In this work precaution was taken to eliminate all ex- traneous white light." The Nernst lamp was used as the source at the red end, whereas the arc was used as the source at the violet end. The limit at the red end is approxi- mately 715 j^M- It is probable that if a greater amount of time had been taken to determine the limit for the chick very accurately, it would have been found to coincide with that of the human being, which is ordinarily placed at 760/^yw. The limit at the violet end is approximately 400 MM p which again is almost the same as that of the human eye, 396mm- I^ a long series of experiments the relative stimulating effect of the different regions of the spectrum was determined for the dark-adapted chick in the following way: All of the monochromatic bands, the wave- lengths of which are shown on the abscissa (Fig. 61), were equated in energy. Experiments were begun upon the chick with the green, 550 MM- By cutting down the in- tensity of the light with the Brodhun sector (see descrip- tion of whole apparatus on p. 76) a point was reached at which the chick could not respond correctly (this can be determined accurately since the chick is positive to light was conducted under the auspices of the Marine Biological Labora- tory of the Carnegie Institution. I am indebted to Dr. Alfred G. Mayer and to the Carnegie Institution for permission to use this summary and the descriptions and cuts of apparatus shown on ^ A word should be said concerning the state of adaptation: The chicks were brought into the dark room in a fairly tightly-woven wicker basket. They were left in total darkness for 15 minutes. The experimenter then entered the room and turned up a 16-c,p. tantalum light, placed 1 meter above the home compartment of the animal box. The chicks were taken out and run in the same order' every day. One chick at a time was placed in the home compart- ment; then the light was turned out. A wait of 5 seconds ensued and then the door was opened and the animal confronted with the stimulus. As soon as the correct response was made the light was turned up. The chick ate of the grains for a few seconds. It was then driven into the home compartment, when the light was again turned out for 5 seconds and the routine repeated. That darkness-adaptation was well advanced appears from the fact that the point of highest stimulating value for the author's eye lay at 520/uy[t. If the threshold had been taken upon the eye in the (approximately) light-adapted state the point under consideration would have shifted over to about oSOfMix to QQOfip..
338 VISION and since only one light was used) or else would not leave the home box at all. The sector reading which is directly proportioned to the energy was noted. When the reaction threshold had been satisfactorily determined in this region, another wave-length was selected and the same routine repeated. We thus obtain the threshold at any given wave- length in terms of the sector opening. For purposes of comparison, an assistant took the author's threshold each day after the work on the chick had been completed. Care was taken to keep adaptation for his eyes the same as for that of the chicks. It is clear that the greater the energy required by any given monochromatic light to stimulate the retina, the less the sensitivity of the retina to that ray; i.e., the reciprocals of the energy value (1/e where e equals the energy necessary for stimulation) will, when plotted against w^ave-length, give the usual sensitivity^ curve. In examining the curve it is well to remember that the higher the point on the curve is above the base line, the less the energy required to produce stimulation. We do not know in absolute terms (ergs) the standard energy carried by each beam with which we worked, but we do know that all had the same initial energy and consequently that the final sector openings at which the various thresholds were obtained, are all directly comparable and proportional to the absolute energy required for stimulation. Hence it is legitimate for us to take the reciprocals of the sector open- ings. Below we give a table showing the actual sector open- ings for chick and man, and the reciprocals of those values.
Wave- Sector in mm., Becip- Sector opening Reciplength chick rocal mm., man rocal * The opening at 660/x/a is of course merely proportional. At all other wave-lengths a smoked wedge was used after the energy was calibrated, the transmission curve of which, from 396w»ji. to 760m/u-» was known (calibrated by Bureau of Standards). At GGOju-/* the efficiency was so low that the wedge had to be removed.
SPECTRAL RANGE OF CHICK If now we allow each millimeter on the ordinate to rep- resent.01 these reciprocals, when plotted against wave- length, yield the two curves shown below in Fig. 61.
Sensitivity Curve ( Dark- Adapted Eye) A, chick; B, man. The ordinate shows the reciprocal (1 mm. nr.01 ) of the energy value necessary to stimulate the retina at the wave-length given on the abscissa.
It follows from these experiments that the chick is not blind to blue and violet rays. On the contrary, these rays are highly efficient in stimulating effect. In general the sensitivity curve for chick and man are strikingly alike. The absolute threshold for man is lower than that of the chick except in the extreme red.
Birds with twilight vision. — Several birds with rod 340 VISION retinge, and especially birds with twilight vision, have been under laboratory observation. Among these are the kestrel (a bird similar to our sparrow hawk), house hawk, small owls, and the horned owls. A bright-adapted kestrel, according to Hess, has a range in the red coextensive with that of man. Its range is short at the violet end. A dark-adapted kestrel has a somewhat wider range at the violet end than the light-adapted animal. But even a dark-adapted animal responds to food stimuli less accurately and quickly in the blue and violet regions than in the red and yellow. "When the kestrel and the chick are tested together simultaneously, the one with bits of veal and the other with grains illuminated by the whole spec- trum, it is found that each will eat the food from the red to the green, but the kestrel will find the food further along in the green than the chick. When tested with food il- luminated by blue and red ray filters it was found that the kestrel rarely ever touched the food illuminated by the blue, even after 20 hours of adaptation, unless the blue was made very intense. The horned owl has a range both in the violet and in the red much more nearly like that of man. Even the owl is slow and clumsy in its movements in blue and violet light. The dark-adapted owl eats first in a spectrum of weak intensity in the regions offering the great- est stimulating effect to the human eye.
Hess' absorption theory. — It is well known that the chick and most other day birds have retinae richly supplied with cones but lacking in rods. Hence we have in the day birds little or no visual purple. Just the reverse is the case with birds possessing twilight vision. Their retina abound in rods but contain few cones. In some forms, like the owl, rods are almost completely lacking. In the cones of the day birds we find red to reddish-orange-colored oil globules and similarly in the retinal rods of the birds with twilight vision we find orange to greenish-yellow-colored oil globules, at times colorless globules. Hess supposes that the shortened range in the violet of the birds and other forms possessing these oil globules is brought about by the physical fact that the short waves are stopped by the SPECTRAL RANGE IN FISH 341 globules. This affords the possibility of accounting, on physical grounds, for the asserted lack of sensitivity in the violet. Our own experiments showing that the spectrum is not shortened in the violet for the chick at least tend to prove that one of two things must be true: (1) either the oil globules do not absorb the short rays as Hess maintains, or else (2) the chick's retina is so sensitive to these rays that the loss is compensated for by greater sensitivity.
In fish. — Hess shows that the dark-adapted fish (Atherina) collect almost immediately in the neighborhood of the yellow-green to green, between th^ lines E to b. That this is the point of highest stimulating value comes out clearly from the fact that when the container is shaded one can drive the fish at will towards the red end or towards the violet end. This method (p. 69) enables one to test sensitivity at the two ends of the spectrum. It is found that their range at the red end is shorter than that of the human being normal in color vision. This was shown by shading all of the spectrum up to orange. As the card is moved along the fish advance before it, collecting in the light. Long before the human limits are reached in the red the fish begin to respond as to darkness: i.e., to swim aimlessly under the shadow of the card, out into the red, and then into the infra-red — i.e., they cease to collect. In general it is found that the common fresh and salt water fish collect in the neighborhood of the yellow-green to green. The stimulating value of the spectrum decreases from this point towards the red end very quickly. The stimulating effect is remarkably decreased by the time the pure yellow is reached. The orange and the red rays have exceedingly little stimulating value. The decrease in stimu- lating efficiency is much less apparent as one goes from the yellow-green towards the violet.
In reptiles and amphibia. — The limit at the red end for reptiles (turtle) is the same as for man, but their range is even more limited at the violet end than that of the chick.^ ^ Again in the case of the reptiles Hess falls back upon his ab- sorption theory. In the retina of the turtle only cones are present. The cones contain reddish-orange oil globules.
342 VISION By methods similar to those Ave have already considered Hess shows that several species of frogs have a range in the spectrum coextensive with man's. This is in agreement with the work of Gotch which appeared several years be- fore that of Hess.^ Gotch showed that in the excised eye of the frog the photo-electric responses fail or become ex- tremely feeble if the eye is stimulated by infra-red or ultra- violet rays. It follows from this that the range of light vibrations within which the frog's eyeball gives definite photo-electric changes corresponds very closely to the range which produces visual response in the human being. The point of greatest stimulating value (spectrum of low intensity) for dark-adapted amphibia is in the neighbor- hood of the yellow-green to green. This point can be made to shift for amphibia as for birds by changing the animal's state of adaptation and by increasing or decreasing the in- tensity of the spectrum, i.e., here also a Purkin^e effect is demonstrable (Hess).
II. Darkness-Adaptation; White Light Introduction. — The current view in sensory physiology holds that darkness-adaptation is connected with the ac- cumulation of visual purple which takes place while the eye is undergoing adaptation to darkness. Two phenomena have been urged in support of this theory: (1) The as- serted lack of adaptive increase in sensitivity in foveal vision, after adaptation; (2) the supposed fact that birds with day vision, whose retinge are sparsely supplied with rods and consequently contain, little visual purple, do not possess the power to adapt to darkness. It has been shown by Piper and others that the fovea in the human being does have the power to adapt to darkness. The work of Hess cited below shows conclusively that the course of darkness-adaptation in day birds is not very different from that found in man or in birds with twilight vision. We have not repeated Hess' work in detail, but of the fact that an enormous increase in sensitivity comes with dark- * But not mentioned by Hess.
ness-adaptation where the chick is the subject there can be no question. We have had in sensory physiology to give over the attempt to make of the visual purple a photo- chemical substance responsible for the general phenomena of color; it would seem necessary also to give up the at- tempt to connect it with the phenomena of adaptation to light.
Darkness-adaptation in birds with daylight vision. — When chicks are bright-adapted for one hour and then taken suddenly into a dark room and placed before grains of food illuminated by white light passing through an Aubert diaphragm (working over a ground glass disk), a very noticeable increase in sensitivity can be shown. If the pecking threshold is taken in terms of the Aubert diaphragm, it may be shown that chicks cease to peck at the points at which the grains become invisible to the human observer. E.g., in one test it was shown that im- mediately after entrance into the dark room, the light- adapted chick ceased to peck when the Aubert diaphragm was set at 21 mm. The grains ceased to be visible to the human observer at 20 mm. After five minutes' adaptation, the limit for the chick was 9 mm. The grains ceased to be visible to the human being at 9 mm. In another case a human being, a chick, and a white pigeon were tested simultaneously. The subjects were bright-adapted and then taken into a dark room. After one hour's darkness- adaptation the limits were taken (in this case with an iris diaphragm). The relations may be shown in terms of the divisions on the iris diaphragm as follows: White pigeon 3 J Chick 7i In all cases one hour in darkness produces complete adaptation. It follows from this that the course of adap- tation to white light in day birds is closely similar to that in man.'^ It is interesting to note that chicks, when long ■^ The increase in sensitivity produced by darkness-adaptation may be, shown roughly in birds by a very easy experiment. If a chick VISION dark-adapted, are not blinded by strong light. Small grains placed before a dark-adapted chick and then sud- denly illuminated by bright sunlight, are picked up readily by the birds. Birds and turtles stand in interesting con- trast to man in this respect. Man is blinded by strong light after darkness-adaptation. Whether this is due to the fact that man possesses rods, and hence visual purple, is not clear at present.
In birds with twilight vision. — Tests by Hess upon the light-adapted kestrel with the Aubert diaphragm (see above, p. 343) show that the animal ceased to eat immediately after entrance into the dark room when the Aubert diaphragm was set at 40 mm. The food ceased to be visible to the human observer at 12 mm. In another test with the iris diaphragm the limits of visibility for the observer were 7 to 8 mm. (di- ameter of opening of iris). After a half hour's darkness- adaptation the kestrel's limit on successive trials was 7J, 5J, and 5^; for the human being, even after 5 minutes in the dark room, the limit on successive trials was 6J, 4J, and 4|; after 10 minutes in darkness, 4J, 4, and 4J. On other specimens the following course of adaptation was obtained: Bright-adapted kestrel on entrance After l^ hour After 114 hours Human being immediately upon entrance After 1/2 hour After 114 hours The range of adaptation to white light in the horned owl corresponds a little more closely to that of the human being. The bright-adapted owl upon entering the dark room had a limit on the Aubert at 24 mm.; the human observer at 18 mm. After a quarter of an hour in darkness the owl's with one eye blindfolded is light-adapted and then brought sud- denly into a dark room and the bandage removed from its eye, it will, when placed before a row of grains, take the food on the side of the dark-ad? pted eye.
limit was 14 mm., and the human observer's was also 14 mm. The owl is not light-shy in the common meaning of that term. The animal sees very well in bright light. With the possible exception of the owl adaptive increase in sensitivity in birds with twilight vision is not greater than in the day birds. Under decreasing illumination they seem to see much more poorly than the human being. The in- crease in sensitivity comes on much more slowly. The total range of adaptation is about the same as in man, although, with the exception of the owl, the absolute threshold is not quite so low.
In fish. — After 10 to 15 minutes darkness-adaptation sensitivity to light is increased over one-thousand-fold in fish. This adaptive increase is in part physical, due to the wandering of the retinal pigment towards the vitreous humor under exposure to light. In darkness the pigment recedes and clears the retinal elements for the reception of the light. In man the pigment is almost stationary. A large part of the effect in fish, however, is due to the gen- eral physiological processes with which adaptation is con- nected (Hess).
In reptiles and in amphibia. — Reptiles (various species of turtles have been mainly worked upon), although pos- sessing a retina totally unsupplied with rods, nevertheless show a marked increase in sensitivity to light when dark- adapted. If a light-adapted animal is placed with its back to the apparatus admitting graded white light, it is found to be an easy matter to find the limiting intensity at which the food will be struck; a dark-adapted animal will strike at a piece of moving meat which is invisible to the bright- adapted human eye. The course of adaptation is almost identical with that of the human eye when the latter is covered with a piece of orange glass (to compensate for the absorptive effect of the red oil globules). The turtle responds to objects surprisingly well in very high illumi- nation. The process of adaptation in the amphibian eye (frogs of various species) seems to follow closely that of the human eye. But the absolute increase in adaptive sen- sitivity seems to be somewhat less. An interesting way 346 VISION 346 VISION to test the process of light-adaptation in such animals is for the observer to light-adapt only one eye. This eye is kept in the bright light for the same length of time that the subject's eyes are light-adapted. Then when the animal is taken into the dark room the experimenter, with his dark-adapted eye, can arrange the apparatus, while with his light-adapted eye he can test the course of adaptation in himself and compare it directly with that of the animal under observation (Hess).
III. Darkness-Adaptation: Monochromatic Light In birds. — Chicks are somewhat similar to man and reptiles in the range of adaptive increase in sensitivity to monochromatic light. On account of the presence of the oil globules in the retinge of the chick man needs, as in testing reptiles, to put an orange-colored glass before the eye when examining the course of adaptation. When man and chick are tested under these conditions, close similarity in the course and the range of adaptive increase can be noted between them from the green region to the red (Hess).
In reptiles and in amphibia. — The course of adaptation to monochromatic light in reptiles is similar to that in the chick. In an actual experiment, where food was illumi- mated by a red ray filter, the intensity of which could be altered by moving the lamp behind it, it was found that to the bright-adapted human eye the food was just invisible when the distance between the lamp and the filter was 8 cm. The reptiles under these conditions snapped at the food ^ith some sureness. After one half hour's darkness-adap- tation the food was visible to the human being with the lamp 17 cm. behind the filter. The turtles, dark-adapted for the same period, had their threshold at 17 cm. In general the course of adaptation is exactly the same in the turtle as in the human being when the latter observes the spec- trum through an orange-colored glass. This holds until the" blue is reached, beyond which point the spectrum of the turtle does not extend — i.e., has such a high threshold that SENSITIVITY TO WAVE-LENGTH 347 comparison between human beings and reptiles is impos- sible. Here again, as in day Birds, we have animals which are lacking in cones, and hence in visual purple, showing a considerable range of adaptive sensitivity to monochro- matic light. What is still more surprising is the fact that these animals are almost exclusively nocturnal. When the frog's eye and the human eye are tested under comparable conditions the course of adaptation for the tw^o forms is closely similar throughout the range of the spectrum. It will be recalled here that the spectrum is not shortened in the violet for the amphibian. The range of adaptive in- crease for the several amphibians tested was practically the same as for man (Hess).
IV. Sensitivity to Wave-length Of mammals. — All of the tests, taken at their face value, upon monkeys seem to point to the fact that they are sensitive to difference in wave-length. Most of the work upon them has been carried out with colored papers or colored foodstuffs as stimuli. It was shown several years ago that the M. rhesus can learn to respond positively to one of a series of colored boxes or glasses, regardless of its position in the series. When the rhesus is presented simultaneously with bits of food dyed with a given color- ing matter and soaked in a solution of quinine, and with bits of food dyed with a different coloring matter and left free of quinine, it learns wdth astonishing quickness to select only the bits free from quinine. The writer's own experi- ments upon the rhesus and cebus, carried out with mono- chromatic light, gave similar evidence of discriminative ability. In one animal the habit of responding negatively to yellow and positively to blue arose very quickly. Popu- lar literature is full of assurances that the dog is sensitive to wave-length difference. When tested with colored papers, colored metal plates, etc., the results are everywhere confirmative of the popular view. The dog can learn to bring red, green, yellow, and blue balls upon command, as was shown long ago by Nagel. But in such experiments 348 VISION there has been no serious effort to control the various fac- tors which must be taken into consideration in every such test. The more careful work of Nicolai, using the method of Pawlow with ray filters as the source of monochromatic light, fails to confirm the view that the dog can respond to wave-length difference. The dog, according to Nicolai, reacts wholly upon the basis of intensity. Even careful tests with colored papers (Smith) show that the dog's sen- sitivity to wave-length difference must be very rudimen- tary. Several other mammals have been tested in the various laboratories by the colored paper method. Among the animals so tested are raccoons, cats, and porcu- pines. The porcupine and raccoon strangely enough show little ability to respond to colored papers in such a way as to indicate sensitivity to wave-length difference. The cat has recently been more carefully tested with colored papers. It is found that yellow and white of the same flicker equivalent have apparently the same stimu- lating value; that Bradley standard Mue has the same value as dark gray cambric. On the other hand, red is not reacted to differently from black; whereas green has the stimulating value of ^^ dark gray." It appears that a gray can be found which will break down the adjustment to any color (Cole). White rats and mice have been tested exten- sively. In tests with colored objects, such as colored yarns, papers, etc., they exhibit, as most animals do, the ability to form habits which have been interpreted as show- ing color sensitivity. More careful experiments upon the dancing mouse, where ray filters were used as sources of light, failed to give any definite evidences of such sen- sitivity. Extensive tests upon albino rats, black and white rats, and rabbits with monochromatic light failed to give positive evidence of sensitivity to wave-length difference. The red-green and yellow-hliie habits were formed after a large number of trials, but it was shown in the red-green habit that the green was the only effective stimulus, the animals probably being completely insensitive to the red rays. In the yelloiv-hlue habit, while both stimuli were effective, it was shown that the yellow was low in stimu-