average of the b-wave and off-effects for four lightadapted and two dark-adapted experiments expressed as a percentage of the maximum. The points in Text-fig. 6 are the average results for all the b-waves and all the off-effects irrespective of the adaptation state. t will be seen that there was no Purkinje shift and that the maximum lies between 52 and 53 m,u. Although the position of the maximum and the long-wave part of the curve are similar to those already published for the grey squirrel, the whole curve is not a narrow one because of the high values for the shorter wavelengths F 6 p >;- -5 -o4- Ga - o.- L -v9 2- Text-fig A Wavelength (mgi) Spectral absorption of the souslik lens. These results were obtained for us by Dr R. A. Weale. 7 DSCUSSON The squirrel and souslik are both Sciuridae and are both unusual among mammals in having diurnal habits. n addition, the structure of their retinae is very similar, both being almost certainly pure-cone. t is not, therefore, surprising that they should have electroretinograms which are so much alike and of the type usually associated with cone activity. Bornschein (1954) has already shown that the souslik, like the squirrel, has a high light-threshold and is capable of very little adaptation to darkness. We have fully confirmed these observations in our own experiments. Further, we have found no more evidence

6 23 G. B. ARDEN AND KATHARNE TANSLEY E 1 9o o O _._E U 8 E 7 U 6 v W 5 wu 4 _~~_ O ED (D ~ _ :> 3 V 2 1 v ' O _ U _ Wavelength (mu) Text-fig. 5. Spectral sensitivity of the souslik. The points were obtained from four light-adapted and two dark-adapted experiments and are given as percentages of the mean maximum., light-adapted b-wave; O, light-adapted off-effect, +, dark-adapted b-wave;, darkadapted off-effect. Equal quantum intensity spectrum. 1 E 9._E E 8 7-6,- 5 L 4 +-f ; 3 u 2 1 -,. 1 1 u Wavelength (mm) Text-fig. 6. Mean spectral sensitivities for the b-wave and off-effect. The points are given as percentages of the mean maximum. The vertical lines represent twice the standard error., b-wave;, off-effect. Equal quantum intensity spectrum.

7 SOUSLK ELECTRORETNOGRAM 231 of a Purkinje shift in the souslik than we did in the squirrel and a maximum spectral sensitivity in the same wavelength range for both. There is, however, one marked difference between the spectral sensitivities of the two species. The mean souslik curve is much broader on the short-wave side of the maximum. The scatter of the points (and therefore the standard errors) is also much greater in this region. When plotting the sensitivity curves for individual experiments we found that these fell into two distinct groups of which typical examples are shown in Text-fig. 7. n some experiments the sensitivity curve was a narrow one, but in others the values for 47 and 492 m,u were unexpectedly high. Text-fig. 7 B shows an extreme example. 1-9 A B - 15 ~~~~~8 ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~ -12.t~6 9 ~7-,8f- -~ C: L 3~~~~~~~~~~~~~~~~~~~~~~ Wavelength (my) Text-fig. 7. ndividual sensitivity curves. A: sensitivity curve of dark-adapted off-effect. Mean of five runs on souslik 1 (1. xii. 54). B: sensitivity curve of light-adapted b-wave. Mean of two runs on souslik 1 (16. viii. 54). The recording of high values at the shorter wavelengths was erratic, although they tended to appear more often in the b-wave than in the off-effect records. This tendency is reflected in the relatively higher values for the b-waves in Text-figs. 5 and 6. These findings suggest that, in addition to a retinal mechanism with a maximal sensitivity between 52 and 53 m,, there is another with its maximum at a shorter wavelength (perhaps near 47 m,u) which may or may not make its effect apparent. f such a second mechanism really exists in this pure-cone retina we are, at present, entirely unable to explain why it should sometimes manifest itself and sometimes not, nor why it should appear to be more closely associated with the b-wave than with the off-effect. This variability of response to blue light with its suggestion of a second mechanism

8 2322. B. ARDEN AND KATHARNE TANSLEY sometimes, but not always, manifesting itself, makes it unlikely that we were justified in assuming similar stimulus-response curves to all wavelengths in all experiments. For this reason we have not drawn a sensitivity curve through the points in Text-fig. 6. Obviously a great deal more work needs to be done on the reactions of the souslik retina to short wavelengths before all the problems of its spectral sensitivity are solved. SUMMARY 1. The structure of the souslik retina is described briefly. t appears to be a pure-cone retina and in many ways resembles that of the grey squirrel. 2. Electroretinograms were taken from the souslik eye in both light and dark adaptation. The record is typical of a pure-cone retina and its shape is not affected by the state of adaptation nor by the wavelength of the stimulus. 3. Spectral sensitivity curves, both for the b-wave and for the off-effect, were calculated, using the amount of light required to produce a constant response as the criterion of sensitivity. 4. The sensitivity curve had its maximum between 52 and 53 mp. There was no Purkinje shift. 5. The sensitivity curve was similar to that of the squirrel in so far as the position of its maximum and its shape between 52 and 612 m,u were concerned. Between 52 and 47 m, the mean sensitivity was higher than in the squirrel. 6. The possibility that a 'blue mechanism' may sometimes be active in the souslik retina is considered. REFERENCES ARDEN, G. B. & TANSLEY, K. (1955). The spectral sensitivity of the pure-cone retina of the grey squirrel (Sciuru8 carolinen8i8 leucoti8). J. Physiol. 127, BOUNSCHE, H. (1954). Elektrophysiologisoher Nachweis einer -Retina bei einem Sauger (Citellus citellus L.). Naturwi8enwschaften, 41, VTER, V. (1954). Histologie et activit6 6lectrique de la retine d'un Mammifere strictement diurne, le Spermophile (Citellu8 citellu8). C.R. Soc. Biol., Pari8, 148, EXPLANATON OF PLATE The souslik retina Fig. 1. Section through whole retina: note the wide inner nuclear and optic nerve fibre layers and the nearly equal numbers of visual cell nuclei and ganglion cells. ntravital fixation with Flemming's fixative; azan stain. x 378. Fig. 2. High-power view through the outer retina showing the cones and their nuclei. ntra-vital fixation with Flemming's fixative; azan stain. x 7. Key: 1, pigment epithelium; 2, cones; 3, outer nuclear layer; 4, outer fibre layer; 5, inner nuclear layer; 6, inner fibre layer; 7, ganglion cell layer; 8, optic nerve fibre layer.

9 THE JOURNAL OF PHYSOLOGY, VOL. 13, No. 1 PLATE Aw, ~~~~~~~~~~~~~~~~~~~ ~~ ~ ~ ~~~ior~~~. '7t~~~~~~~~~~~~~~ Fig. 1. * :~~~~~~~~~~~~ -- R - l~~~~~~~~~~ 3 Fig. 2. (Facing p. 232)