A WORKING MODEL FOR DEMONSTRATING THE MOSAIC THEORY OF THE COMPOUND EYE

Size: px

Start display at page:

Download "A WORKING MODEL FOR DEMONSTRATING THE MOSAIC THEORY OF THE COMPOUND EYE"

Marian Griffith
5 years ago
Views:

1 A WORKING MODEL FOR DEMONSTRATING THE MOSAIC THEORY OF THE COMPOUND EYE BY EDGAR ALTENBURG, The Rice Institute, Houston, Texas. (With six Text-figures.) (Received 27th February 1926.) THE confusion in regard to the mechanism of image formation by the compound eye of insects and of other arthropods has been due in part to the fact that no simple working model has been constructed for demonstrating the way in which this type of eye works. I have found it possible to construct such a model by means of numerous tubes of small diameter. Straws are very suitable for this purpose. In the construction of the model, the straws are packed in a box with their long axes parallel to each other (see Fig. 1). The straws should first be blackened so as to prevent Stra ws m box. frosted glass at end of straws Fig. 1. light diffusing through their sides. The box is open at one end (front) to allow light to enter the straws, and the opposite end (back), is covered with a piece of frosted glass or other translucent material which is used as a screen and on which the image is thrown. The "back" ends of the straws touch the frosted glass. In other words, the model is so constructed that light from an object enters the straws at one end and passes out through the other end, which rests against the frosted glass. I have found that a model which contains about 10,000 straws is of venient size and gives perfectly good results. The object to be projected is placed directly in front of the above-described model and not more than a foot from the ends of the straws. The object is well

2 Model for Demonstrating Mosaic Theory of Compound Eye 39 just as for opaque projection with a lens. One or more lamps of at 150 watts each suffice for this purpose (attached to an ordinary desk lampstand equipped with a shade). The lamp is placed so as to illuminate the part of the object facing the straws, and to do this it is necessary to place the lamp closer to the model than is the object, and at the same time in such a position that it will not be in the way (the lamp must not be in the direct path between the object and the model, but a little to one side). The lamp-shade is used to keep the light from striking the straws directly, so that only light reflected from the illuminated object may reach the model. The observer is in a darkened room and light is prevented from escaping from the lamp into the room by a dark cloth suitably enclosing the lamp, object and front portion of the model. An arrangement that is permanent can easily be made by having, in place of the cloth, a box the inside of which is blackened, and in which the lamps are fixed in about the same position, relative to the model and object, as above described. In case it is inconvenient to darken the room, the observer can enclose his head and the back end of the model in a dark cloth and thus prevent outside light from striking the frosted glass that acts as the screen. It should be emphasised at this point that the images that appear on the screen of the model are really images and not merely the objects "shining through" a translucent screen. In order to show this point, a handful of straws are removed from some convenient part of the model (say a part near the upper righthand corner) and replaced by a piece of paper wrapped in the form of a cylinder. No image formation occurs in this part of the model, and this part, therefore, serves as a "control." Objects that are projected on the screen of the "straw" model above described show a fair amount of detail. It is, for example, possible to see the face of a person in sufficient detail to recognise him on the screen. The efficiency of the model depends in part upon the intensity of illumination and on the exclusion of outside light from the screen, as is the case with apparatus ordinarily used for opaque projection. It depends also upon the relative length and diameter of the straws. The longer and narrower they are, the greater is the detail shown by the image. The straws in the model obviously correspond to the units that constitute the compound eye the ommatidia. The number of straws used in the model (10,000) is about the same as the number of ommatidia in some compound eyes, but there may be several times this number in the eye. The formation of an image by the compound eye is dependent primarily, not upon lenses, but upon the fact that the ommatidia are comparatively long and narrow and have a regular arrangement with regard to each other, substantially as in the model. Just how this simple mechanism is sufficient for image formation can be made clear by means of the following analogy. Suppose that the windows building had very deep casements, like those of a castle, on account of the of its walls. Suppose, further, that these windows were rather small, numerous and close together, and that a person were sitting behind each window. Then some object in front of the castle, say«tree, could be seen by only a limited

40 EDGAR ALTENBURG number of the persons in question, and furthermore, each person would see a limited part of the tree. Any one person could see only what was directly a of him.

3 40 EDGAR ALTENBURG number of the persons in question, and furthermore, each person would see a limited part of the tree. Any one person could see only what was directly a of him. He could not see anything above or below or to the right or left, because the sides of the casement would obstruct his view. In other words, he could see only an outside object, or that part of it, which happened to be in line with the "long axis" of his casement. Suppose, now, that the persons who could see some part of the tree should have white clothes, and the others black, making them readily distinguishable, and that a magician should convert the castle into glass, making all persons easily visible to an outside observer. The combined assemblage of persons in white would then present a figure which would roughly conform in outline to that of the tree; in other words, they would constitute a rough image of the tree. The above-described "image" is due obviously to the fact that any given person in white has the same position relative to the others in white as has the part that he sees of the tree to the other parts of the tree. This correspondence is made possible by the interposition of the casements between the tree and the persons that view it. Without the casements, any one part of the tree would be visible to all persons, and there would be no "image" formed by the aggregate of persons who saw the tree. But because of the casements, there is a spatial correspondence between a given person and the part that he sees of the tree. Hence, an "image" is formed. The mechanism of image formation by the compound eye is precisely analogous to that just described. The casements in the castle analogy correspond to the ommatidia, and the inside observers correspond to the cells of the retina. In the model, the corresponding parts would be the straws and the frosted glass or other screen at the "hind" end of the straws. Light coming from any point on an object strikes the entire surface of a compound eye that is viewing it. But a given ommatidium sees, so to speak, from its "hind" end only a limited portion of an object, really almost only a point of the object. This point is one that is directly in line, with its own long axis. Or, to put the matter in another way, only such rays can enter an ommatidium as are in line with its long axis. Therefore, light coming from a given point in an object can enter only a particular ommatidium, namely, the one that has its long axis parallel to the entering light. All other light will strike the sides of the ommatidium and fail to reach its hind end, where the retina is situated. Consequently, an image will be formed on the retina. Contrary to what has sometimes been maintained, no one ommatidium will contain an entire image. It will have on its retina merely a spot of light. The image itself is built up by the aggregation of spots in the various ommatidia that receive light from the object. In other words, the image is a sort of patchwork, or mosaic. The way in which the model of the compound eye works can, possibly, be easily understood by means of the following consideration. Suppose a person's ha should be illuminated, by means of a beam of light, in a room that was other dark, and that he should hold his hand with its illuminated surface facing a screen. Light would be reflected from his hand in au directions and would be shed on the

4 Model for Demonstrating Mosaic Theory of Compound Eye 41 entire screen without producing any effect other than making the screen in general ^Hitly bright. But if his hand had the capacity to select the rays that were to be reflected, and if it could reflect accordingly only such rays as were parallel to each other and at right angles to the plane of the screen (see Fig. 2), then the screen would be dark except for the part that was directly "opposite" the hand. On this part, there would be a lighted area having the form of the hand. In brief, an image of the hand would be produced on the screen, due to the selection of Fig. 2. certain rays. The model of the compound eye has precisely such power of selecting the rays that are to reach the screen, and, so, of causing image formation. If the above account is correct, it will be seen that lenses take no essential part f in the mechanism of image formation in the compound eye.. Nevertheless, there lens at the front end of each ommatidium. What is its meaning? The answer this question involves a detail connected with the structure of the eye and the way it works. The sides of the ommatidia converge towards their attached ends.

5 42 EDGAR ALTENBURG Such convergence makes possible a larger field of vision than if the sides were parallel. In the working model, the sides of the straws are parallel and the ^P of the image is the same as that of the object. Thus, the two ends of, say, a pencil, are as far apart in the image as in the object itself (see Fig. 3 a). In order to "see" an entire pencil, the eye would have to be at least as large as the pencil, provided the sides of the ommatidia were parallel. But if the ommatidia converge towards their inner ends, then light from the extreme ends of an object much larger than the eye can enter them, and the eye will form an image of reduced size (see Fig. 3 b). The field of vision of the eye is accordingly increased, and an insect is enabled to see an appreciable portion of the outside world at a single "glance," so to speak. But to get such a result each ommatidium must be broader at its outer end than object straws Fig. 3 a. image L object ommatidia Fig image at its attached (retinal) end, and it is in this connection that a lens is of use. For, without a lens, some of the parallel rays that entered the front of an ommatidium would strike its converging sides and would fail to reach the back end, but by means of a lens the rays may converge and all reach the retina (see Fig. 46). In this way, the lenses increase the brightness of the image. No one lens, however, forms the complete image, and in fact the image could be formed equally well without the lenses, excepting that it would not be as bright. The lenses, in other words, take primary part in image formation. The image is a mosaic, made by the aggregate ommatidia that receive light from the object. It is true that an image could be obtained by means of the type of lens possessed

6 Model for Demonstrating Mosaic Theory of Compound Eye 43 ommatidium, but the focal distance would, of course, have to be properly d. It is probably in this way that micro-photographs can be obtained showing complete image formation by each lens and hence a myriad of images by the compound eye as a whole. It is obvious, however, that images could not ordinarily be produced by the individual lenses, since there is no mechanism, such as that of accommodation in the vertebrate eye, for adjusting the focal distances of the lenses. An eye that depends upon a lens must have some such mechanism; otherwise, it would be worthless for the same reason that a camera would without some focussing device. The detail in an image formed by a compound eye is dependent upon the number of ommatidia in an eye of given size. A single ommatidium shows no detail; that is, its sensory cells, when stimulated, cause the animal to see just a small spot of light, regardless of the shape of the object from which the light comes. Thus, if an insect viewed, say, the letter "S," held at a given distance, and if an om- Fig. 4 a. Fig. 4 b. matidium were so large as to include the entire " S " within its field of vision, then the insect would receive merely the impression of a spot or blotch of light. But, if this single ommatidium were subdivided so as to give a larger number of ommatidia, each of smaller size, then the curves of the "S" would be seen, and the smoothness of the curves and their general detail would depend upon the extent of the subdivision (see Fig. 5). The same principle applies here as to mosaics generally. The smallness of the "pieces" determines the extent to which detail can be shown by the compound eye, just as is true of the artist who is making a mosaic of inlaid stone. The detail in an image formed by a compound eye also varies with the disce that the object is from the eye. Consider a person looking through the small of a tube shaped like an ommatidium. He would see just a small part of a large object, say a building, that was close to him, but if the building were at a sufficiently great distance, he could see the whole of it through the tube. From

7 44 EDGAR ALTENBURG this consideration, it is apparent that the further an object is removed fror compound eye, the greater is the part of the object that is included within the f of vision of a given ommatidium. Therefore, the further an object is removed from the eye, the fewer are the ommatidia that" cover " the object as a whole, or any part of it, and accordingly the less the detail that can be seen in it by an insect or other animal with a compound eye. The vertebrate eye labours under no such difficulty and is probably much more efficient in showing the detail of objects at a distance than is the insect eye. i 4 16 (i.e. 3 complete plus 3 thirds) Fig. 5 a. Number of ommatidia covering a given field (including the letter "S") (No image) (No image) (Poor image) (Image with fair amount of detail) Fig. s b. Appearance of the illuminated area of the retina. There is another reason, in addition to the one just mentioned, why the vertebrate eye is more efficient than the insect eye. All the light that strikes the lens of a vertebrate eye from a point can be brought to a point on the retina (see Fig. 6a). But, in the case of the insect eye, only a small fraction of the Fig. 6 a. Vertebrate eye.

8 Model for Demonstrating Mosaic Theory of Compound Eye 45 that strikes the eye, and that comes from a point, can reach the retina; namely, the rays that are parallel to the long axis of some particular ommatidium (see Fig. 6 b). All other rays that strike the surface of the eye fail to reach the Fig. 6 b. Compound eye. retina. Accordingly, an image formed by a compound eye is not so bright as one formed by a vertebrate eye (making due allowance for difference in the size of the eyes in question and amount of light received by them). The insect eye probably gets round this difficulty to some extent by having more highly sensitive retinal cells. However, even then, the insect probably cannot see as well in dim illumination as can a vertebrate, a conjecture which, if true, might account in part for the lowered activity of insect life on cloudy days. SUMMARY. A working model of the compound eye can be constructed by the parallel arrangement of numerous tubes (e.g. straws) packed in a container. Lenses are not a necessary part of the model. The defining power of the model depends upon the number of tubes per unit area the greater their number, the more detailed is the image formed by the model. The manner in which the model works would indicate that the image formed by the compound eye is a mosaic, or patchwork; and that lenses take no primary part in image formation by a compound eye, but are merely accessory structures which bring about an increase in the amount of light that strikes the retina.

EXPERIMENT 4 INVESTIGATIONS WITH MIRRORS AND LENSES 4.2 AIM 4.1 INTRODUCTION

EXPERIMENT 4 INVESTIGATIONS WITH MIRRORS AND LENSES Structure 4.1 Introduction 4.2 Aim 4.3 What is Parallax? 4.4 Locating Images 4.5 Investigations with Real Images Focal Length of a Concave Mirror Focal