The Animal Mind: A Textbook of Comparative Psychology

(802). Special vigor and speed generally characterize reactions to contact with moving objects. In eliciting the scratch-reflex of dogs, an object drawn along the skin is decidedly more effective than one pressed against the skin for the same length of time (681, p. 184). The physiological effect is probably, Sherrington says, the same as that involved in the “summation” of successive slight stimuli applied at the same point. As is well known, the latter will bring about a response of considerable violence, though each stimulus acting alone would apparently be without effect.

Is it likely that these responses to moving stimuli in contact with the skin involve the perception of movement as a form of space perception; that is, a perception of the successive positions occupied by the stimulus and their relative direction? I think we may say that they probably do not, in the lower animal forms at least. And a chief reason for saying so lies in the fact that the reactions are so rapid. To perceive the spatial relations of stimuli, or any other relations, is a process not favored by great speed of response. The quicker the reaction, the less clear the perception of its cause: such seems to be the general law. The sensation accompanying contact with a moving object may differ in intensity from that accompanying a resting stimulus; it may, in the lower forms, differ qualitatively in some way not represented in our own experience, but it can hardly be connected with the more complex psychic processes involved in any form of space perception.

In vision, also, there are special arrangements for reacting to moving stimulation. The sensitiveness of many animals to changes of light intensity, although not a direct adaptation to the spatial characteristics of a stimulus, serves the same purpose, for changes in light intensity are oftenest brought about by objectsin motion. In the mollusk Pecten varius, a transition from shadow vision to movement vision is illustrated: the animal closes its shell when a shadow is moved so as to fall on its eye spots in rapid succession (628). Generally speaking, the simple invertebrate eye, however, is adapted to respond to changes in light intensity rather than to moving objects. Plateau found that caterpillars, which have only simple eyes, could see moving objects no better than those at rest (597), and Willem was inclined to think snails saw resting objects better than moving ones (788). On the other hand, the compound eye (see page 219) is specially formed to be affected by moving stimuli. The crayfish will react to anything of fairly good size in motion, but is apparently unable to avoid stationary objects in its path (40). The poor vision of the compound eye for resting objects isshown by the ease with which insects may be captured if the movements of the captor are very slow. They may be readily approached, also, if the movements are all in the line of sight, that is, directly toward the insect, so that successive facets of the compound eye are not affected, as would be the case in lateral movements. Let the reader try bringing the hand slowly straight down over a fly, and see how much closer he can come before the fly is disturbed than he can if the hand is moved from side to side. Plateau, from experiments on different orders of insects, concludes that “‘visual perception of movement”’ is best developed in the Lepidoptera (moths and butterflies), Hymenoptera (ants, bees, and wasps), Diptera (flies), and Odonata (dragon-flies); that the distance at which movements can be seen does not exceed two metres, and averages 1.5 metres for diurnal Lepidoptera, 58 cm. for Hymenoptera, and 68 cm. for Diptera (599).

through the eye may be accompanied by spatial perception of movement, although if the eye is compound, the experience must differ from our own visual movement perception. By an image is meant the perception of simultaneously occurring but differently located stimuli as having certain spatial relations to each other. Through its means, or that of the nervous processes underlying it, there arises the possibility of adapting reaction not merely to the location of a single stimulus, but to the relative location of several stimuli. Responses may thus be adjusted not only to the direction of an object but to its form. On the basis of such adjustments a whole new field of possible discriminations is opened up.

The commonest arrangement for the production of a visual image is the double convex lens, which collects the rays of light diverging in their reflection from an object and brings them together again upon the sensitive retina. The lenses found in many simple invertebrate eyes seem, however, very ill adapted to the image-producing function. It is probable that they serve rather to intensify the effect of the light rays by bringing them together, than to give a clear-cut image (523). In the eye of certain invertebrates, such as the Nautilus, a cephalopod mollusk, while there is no lens, the opening admitting the light rays is so small that an inverted image might be formed through it, such as may be obtained through a pinhole. It is unlikely, however, that this eye is really an image-producing organ. Hesse includes under image-forming eyes only the camera or convex-lens eye, the mosaic eye, and the superposition eye. The last is a peculiar form of com-

pound eye where light can pass from one section to another, and where the image is formed by the codperation of various refracting bodies (324). The simplest and vaguest conceivable visual image would be that of a visual field whose different parts should differ in brightness. An eye capable of furnishing indications merely of the direction from which the greatest illumination comes might produce this kind of an image, which would of course 4 ! not allow the Fic. 11. — Diagrammatic representation of the compound

: 7 eye of a dragon-fly. C, cornea; K, crystalline cone; perception of P, pigment; R, nerve rods of retina; Fd, layer of ob jects, only las G, bee “ ae ti rae Rf, retinal fibres; . , crossing of fibres. er Claus. that of brightness distribution. The compound eye found in crustaceans and insects would seem to be adapted chiefly for the perception of light direction and of moving stimuli. It consists essentially of a number of simple eyes so crowded together as to produce a common faceted cornea, each facet belonging to an eye. These facets are lens shaped, and back of each lies a refractile crystalline cone. Behind these, in turn, are nervous structures, the rods or retinulz, each separated from its neighbors by a pigment sheath. Light rays passing through each corneal facet probably produce a single spot of light on the retinula, and the

total image may thus be a mosaic formed of these spots (Fig. 11). We have already seen that the orientations of certain animals to light seem to be produced through a tendency to take such a position that the two eyes shall be equally illuminated. If the two visual fields are combined in the case of such animals, as they are in our own binocular vision, under ordinary conditions the oriented position would give a field whose brightness is uniform throughout, while any other position would give greater brightness at one side of the field. If they are not combined, if there is no binocular vision, we cannot imagine what the resulting perception is. In the case of the starfish, we have an animal which seems to “see” a vertical white wall or dark wall that does not cast any actual shadow upon it; the starfish will direct its movements to or from such objects. Since the starfish has only eye-spots on the tips of its arms, with no arrangements for the formation of an image, and since the eye-spots are not arranged close enough together so that differences of illumination in different parts of a field could be represented by the different illumination of different eye-spots, we can explain the reaction to walls only, as Cowles (156) does, by supposing that those eyespots and portions of the body nearest, say, a white wall, are more strongly illuminated than those furthest away. The response would then be one to different intensities of stimulation on different parts of the body, and these differences would not be seen as composing a visual field.

That the direction from which the light comes influences ants in finding their way is the opinion of Lubbock (441), Turner (722 a), and Santschi (654). The first named found that ants which had learned the way back to an artificial nest were confused when two candles which had stood near the nest were moved to the opposite side. Turner made a similar observation, and Santschi suggests that the compound eye may perceive the direction of light by acting as a kind of sundial. He was able to make ants reverse their course when he altered the light direction. by the use of mirrors.

Various methods of solving the problem as to the nature and accuracy of an animal’s visual images have been used. One method consists in a study of the sense-organ itself, removed from the body. For example, Petrunkevitch (575) has thus investigated the sense of sight in spiders. These animals do not have the compound eye, but a number of simple eyes placed in groups. By a careful measurement of the possible minimal angles of vision in two spiders, Phidippus and Lycosa, the conclusion is reached that while a creeping insect about one square centimeter in size would be to the human eye so clearly visible at a distance of three metres that its species could be recognized, it would be only an indefinite moving speck to the eye of Phidippus and wholly beyond the range of vision of Lycosa,

Again, inferences are drawn as to the visual powers of animals from miscellaneous peculiarities of behavior. Thus Petrunkevitch (576) reports that a male spider of the species Dysdera crocata, in the courting stage, ‘‘watched” the movements of the end of a hatpin with which the observer was breaking clumps of earth, and when the movement ceased the spider approached the spot and scratched it with his front legs. The sight of a female spider digging had the same effect upon him, so evidently the visual

image which he received was hardly more definite than one of general size and movement. Bauer (28) reports of the mollusk Pecten, which has eyes of peculiar and complicated structure, that when a small quickly moving shadow is cast upon it, the tentacles are quickly withdrawn; large or slowly moving shadows have no effect, but a small, slowly moving shadow makes the animal stretch its tentacles and eyes towards the shadow. In this way, Bauer thinks, it is enabled to ascertain the nearness of its worst enemy, a starfish: apparently he supposes that the movement of the eyes towards the shadow gives an opportunity for visual perception of its form or characteristic movements. Wenrich (777) has recently obtained the following evidence of the formation of an image in Pecten. The bivalve normally responds only to a decrease in illumination, not to an increase. Jf a white card is moved across a black one, the card being not less than fifteen millimeters square and its distance not greater than thirty-five centimetres, Pecten responds by closing its shell, although the illumination is increased rather than diminished.

The chief lines of evidence, however, from which the nature of the visual image can be concluded are three: experiments on the visual perception of size, experiments on the visual perception of form, and experiments or observations on the recognition of visual landmarks in homing. Bohn’s observations on the mollusk Littorina show that its reactions are influenced by the size of the illuminated or darkened surface, as well as by the intensity of the light. When neither very wet nor very dry, Littorina will react to small objects in its neighborhood, whereas in an extreme

state of “hydratation” or desiccation it responds to the attraction or repulsion of the larger screens with fatal uniformity (80). Plateau attempted to test the responses of certain Diptera to the size of an opening admitting light, by placing them in a dark room, into which light entered from two sources. One was a single orifice large enough to let the insects out; the other was covered with a net whose meshes were too fine to allow them to pass. The amount of light from the two sources could be made equal. When this was done, the insects, which were positively phototropic, sought the two equally often; if the light from either was made more intense, they went to that one. Plateau concluded both that the flies could not see the netting and that the area of the light source did not affect them (592). On the other hand, Parker found that the mourning-cloak butterfly did discriminate areas, flying to the larger of two sources of equally intense light (537).

This method of testing the image-forming power of an animal’s eyes has been elaborated by L. J. Cole. He subjected animals with decided positive or negative phototropism to the influence of two lights made equally intense but differing in area, one coming through a piece of ground glass 41 cm. square, the other a mere point. Eyeless animals, the earthworm, for example, reacted equally often to each light. Animals whose eyes from their structure have been judged capable of perceiving merely the direction of light rays, such as the planarian Bipalium, confirmed the argument from structure by showing little more discrimination than the eyeless ones. On the other hand, animals with well-developed compound or camera eyes, for example certain insects and frogs, did distinguish between the lights, going, if positively phototropic, toward

the one of larger area; if negatively phototropic, away from it (132). Turtles showed a remarkable keenness of discrimination in the study made by Casteel (119), in which they were offered the choice of two compartments faced with cardboards carrying black lines on a white ground. Two turtles learned to discriminate between vertical lines eight millimeters in width and vertical lines two millimeters in width, and one gifted animal learned to distinguish, first, lines eight millimeters wide from lines one millimeter wide, then between a width of four millimeters and a width of one millimeter, then between four and two, and finally between three and two millimeters. Chicks proved equal to a discrimination between a standard circle six centimeters in diameter and one from one-fourth to one-sixth larger. The relative brightness of the circles was varied so that the chicks could not use this as a basis for their choices (102). White rats can discriminate circles thirty millimeters in diameter from circles fifty millimeters in diameter, and squares twelve centimeters a side from squares one centimeter a side (411). Discrimination of boxes differing in size but alike in form, placed in a row along a board, food having been put in one, was imperfectly learned by two Macacus monkeys; the errors leaned in the direction of taking the larger vessel (401). Raccoons were taught to distinguish perfectly between two cards, one 63 X 6} inches square and the other 43 X 43, shown successively. The animals were to climb on a box for food when the larger card was shown and to stay down when the smaller one appeared. As we shall see later, L. W. Cole, the experimenter, thinks the learning gave evidence not only of a spatial image, but of a memory image (134).

to the distance at which an object produces a reaction. Caterpillars, for example, are described as giving evidence of seeing a slender rod extended toward them at a distance of about a centimeter; large masses they reacted to at somewhat greater distance (597). It is highly doubtful whether this means that the simple eye of the caterpillar could give a perception of two objects as differing in size if they were equally distant. Myriapods, which make very little use of sight and do not perceive their prey until they touch it, give evidence of seeing an obstacle having a rather broad surface, the size of a visiting card, at a distance of about ro cm., if it is white and reflects much light, or if it is blue; put not if it is red.

The second method of studying visual images tests an animal’s power to discriminate forms. Bumblebees were thought by Forel to evince a capacity to distinguish a blue circle from a blue strip of paper when they had previously found honey on a blue circle, even though the two had been made to exchange places. They flew first to the place where the blue circle had been, but did not alight upon the strip. Wasps also, according to Forel, distinguished among a disk, a cross, and a band of white paper, going first to the form on which they had last found honey (231). Turner (726) reports the ability of the honeybee to distinguish, in the open air, among “artefacts” of various forms (disks, cornucopias, and boxes), covered with various patterns such as transverse and longitudinal stripes, mottled surfaces, and spotted surfaces; if the bee had found honey in an artefact of a certain pattern it would select that pattern from among other patterns or

_ criminate patterns like those of flowers, but fail with those very unlike flower patterns. This evidence, taken at its face value, indicates that the compound eye is able to furnish a fairly clear image, and not merely discriminations of light direction and movement. Among vertebrates, various species of birds were experimented on by the method of placing cards carrying simple designs over glasses covered with gray paper, food being found always under the same card. The English sparrow and the cowbird both learned to distinguish a card bearing three horizontal bars and one bearing a black diamond from each other and from plain gray cards. On the other hand, the sparrow, curiously enough, did not succeed in discriminating vessels of different form; the cowbird was not fully tested with these, but gave some evidence of learning (610, 611). Pigeons were only moderately successful in a similar test (647). Breed (1oz, 102) and Bingham (56) investigated the form discriminations of the chick, using the more accurate method of offering a choice between compartments illuminated through openings of different forms. One out of three of Breed’s chicks succeeded in discriminating between a circle and a square: Bingham’s chicks distinguished between a circle and a triangle when the apex of the triangle was on top, but the discrimination broke down when the triangle had its base uppermost. The most careful work that has been done on the discrimination of forms or patterns by animals, up to the date of publication of this book, is that of Johnson (386). His apparatus allowed the presentation of two illuminated fields whose intensity could be perfectly controlled, with black bands across them whose width could be varied at will. He proposed four problems: (1) the width of stripes

necessary to make a striped field just distinguishable from a uniform field; (2) the just noticeable difference between the width of stripes on two fields; (3) the just noticeable difference in the direction of the stripes on two fields; (4) the just noticeable difference in brightness between two fields, one of which has stripes of equal brightness, while the stripes on the other are of unequal brightness. The chick’s ability to distinguish a striped from a plain field proved to be about one-fourth that of a monkey or human being; when the problem of distinguishing between striped fields whose stripes were of different widths was presented, the monkey did ten times as well as the chick. For differences in the direction of stripes, the threshold of the chick was between twenty-five and thirty degrees; the monkey’s was between two and five degrees: moreover, the monkey learned the discrimination in twenty trials, while the chick required 585. It seems practically certain that the chick is not a fair representative of the bird family as regards the clearness of its vision for form and size; the eye of a hawk is a proverb for keenness, and the ability of birds to find their food by vision demonstrates the high development of their eyes in image-forming power. Among mammals, many dogs have been taught to distinguish printed letters on cards; Sir John Lubbock’s poodle “‘Van” is a familiar example. Van learned to pick out cards marked “Food,” “Bone,” “‘ Out,” “Water,” and the like, and to present each on its appropriate occasion. It took him ten days to begin to make the first step of distinguishing between a printed card and a plain one; in a month this was perfected and in twelve more days, when he wanted food or tea, he brought the right card one hundred and eleven times and the wrong one twice.

The second mistake consisted in bringing the word “Door” instead of “Food,” which was taken as indicating that he really was paying attention to the look of the words (444). Such observations, however, are very inconclusive when compared with modern experimental studies where all the sources of error, from smell, for example, are carefully controlled. In Johnson’s (386) study of the visual acuity of the dog, while two chickens and a monkey learned to distinguish a striped from a plain field in from three hundred to four hundred trials, dogs failed to learn in over a thousand trials, although the stripes were made nearly six times as wide. The dog could not distinguish between two visual fields unless they differed in intensity. Thus his visual images would seem to be far from clear. The eye of the dog, it may be noted, does not possess a fovea. Johnson thinks the dog’s vision is useful chiefly for the perception of moving objects. Szymanski (7o2) finds that when dogs and cats have been trained to go to a box in a certain corner to get food, and the box is moved, the dogs show their lack of dependence on vision by displaying little tendency, as compared with the cats, to use this sense in finding the new situation of the box. Orbeli, however, obtained evidence by Pawlow’s method that dogs could appreciate form and size differences (532 a).

The dancing mouse could not learn to distinguish two equal illuminated areas of different forms (820). Raccoons learned to discriminate a round card from a square one (134). Thorndike taught the two Cebus monkeys under his observation to come down to the bottom of the cage for food when a card bearing the word “Yes” printed on it was exposed, and to stay up when one bearing the letter ““N”? was shown. The conditions seem to have been complicated, however, by the fact that the two cards were not placed in quite the same position. Further

tests with cards carrying various designs showed varying degrees of capacity to distinguish them on the part of the monkeys (708). Kinnaman got negative results with his two Macacus monkeys in attempting to train them to distinguish cards such as those used in the later experiments of Porter on birds. His monkeys, however, proved able to distinguish vessels of different forms, “a widemouthed bottle, a small cylindrical glass, an elliptical tin box, a triangular paper box, a rectangular paper box, and a tall cylindrical can.” These vessels differed in size as well as in form (401). Johnson’s far more accurate experiments with the striped fields give the monkey a visual acuity about equal to that of man.

The question has been raised as to just what is meant by the term ‘“‘form” in connection with the visual perceptions of an animal. When Bingham (56, 57) found that a chick failed to recognize a triangle whose base instead of its apex was uppermost, he suggested that the chick’s previous discrimination of the triangle from a circle was not a discrimination of form in the true sense of the word, but based “on the unequal stimulation of different parts of the retina.”” Hunter (352) thinks that the animal in such a case is really discriminating pattern rather than form, and by pattern he means the whole design presented by the lighted forms and their surroundings. That is, a square lighted area inside a round tunnel would present to the animal a different pattern from a square lighted area inside a square tunnel; an animal might fail to recognize that the forms of the squares were identical when they were presented as parts of such different patterns. The writer of this book suggested in a review of Bingham’s work ! that his chicks, in failing to recognize that a triangle

with apex down is the same form as a triangle with apex up, were demonstrating not their deficiency in form vision, but their lack of an abstract idea of triangularity. It may well be that such a perception of form in the abstract, such an ability to analyze forms out of patterns, depends upon the association between visual impressions and movements like the hand movements of a human being; few lower animals, of course, have the same kind of motor experience of objects that man possesses.

Special evidence of the comparative development of the visual image in different genera of ants is suggested by Wasmann to be furnished by the facts of mimicry. Certain insects belonging to orders other than the Hymenoptera inhabit ants’ nests, and have in many cases become more or less modified to resemble their hosts. 'Wasmann thinks that these resemblances, which have been established on account of their protective value, are in insects living among ants of well-developed visual powers, such as would deceive especially the sense of sight, while in the “guests” of ants whose vision is poor, the mimicry is adapted to produce tactile illusions (762).

The ability to find their way back to their dwelling place, or to any other locality that has a vital significance for them, is a power widely distributed among the most various forms of animals. We have considered, in the chapter on the Chemical Sense, the part which smell plays in this process, and on page 100 we noted the fact that the perception of light direction is not wholly without influence in some cases. The common human method of path-finding is by the recognition of visual landmarks: when we

set out from a familiar region into a strange region, we fix our attention on the appearance of the surroundings at critical points and turnings, and on the homeward journey guide ourselves by identifying these points through vision. Where it can be shown that animals are influenced in their homing journeys by the appearance -of the surroundings, we have evidence that their vision must involve some perception of the form and detail of objects. The fiddler crab “remembers” the location of its nest, but just what the memory depends upon is not clear. On one occasion the observer, Pearse (568), covered the nest with his foot; the female crab to which it belonged waited fifteen minutes until he moved his foot, and then dashed for the nest and tried to reopenit. Lubbock’s (441) demonstration that ants do not use visual landmarks on frequented roads will be recalled (see page 97).

In the case of bees, on the other hand, there is a good deal of evidence in favor of the use of visual landmarks in homing. It is true that Bethe (51) was unable to note any disturbance in the flight of bees back to the hive when he altered the appearance of the hive, or when a large tree that stood near the hive was cut down. But in this case the bees had thoroughly learned the location of the hive and had probably ceased to need landmarks in its immediate ~ environs. Lubbock found that bees from a hive near the seashore, when taken out on the water and liberated, were unable to find their way home, although the distance was less than their usual range of flight on land; and he ascribes their failure to the lack of visual landmarks to guide them (441). Bethe, who thinks bees are guided home neither by vision nor by smell, but by an unknown force to which they respond reflexly, also liberated some bees at sea about 1700-2000 metres from their hive, which was

near the foot of Vesuvius and beside some very tall and conspicuous trees. The bees failed to return, and Bethe thinks, if they were guided by vision, the mountain and the trees should have aided them to do so (53). It may well be, of course, that bees cannot see objects at such a distance. Besides his observation that changing the appearance of a hive did not disturb the bees in their homing flight, Bethe urges against the visual memory hypothesis an observation on a hive which had on one side of it a garden, and on the other side a town, which he thinks the bees never visited, as food was to be had in abundance in the garden. Yet when liberated in the town they flew back to the hive with an accuracy certainly not born of their acquaintance with the locality (51). Von Buttel-Reepen, however, doubts whether the bees really never visited the town. Bethe’s most striking illustration of his unknown force, however, is derived from his ‘‘box-experiments.”’ If a number of bees are carried in a box some distance from the hive, on being liberated they fly straight up in the air. Some of them will return to the hive, but if the distance is great enough, many will drop back upon the box. Now if the box has moved only a few centimeters away during the flight of the bees, they will drop back to the precise spot where it was, and take no notice of its new location. If they were guided by vision, Bethe urges, they could easily see the box (51, 53). This, says von Buttel-Reepen, is arguing that their visual memory must be like ours if it exists at all; it may be a memory, not of the appearance of the box, but of its locality. He himself, repeating Bethe’s experiments, observed the bees on dropping back after their upward flight, hunting not at the place where the box had been, but at a height which was about that of their home hive entrance. He thinks that an important