during a stay in the dark and therefore, since the photochemical effect light-sensitive substance, it seems probable that the regeneration of

Transcription

1 6I [ II + 6I7.75I.7 THE REGENERATION OF VISUAL PURPLE: ITS RELATION TO DARK ADAPTATION AND NIGHT BLINDNESS. BY KATHARINE TANSLEY'. (From the Department of Physiology and Biochemistry, University College, London.) INTRODUCTION. THE duplicity theory explains the striking differences between the reactions of the eye to illuminations above and below.1 f.c. by assuming that the cones are responsible for vision at high illuminations and the rods for vision at low illuminations. Further, it ascribes the reactions of the rods to low illuminations to the bleaching of the photochemical substance, visual purple. In the eye visual purple is bleached by light, but reappears slowly in the dark (Kiihne [1877] found three hours necessary for regeneration in the frog); at the same time the eye becomes slowly more sensitive during a stay in the dark and therefore, since the photochemical effect of a given amount of light is proportional to the concentration of the light-sensitive substance, it seems probable that the regeneration of visual purple is responsible for dark adaptation. Most of the workers on dark adaptation have confined themselves to a study of the subjective changes which occur, and up till now there has been no attempt at a quantitative study of the regeneration of visual purple which might link up the two phenomena more satisfactorily. This paper is a description of an attempt to obtain a curve of the regeneration of visual purple from rats during a stay in the dark, and to compare this curve with that of normal dark adaptation. The condition known as night blindness is in many cases simply a failure in dark adaptation, though whether the failure consists of a slowing-up of the process, in which case the patient would merely take longer to reach the normal level, or whether the normal level is not 1 Working with a grant from the Medical Research Council (Physiology of Vision Committee).

2 REGENERATION OF VISUAL PURPLE. 443 reached at all is not clear. Probably both types of failure occur. v. Kries [186] considered that night blindness was caused by a failure of the rod apparatus just as cbmplete colour blindness was due to a failure of the cones. Parinaud [1881] did a good deal of work in the light-sense and colour-sense in night blindness and considered that the relation between these two formed one of the most essential supports of the duplicity theory. He believed the poor dark adaptation in this condition to be due to a faulty regeneration of visual purple which prevented the eye from reacting to low intensities. One common cause of night blindness is now known to be vitamin A deficiency, and it readily yields to treatment with cod-liver oil. The complaint has always been common among peoples living under poor conditions and the use of liver as a traditional cure is very widespread. The earliest reference we have is in Eber's Papyrus [see v. Leersum, 124] dated about 15 B.C. It was also recommended by Hippocrates. Night blindness has received very little attention in the laboratory, probably owing to the difficulty of demonstrating the condition in animals. Holm [125] showed that vitamin A-deficient rats would not jump off a pedestal on to a dimly lighted shelf where normal rats showed no difficulty, and Sugita [125] found that deficient rats could not find their way out of a dark box when the illumination on the exit was low, but had no difficulty when it was raised. Control animals escaped easily at the low illumination. Since vitamin A deficiency certainly causes night blindness in man and probably also in animals one would expect it to affect the production of visual purple in the retina of the dark-adapted animal. Work on these lines is made extremely difficult by the fact that the colours of the retinas have to be compared in the light, in which they are rapidly bleached. Yoshiere [125] working on dogs attempted to compare the retinae of vitamin A-deficient animals after 1 hour in the dark with those of normal animals under the same conditions. He was unable to get any results owing to great variations in colour in the normal retinae. Fridericia and Holm [125] were more successful. They used albino rats and compared the retinas not with each other, but with a modification of Garten's [16] colour chart. They found a gradual increase of visual purple in both normal and deficient animals during a stay of 3 hours in the dark. At any time during these 3 hours there was always less colour in the deficient retinas than in the controls. They found, however, that after 12 hours in the dark there was no difference between the two types of retina. These results support the view that the failure in dark adapta-

3 444 K. TANSLEY. tion, at least when induced by vitamin A deficiency, is merely one of retardation and that the normal level will be reached in time. Further work on these lines obviously requires a better method for the estimation of the retinal content of visual purple, and for this purpose a solution of visual purple is necessary. In this investigation observations have been made on the effect of vitamin A deficiency on the visual purple content of the retinae of albino rats. By making extracts of visual purple from animals which have been in the dark for varying periods after light adaptation and measuring the density of these extracts by the photographic method devised in collaboration with Lythgoe [12], it has been possible to plot regeneration curves for both deficient and normal rats. The description of this work will be divided into two parts, one dealing with the method of making an extract of visual purple, and the other with the way in which the regeneration curves were obtained. PART I. THE EXTRACTION OF VISUAL PURPLE. Kiuhne extracted visual purple from the eyes of frogs by removing the retinae, treating them with a bile-salts solution and filtering off the debris, but this method is useless for quantitative work since only a fraction of the visual purple is obtained. Hecht [12 b] improved this method by centrifuging down the debris. He was able to obtain quantitative results on his solutions. Both Hecht and Kiihne used sodium glycocholate as an extractive, and both are emphatic that it must be specially pure, in the form of a white powder that will dissolve in water to give a clear, colourless solution, and that commercial preparations are useless. Unfortunately neither gives experimental details for the preparation of this pure form, and none of the preparations obtained was satisfactory as an extractive. Certain specific properties (haemolytic and cytolytic) of bile salts are shared by the glucosides saponin [Ponder, 124] and digitonin [Loeb, 113], and it was found that both these substances will extract visual purple from the retina. Of the two digitonin is the more powerful extractive and is also more convenient in use, since it will dissolve in hot water to give a clear, colourless solution which will keep for any length of time. Saponin solutions have to be kept in cold store and even then become acid on standing. For the following experiments, therefore, a 2 p.c. solution of Me r c k's digitonin in water was used. The digitonin is dissolved by boiling it with water. After several days it may reprecipitate and also become acid by absorbing C2, but it can be restored to its original condition by boiling.

4 REGENERATION OF VISUAL PURPLE. 445 Using such a digitonin solution practically all the visual purple can be extracted from the retina. Visual purple cannot be extracted from the retinal tissues by simple osmotic cytolysis-distilled water, even after several hours' application, has no effect whatever. Kiihne showed that the ordinary extractives in use in biochemical work (e.g. alcohol, chloroform, ether) destroy the purple colour. This is also destroyed if the solution is made markedly acid or alkaline, the best results are obtained atph 7. Bile salts and saponin are powerful haemolytic agents which not only rupture the envelope of the blood corpuscle but also completely destroy the organized elements of the cell-" stromatolysis" [Ponder]. The power of lowering the surface tension of a solution is also common to these two substances and digitonin, but it was found that the extraction of visual purple does not seem to be correlated with the surface tension of the solution. Ponder also found that the heemolysis of the red-blood corpuscle by saponin and bile salts is not dependent on the surface tension. There is also some evidence that the effect of temperature is similar in the two cases. The rate of h.emolysis by saponin is greatly increased by raising the temperature, and it was also found that the extraction of visual purple took place more quickly at 3 C. than at 15 C. If the temperature is raised above 3 C. proteins, which are also present in the solution, coagulate. It would seem, therefore, that the extraction of visual purple from the retina and the heemolysis of the red-blood corpuscle with subsequent stromatolysis are similar processes. The extraction of visual purple from rats' eyes. Nine albino rats (Wistar strain obtained from the Glaxo laboratories), of average age 8 days and average weight 15 g., were put into the dark overnight. Each animal was lightly aneesthetized with chloroform and its back broken. The retina was removed from each eye and dropped immediately into 1 c.c. of freshly boiled 2 p.c. digitonin solution in a small test-tube. The retina can easily be removed from the eye without taking the latter out of the head. The points of a pair of fine curved forceps are inserted behind the eyeball, which is thus made to protrude and at the same time kept steady. While the forceps are held in position the cornea is cut away with scissors. Usually the lens comes away with the cornea, but if it does not it may be picked out with forceps. The retina can now be lifted out of the eye. An advantage of this way of removing the retina is that the compression of the blood vessels and optic nerve by the forceps prevents blood getting into the preparation. By this method both retinae can be taken out of a rat in under 2 minutes. The retinee were stirred up and kept in a thermostat at 28 C. for i hour. The retinal debris was then separated by centrifuging for

5 446 K. TANSLEY. 1 minutes at 2 r.p.m., and the supernatant fluid containing the visual purple poured off. All work, including centrifuging, was carried out in a dark room with a red light (Wratten filter No. 7). The solution obtained by this method is deep red in colour and quite clear. It bleaches to a light yellow in daylight. This solution cannot be made to regenerate its colour in the dark. Digitonin solution can also be used to make preparations of visual purple from frogs' eyes. In this case the frogs were killed by cutting off the head behind the eyes as in a decerebrate preparation. The head was divided by a longitudinal cut between the eyes and the retina removed from each eye in 8ttu as in the rat preparation. Visual purple solutions obtained from frogs are not suitable for estimation by the photographic method, as they are usually contaminated by a small amount of black pigment from the choroid. PART II. THE REGENERATION OF VISUAL PURPLE. The regeneration curves were obtained by plotting the density of the visual purple extract obtained from a batch of rats against the length of time for which the batch had been in the dark. Each batch was light adapted before being put into the dark. The first value was obtained immediately after light adaptation and the final value after a whole night (17-1 hours) in the dark. Rats were used because so much is known about their dietary requirements, and it is easy to induce vitamin A deficiency in them. The albino eye gives a good extract of visual purple free from pigment and can also be light adapted without the use of atropine- to dilate the pupil [Fridericia and Holm]. Method. The rats used were of the Wistar Institute Tyler strain, bred in the Glaxo research laboratories. They were about 27 days old when they arrived and were put straight on to the special diet. Batches of nine animals were used with the litters and sexes represented as evenly as possible in each. Both lots were put on to the vitamin A-deficient diet which was made up as follows: Rice starch (heated at 15 C. for 12 hours) 7 g. Glaxo casein (tested for vitamin A deficiency) 2,, Salt mixture ,, Dried yeast (to provide vitamin B)... 5,, Each animal was also given 1 mg. irradiated ergosterol in liquid paraffin twice a week to provide vitamin D and each control animal received its vitamin A in one drop of cod-liver oil per day. Rats do not require vitamin C.

6 REGENERATION OF VISUAL PURPLE. 447 As soon as the mean weight of a batch of rats showed a definite fall as compared with the controls, the animals were taken for an experiment. This loss of weight occurred after about 4 days. Xerophthalmia was unusual, but eyes which showed even a slight degree of corneal opacity were discarded, since this interfered with light adaptation. The apparatus used for light adapting the rats consisted of a wire platform supporting the cage containing the rats, lit from above and below through ground glass. In order to prevent the temperature inside from becoming too high (above 25 C.) the platform round the cage was packed with ice and there was an electric fan on the platform. The water from the melting ice, together with the rats' droppings, fell through the wire and between two slanting sheets of glass into a pan on the floor. The fourth side of the box was detachable so that the rats were easily put in and taken out. The inside walls were painted white. The illumination on the rats' eyes was about 4 f.c. It was found that j hour inside this box was sufficient to cause almost complete bleaching of the animals' retina. The rats did not seem to mind the light and kept their eyes open during their stay in the box. All nine animals were placed in the box and after i hour the first one was removed into a small numbered cage and taken into the dark room. Three minutes later the next was put into the dark and so on. After the required interval (e.g. 2 hours) in the dark the first rat was killed, its retinas removed and put into the digitonin solution, and 3 minutes after the death of the first animal the second was killed. In this way each individual was in the dark for exactly the right length of time. As soon as the extraction was completed the solution of visual purple was transferred to the cell of the photographic apparatus and three photographs taken of it. The visual purple, still in the cell, was then bleached in daylight and three more plates exposed. Thus, each point on the curve represents the difference in density between the unbleached and bleached solution. The depth of the cell used in these experiments was 1-15 cm. Five points were taken for each curve: immediately after light adaptation, after i hour, 1j hours, 2 hours, and 4 hours in the dark. Each point was repeated once. Each control batch was used as soon as possible after the deficient one-usually on the following day. The complete set of readings for both curves was, repeated some months later. In this second series of experiments it was surprising to find that the amount of regeneration was consistently less than in the first series, but for both normal sets the general shape of the curves was the

7 448 K. TANSLEY. same. The possible reasons for this discrepancy between the results for the two series will be discussed later. Results. The results of both series for normal rats are given in Table I exactly as they were obtained-those for deficient rats in Table II. In Tables III and IV the normal and deficient values for the quantity of visual purple obtained are given in order of the time for which the rats were in the dark and these values are corrected, first, for the number of eyes used and, second, for the age of the rats when they were killed. TABLE I. Quantity of visual purple extracted after varying times in the dark. Initial light adaptation; 3 min. to 4 f.c. illumination. Normal rats Density of unbleached solution on arbitrary Time scale (mean Age Number in dark of three Date in days of rats (mins.) readings) 15. x x xi xi xi xi xii xii xii xii iii iii iv iv v vi vi vii * 7j* 8 8 1st series nd series * One rat with only one eye. t Assuming that Beer's law holds in this case, concentration * * *8 4*87 3* Density of bleached solution on arbitrary scale (mean of three readings) 1-4 1* *55 2* * *55 2*63 7 Density of visual purple (difference between cols. 5 and 6)t * * *25 1* * these figures are proportional to the The first correction was based on the assumption that the visual purple is distributed evenly between the retinal tissues and the digitonin solution when in the test-tube, and that the average volume of one retina is 16 c.c.

8 REGENERATION OF VISUAL PURPLE. 44 TABE II. Quantity of visual purple extracted after varying times in the dark. Initial light adaptation; 3 min. to 4 f.c. illumination. Vitamin A-deficient rats Density of Density of unbleached bleached Density solution on solution on of visual arbitrary arbitrary purple Time scale (mean scale (mean (difference Age Number in dark of three of three between Date in days of rats (mins.) readings) readings) cols. 5 and 6)t 1st series 12. xi * xi xi x21 2x6 x xi xii *87 2. xii xii nd series 18. iii *34 1'5 '75 1. iii j* iv ?* iv v vi *68 - * One rat with only onee e. t Assuming that Beer's law holds in this case these figures are proportional to the concentration. A further correction must be made for the age of the rats at the time of the experiment. Jackson [113] has shown that the weight of the eyeball varies with the age rather than with the weight of the animal, particularly under poor conditions of nutrition. A graph was constructed from Jackson's results on the variation of the weight of the eyeball with body weight-these have been confirmed at the Wistar Institute [Donaldson, 124]. For body weight the corresponding age in days for normal animals was substituted as found by Donaldson, and for the weight of the eyeball a value for the area of the eyeball (weight*) was substituted. The correction to a standard age of 8 days was then read off directly from the graph. The final corrected values were plotted against the time in the dark, the curves for both deficient and normal animals of the first series are plotted in Fig. 1, for the second series in Fig. 2. It will be seen that in every case the value for the deficient rats is well below that for the normal rats taken from the same litters. In order to be sure that a high value for the visual purple extract did not merely mean that a larger proportion of the retinm had been

9 45 K. TANSLEY. el PW =1 Time in dark (hours) Fig. 1. The regeneration of visual purple in animals kept in the dark for different periods. 1st series. Normal rats: o vitamin A-deficient rats. The curve was calculated from the equation for a bimolecular reaction. C) Ca 1 Time in dark (hours) Fig. 2. Similar to Fig. 1. 2nd normal rats: vitamin A-deficient rats.

10 REGENERATION OF VISUAL PURPLE. 451 TABLE III. Corrected values for the quantity of visual purple extracted. Normal rats Density ol Value from Densities calculated from culated from Densities cal- visual purple col. 3 further Time Density of solution coir- corrected the formula the formula in dark visual purple rected foi for age from Table I number of rats of rats K=1 x (min) t A (A-x) X tlogi (A $) 1st series Exact age unknownl Exact age unknown 2nd series removed, the test-tube with the digitonin solution was weighed before and after the retina were added. It was found that there was no difference between the average weights for normal and deficient retina. The actual values, which remained very constant in fourteen experiments, were 16, 16, 15, 16, 13, 15, 17, 18, 16, 1, 15, 15, 16 and 18 mg. The remarkable agreement between the results of the first series is, of course, an accident. It is impossible, considering the errors that there must be in an experiment of this type, for the results to agree so closely from day to day. The readings on the normal curve for the second series are distributed in a more likely manner. DIsCUSSION. It will be seen from a comparison of Figs. 1 and 2 that the readings for the first series were consistently higher than the corresponding readings for the second series. Although the actual experiment was

11 452 K. TANSLEY. TABLE IV. Corrected values for the quantity of visual purple extracted. Vitamin A-deficient rats Density of Value from visual purple col. 3 further Densities for Time Density of solution cor- corrected normal rats from in dark visual purple rected for for age same litters (min.) from Table II number of rats of rats (Table III) 1st series (3 min.) (3 min.) nd series (114 min.) carried out in exactly the same way in the two cases, there was a variation in the previous treatment of the rats. The second series was done in the spring, the room was well lighted and its temperature constant at 21 C., whilst the first series was done in the autumn, the animals were habituated to a lower illumination in a darker room and the temperature was lower, varying between 1 and 17 C. Rats, even at maturity, have a poor mechanism for the regulation of their body temperature, and animals brought from a cold to a warm room will show a higher temperature than those habituated to the warm room. It was thought possible, therefore, that the rats which were kept at a lower temperature might show an increased metabolism in the higher temperature of the light box and dark room compared to those which were used to a temperature of 21 C., and that this increased metabolism might accelerate the regeneration of visual purple. Some experiments in which the body temperatures of a set of rats were taken under both conditions showed, however, that although the "cold" rats did show a higher mean temperature in the dark room than the " warm " ones, the difference (about.7 C.) was not sufficient to account for the differences between the two curves. It is possible that an eye which is accustomed to a high illumination

12 REGENERATION OF VISUAL PURPLE. and, therefore, to a continual bleaching of the retina, may have its capacity for regenerating visual purple impaired. If this were so one might expect the amount of visual purple produced in such an eye in the dark to be less than that produced in an eye habituated to a lower illumination. There is a theory, for instance, that night blindness is due solely to a continual bleaching of the retina by bright lights, e.g. the sun reflected off snow or water, and that it can be cured by keeping the patient in the dark for some hours [Netter, 1858, 1863]. The difference between the two sets of results for the deficient animals of both series is very striking. In the first series they show a steady rise to a value (after 4 hours) very close to the normal, while in the second series they do not fall on a curve at all but are merely a set of isolated points, some of which are extremely low. This discrepancy is, I think, due to the fact that the figures represent the values for different degrees of deficiency. In a batch of nine rats, however uniformly they are treated, it is impossible to get all the animals to the same stage of deficiency at the same time. One or two animals may die while some are still gaining in weight, and in nearly every case, although the average for the batch had begun to fall, there were a few animals still gaining in weight1. In my opinion, the rats of the second series were more deficient than those of the first, and this, together with the changed conditions of temperature and illumination, may account for the much lower values of this series. These results show that a method which involves the use of more than one animal for each reading is not suitable for work on vitamin A deficiency. They also show, however, a qualitative decrease in the amount of visual purple in the dark-adapted retina of the deficient rat. It may be that vitamin A deficiency causes a slowing of the regeneration of visual purple varying with the degree of deficiency until, when this is sufficient, regeneration is stopped altogether. This view would help to reconcile the two theories of night blindness, one, held by Netter, Parinaud, and Fridericia and Holm, that regeneration is slowed but that the normal value is reached after long enough in the dark, the other held by Hess [1], that the final normal value is never reached at all. In this connection it is of interest that several observers, among them Forbes [1811], Berry [1886] and Dumas [188], have reported that night blindness may come on gradually, the patient taking a little longer 1 The difficulty of obtaining young rats which behave in a uniform manner on diets deficient in fat-soluble vitamins has been dealt with by Chick [126] and Chick and Smith [126], PH. LXXi

13 454 K. TANSLEY. to become dark adapted each night. In all the cases mentioned by these authors the night blindness was probably due to vitamin A deficiency. The normal results for each series are far more consistent for, although the second curve is lower than the first (Fig. 3), they are both of the same shape. Aubert [1865], Charpentier [1886], Piper [13] and Nagel [111] have all shown that, for the human eye, dark adaptation begins quickly and, then, after the first i hour gradually slows down. Unfortunately none of these workers took systematic readings after the first hour, so that their results cannot be compared with mine. A c hmat o v Time in dark (hours) Fig. 3. The regeneration of visual purple in normal rats kept in the dark for different periods. ) 1st series: 2nd series. The curves were calculated from the equation for monomolecular reactions. [126] took readin:gs for dark adaptation over a period of 24 hours. He uses the reciprocal of the light threshold to represent the sensitivity of the eye (E) at.any given time and plots this value against the time in the dark (Fig. 4). In a photochemical reaction the amount of breakdown of the sensitive substance is proportional to the intensity (I) of the light and the concentration of substance present, that is, for visual purple Breakdown = KI [v.p.]... (i). If the breakdown of visual purple by light is the basis of the stimulation of the eye, there must be some relation between the amount decomposed

14 REGENERATION OF VISUAL PURPLE. 455 and the resulting stimulus. The simplest relation that we can assume is that a constant amount of visual purple must be broken down to produce a stimulus. If the breakdown is kept constant in equation (i) we get [v.p.] = K"II...1/ (ii) = K"'E...(iii). In this calculation an important factor, time, has been omitted. Equation (i) should, of course, read Breakdown =KI [v.p.]. t...(iv) _4) P_z ) C. '. *Q I q Time in dark (hours) Fig. 4. The dark adaptation of the human eye (Achmatov). I=least illumination visible at different times after going into the dark. The curve was calculated from the equation for a bimolecular reaction. Unfortunately, no dark-adaptation curves appear to have been taken with the time of exposure of the stimulating light kept constant so that it is impossible to include the time in any calculation based on dark-adaptation readings. However, it seems likely that one is justified in neglecting the time factor in connection with these curves since, in practice, variations in the time of exposure are probably very small compared to the wide variations in intensity. Therefore, if this assumption is correct, the sensitivity of the eye is proportional to the amount of visual purple present, and a curve showing

15 45i6 K. TANSLE Y. the change in sensitivity during a period in the dark is comparable to a curve showing the change in amount of visual purple under the same conditions. It is clear that the curves in Figs. 1, 2 and 4 compare very well. Attempts have been made further to explain the mechanism of dark adaptation in terms of the laws of mass action. Hecht [118, 11, 127], working on the "dark adaptation" of the ascidian Ciona intestinalis, the clam Mya arenaria and the lamellibranch Pholas dactylus, obtained curves representing the process and he found that these curves all had the same equation as that for a bimolecular reaction. It was on these curves that he based his theory that the process depended on the bimolecular formation of a photochemical substance, although no such substance has yet been found in any of these animals. Hecht [12 a] also believes that the formation of visual purple in the dark is a simple bimolecular reaction according to the equation I K = t'a A (A -x) -$... (V), () where K is a constant representing the velocity of regeneration, t the total time in the dark, A the amount of visual purple after complete regeneration, and x the amount of visual purple regenerated in the time t. If the regeneration of visual purple is a bimolecular reaction my results should fall on curves represented by this equation. In calculating such a curve from the equation a suitable value for A must be assumed which will give the most constant value for K. For the first series of experiments the best value for A is 2*56 which makes K =, and in the second A = 2. when K becomes * again. The curves calculated from these results are given in Figs. 1 and 2, and it will be seen that they agree very well with the observed values of x. Therefore, it is certainly possible that visual purple is being produced as the result of a bimolecular reaction in my experiments and, if this is the case, that the velocity of the reaction (K) is the same in both series. It is also possible to fit Achmatov's results for the human eye for the first 4 hours to a curve calculated from the same equation. The curve in Fig. 4 is calculated on the assumption that A = 2-8 and K = 7, the points are Achmatov's own readings. His later readings, however (between 5 and 24 hours), will not fit any curve that can be calculated from this equation. It is doubtful whether these calculations have any real value in considering the problems of dark adaptation. In each series of results there are five points only and each one has

16 REGENERATION OF VISUAL PURPLE. a considerable experimental error. The curve of the equation which it is proposed to fit to these points is one whose general shape is suitable. This equation contains two constants (A and K) whose values are painstakingly adjusted until the best fit is obtained. In making the calculations one is surprised at the latitude possible in the values for A which do not involve marked inconstancies in the resulting K. It is unfortunate that, in determining the probability of a given chemical reaction being bimolecular, it is most important to have an accurate value. for "A." This can be found in the chemical laboratory, but is not possible in this type of physiological experiment unless a very large number of readings is taken. It is, therefore, particularly unfortunate that in both curves the erperimental values corresponding to A are poor. In series I the exact age of the rats was unknown and in series II there was only one reading. It is also possible to fit the experimental values with a curve for a monomolecular reaction, that is, K = - o A )... V) K= log1(a -x) (vi), 457 where the symbols have the same meaning as in equation (v), K = 7 and A = 2-25 for series I, while K = 6 and A = 1-7 for series II. The curves plotted from these values are shown in Fig. 3 from which it will be seen that the fit is slightly better than for a bimolecular reaction equation. An equation of the type (vi) does not exclude the possibility of a purely physical reaction such as diffusion under certain conditions. The better fit of the theoretical curves in Fig. 3 as compared with those in Figs. 1 and 2 is perhaps more apparent than real owing to the uncertain values after 1 hours. Until an experimental value corresponding to A can be fixed it is impossible to choose between the two forms of curve. Altogether, I am unwilling to draw from my curves conclusions which are based on mathematical reasoning, since I do not consider that the results are sufficiently accurate to allow of such treatment. In any case, however, the fact that these curves for the regeneration of visual purple in vivo so closely resemble the corresponding part of the curve for the dark adaptation of the human eye is certainly very striking. SUMMARY. 1. A simple method of extraction of visual purple from the eyes of frogs and rats by digitonin is described. 2. Using this method the course of regeneration of visual purple in the rat has been followed. 3. A comparison has been made between the regeneration curves for normal and vitamin A-deficient rats. 4. The relation between the normal regeneration curve of visual purple and the dark adaptation of the human eye is discussed togethi with a possible chemical mechanism.

17 458 K. TANSLEY. I have to express my thanks to Dr R. J. Lythgoe for the valuable advice and assistance which he gave me throughout the course of this investigation. REFERENCES. Achmatov, A. S. (126). Pfluegers Arch. 215, 1. Aubert, H. (1865). Physiologie der Netzhaut. Breslau. Berry, G. A. (1886). Edinb. Med. J. 31, 125. Charpentier, A. (1886). Arch. d'ophtal. 6, 24. Chick, H. (126). Biochem. J. 2, 11. Chick, H. and Smith, H. H. (126). Biochem. J. 2, 131. Donaldson, H. H. (124). The Rat. Philadelphia. Dum as, A. J. A. (188). These de Paris. Forbes, J. (1811). Edinb. Med. J. 7, 417. Fridericia, L. S. and Holm, E. (125). Amer. J. Physiol. 73, 63. Garten, S. (16). Graefes Arch. 63, 112. Hecht, S. (118). J. Gen. Physiol. 1, 147. Hecht, S. (11). J. Gen. Physiol. 1, 545. Hecht, S. (12 a). J. Gen. Physiol. 2, 4. Hecht, S. (12 b). J. Gen. Physiol. 3, 1. Hecht, S. (127). J. Gen. Physiol. 1, 781. Hess, C. (1). Arch. Augenhlk. 62, 5, trans. in Arch. Ophthal. 3, 482. Hippocrates. (Wuvres d'hippocrate. Trans. by Littr6, 1861,, 15. Paris. Holm, E. (125). Amer. J. Physiol. 73, 7. Jackson, C. M. (113). Amer. J. Anat. 15, 1. v. Kries, J. (186). Graefes Arch. 42, 5. Kiihne, W. (1877). Heidelb. Untersuch. 1, 1 et seq. v. Leersum, E. C. (124). Ned. Tijcischr. v. Geneesk. 18, reference to Eber's Papyrus. Loeb, J. (113). Artificial Parthenogenesis and Fertilisation. Chicago. Lythgoe, R. J. and Tansley, K. (12). J. Physiol. 68, 45. Nagel, W. (111). v. Helmholtz, Handbuch der physiologischen Optik, 2, 264. Hamburg. Netter, A. (1858). C.R. Acad. Sci. Paris, 4, 842. Netter, A. (1863). Des cabinets teme'breux dams le traitement de l'hemeralopie. Paris. Parinaud, M. (1881). Arch. ge'n. Me'. 1, 43. Piper, H. (13). Z. Psychol. Physiol. Sinnmesorg. 31, 161. Ponder, E. (124). The Erythrocyte and the Action of Simple Haemolysis". Edinburgh. Schryver, S. B. (112). J. Physiol. 44, 265. Sugita, Y. (125). Graefes Arch. 115, 26. Yoshiere, S. (125). Arch. Augenhlk. 5, 14.