The Functions of the Ocelli of Calliphora (Diptera) and Locusta (Orthoptera)

Investigations into the visual capacities of the ocelli have been made by many previous workers, and various functions have been attributed to these organs. In the present work Calliphora erythrocephala Meigen and Locusta migratoria L. were employed to investigate the functions of the ocelli in form perception and as adjuncts to phototaxis. The optical powers of the ocellus were considered, and behaviour experiments were performed under critically controlled experimental conditions.

The culture of Calliphora was developed from a single batch of eggs in order as far as possible to obtain standardized insects. The locusts were offspring of gregarious parents and the methods of rearing were also standardized.

The biconvex lens of Calliphora is supported by a thin layer of corneagen cells containing minute pale brown pigment granules. The sensory layer lies immediately below the corneagen cells and is composed of elongate retinal cells. Rhabdoms were not identified with certainty, and it is therefore assumed that the retinal cells are sensitive to light along their entire length. The volume occupied by the sensory layer, 0·06 mm. deep, constitutes the retinal space (Fig. 1).

The lens of Locusta is also biconvex with a slight depression at the centre of the inner surface. The corneagen cells are extremely elongate and collectively they form a hemispherical body below the lens. Ovoid dark brown pigment granules are confined to the peripheral corneagen cells which curve inwards radially below the lens to form a pigment iris. The sensory cells are pear-shaped and aggregated in groups of four forming a distinct cup-shaped layer. Rhabdoms are again apparently absent, and the length of the sensory cells constituting the retinal space does not exceed 0·05 mm.

The possibility of form perception by the ocellus depends on the ability of the lens to form an image, the location of the image plane relative to the retinal space and the ability of the sensory layer to receive and utilize the image so formed.

If an ocellar lens of either the blowfly or the adult locust is removed and a drop of water placed on the inner surface of the lens is allowed to evaporate almost to dryness, well-defined inverted images may be seen through the microscope. Such images may be obtained of objects placed between 2 cm. and 5 m. from the lens of Calliphora, and less well-defined images obtained of objects placed as far as 1 m. from the lens of Locusta. If, however, the inner surface of the lens is perfectly dry it is impossible to obtain these images because the roughness of the inner surface of the lens scatters the light in all directions. It is reasonable to assume that the inner surface of the lens in situ is in contact with body fluid and that light rays passing through the lens are converged to a focus as in the isolated lens.

Parry (1947) showed that the principal focus of the ocellar lens of Locusta occurs 5 times as far behind the lens as the retina, and Homann (1924), working on a number of insects, showed that in no case were the focal and retinal planes coincident. Both these workers conclude that form perception by the ocellus is impossible.

The radius of curvature of the outer surface of the lateral and frontal ocellar lenses of Calliphora was measured by a modification of Homann’s method (1924). From six-determinations a mean value of 0·07 mm. (± 0·004) was obtained. Using the same method for the adult locust, nineteen determinations provided the mean value of 0·47 mm. (± 0·024). The radius of curvature for Locusta was also obtained graphically from medianly cut hand sections of lenses placed in a micro-projector. By this second method, twenty-five determinations provided the mean value of 0·44 mm. (±0·016). The mean obtained by these two methods in Locusta was therefore 0·46 mm. The refractive index of the ocellar lens was determined by Dethier’s method (1942) and the values of 1·55 and 1·48 were obtained for Calliphora and Locusta respectively. The lenses were observed to be of uniform refractive index except at the periphery where slightly higher values were obtained. Assuming that light refraction does not occur at the inner surface of the lens in contact with the corneagen layer the physical constants determined above provide focal length values of 0·12 mm. for Calliphora and 0·96 mm. for Locusta.

The distance of the outer limit of the retinal space from the outer surface of the lens of Calliphora, 0·04 mm., was ascertained by focusing the microscope first on to the outer surface of the lens, and secondly, focusing through the lens to the distal ends of the retinal cells. The measurements obtained from a vernier scale were then corrected for the difference in the refractive indices of air and the lens. Parry (1947), using this method for Locusta, apparently without the refractive index correction, said that ‘the distance between the front of the lens and the retina was never found to reach the value of 0·09 mm.’ In the present work, the mean thickness of the lateral and frontal ocellar lenses of the adult locust, ascertained from forty-six medianly cut sections, was found to be 0·18 mm. (+ 0·008). From sixteen determinations of the distance of the distal ends of the sensory cells from the inner surface of the lens, measured along the central axis, a mean of 0·27 mm. (± 0·003) was obtained. The mean distance of the outer limit of the retinal space from the front of the lens of the locust was therefore 0·45 mm. The physical constants determined above provide means of 0·10 and 0·50 mm. for the distance of the inner limit of the retinal space from the outer surface of the ocellar lenses of Calliphora and Locusta respectively.

These investigations have shown that the plane of the principal focus of the ocelli of Calliphora and Locusta is located deeper in the ocellus than the limits of the retinal space. Light rays impinging upon the ocellar lens from objects close to the insect will converge to a focus even further outside the retinal space than parallel rays entering the lens from objects at infinity. Therefore, as the positions of the image space and retinal space are not coincident, the possibility of form perception of any degree of accuracy by the ocelli of these species must be excluded, and the ability of the sensory cells to receive and utilize the images formed need not be considered. It may be noted, however, that in both species, the centre of curvature of the outer surface of the lens occurs within the retinal space, but the implications of this are not clear.

The visual fields of the ocelli were investigated by observing the reflexion on the ocellar lens of a moving light source. When this reflexion was not obtained the source was known to be outside the visual field. This method provides the maximum possible field limited by the position of the ocelli on the head, and assumes that the sensory cells are stimulated by light incident anywhere on the lens. The presence of a cup-shaped sensory layer in Locusta and of retinal cells in Calliphora of which the distal ends are very close to, and have the same curvature as the hind surface of the lens, suggests that even oblique rays will be received. The wide visual fields of the ocelli (Figs. 2, 3) afforded by their positions on the head, and the structural arrangement of these organs support the suggestions of previous workers that the ocellus is essentially a light-gathering and not an image-forming organ.

Behaviour experiments with Calliphora were carried out at a constant temperature of 21°C. and relative humidity of 70%, and experiments with Locusta at a constant temperature of 27·5°C. and relative humidity of 75 %. The room used was without windows and the four walls and ceiling were papered with a non-reflecting black paper. The bench was also covered with this paper, and as a further precaution the edges of the bench were boarded to a height of 6 in. with cardboard covered with black velvet. No electric light sources were permitted except those essential for the experiments, and in all cases these were screened by $212$ in. thick water filters to obviate temperature effects. To remove reflexions, the glass filters were covered with black paper and all other apparatus was also blackened. The apertures of the light sources in experiments with Calliphora were 3·8 cm. with the centre at 5 cm. above the bench surface, and those in experiments with Locusta were 10 cm., with the centre 6·5 cm. above the bench surface. When recording tracks of walking insects, gloves and a coat of black material were worn to reduce reflexions from the hands and clothes.

The part played by the ocelli in the phototactic reactions of Calliphora and Locusta was investigated by a series of experiments in which the effect of the occlusion of either the ocelli or compound eyes on the directional response was obtained. The eyes were occluded with an almost odourless and extremely quick drying lacquer^* which was applied to unetherized insects under a binocular microscope. In Calliphora there was no significant difference in the intensity of reaction between the sexes and as in the males the ocelli and compound eyes are very close together, only females were used in experiments where ocellar occlusion was required. A period of 30 min. was allowed after occlusion before starting an experiment, during which time the extra cleaning movements induced by handling and occlusion were effected. At the end of an experiment all insects were examined under a binocular dissecting microscope to check the complete occlusion of the sense organs. Results for insects with incomplete occlusion were discarded. In experiments made with walking flies the wing apices were removed and again a 30 min. period was allowed for the completion of cleaning activity. Locust hoppers were used to investigate walking reactions because adults showed a strong tendency to fly even with the wings severely clipped.

It was first shown that Calliphora does not exhibit telotaxis. Two light sources of equal illumination and screened by water filters were placed 35 cm. apart, and paths made by intact flies released about 1 m. from the lights were seen to approximate to a line drawn symmetrically between the sources. Twenty intact insects were then released at the same distance from a single light source and in every case movement was made towards the source. On blackening one compound eye, repeated circus movements were made and from these observations it was concluded that Calliphora exhibits positive photo-tropotaxis.

To investigate the phototactic reaction of Calliphora, twelve normal flies were introduced into a 15 cm. Perspex cube on which a line was drawn dividing the cube into two equal halves. Random distribution of the flies was allowed and with incident light from above, the number of flies in the upper half was noted at intervals of 30 sec. The ocelli were occluded and the experiment was repeated. The procedure was again repeated subjecting specimens with ocelli functional and occluded to incident light from below and also from one side. The intensity of reaction was calculated as a percentage of the flies present in that half of the cube nearer to the light source. The results shown in Table 1 indicate that when the ocelli are occluded there is a small decrease in the intensity of reaction for all directions of incident light. The difference between the intensities of reaction to incident light from above and below is attributed to the association of the positive phototactic response with a negative geotactic response. To investigate more critically the effect of occluding the ocelli on the phototactic response, the following experiment was conducted in which the length of path made by walking flies was recorded.

An insect which exhibited a completely efficient phototactic response would move in a straight line towards a light source. By measuring with an opisometer the actual distance travelled along a curved path (Fig. 4), and comparing this with the length of the straight line, a measure of the ‘efficiency’ of response is obtained. Paths of males and females with functional ocelli were measured, and the results shown in Table 2 indicate that there was no significant difference between the efficiency of movement of the sexes. Females only were used when path measurements were made of flies with occluded ocelli. Using a bright light source occlusion of the ocelli produced a significant decrease in the efficiency of response. Paths made by females with functional and occluded ocelli are represented to scale in Fig. 4. The angular paths made with occluded ocelli resulted from the two modes of progression shown in Fig. 5. When a less luminous source was used no significant difference between the efficiency of response of normal and anocellate insects was obtained. With this source the paths and modes of progression made with functional and occluded ocelli resembled those made by flies with occluded ocelli when subjected to bright illumination.

The next experiment was designed to investigate the effect of ocellar occlusion on the phototactic response to a change in direction of incident light. Two light sources screened by water filters were placed at the positions B and D as shown in Fig. 6. Flies released at A were allowed to walk towards the illuminated source B, but on reaching C the source B was extinguished and source D simultaneously illuminated. The insects continued along curved paths reorientated to the second source. The points E on the curves farthest from the line CD were recorded and their distance from the line CD was measured. These operations were conducted on females, first with functional ocelli, and secondly with ocelli occluded. The light source B orientates the fly at C so that on the illumination of source D only one compound eye is subject to stimulation. Measurements were made only on flies which travelled without stopping from the starting position to source D. Occasionally, on the simultaneous changing of the sources a fly at C momentarily became stationary, turned axially through an angle of 90 degrees and then proceeded towards source D. Records of flies effecting such movements were not included.

Table 3 shows that when Calliphora is subjected to a change in the direction of incident light there is a significant decrease in the efficiency of reorientation by flies with occluded ocelli. The average paths effected by insects with ocelli functional and occluded have been represented in Fig. 6.

Finally, the role played by the ocelli in the phototactic response was investigated by occluding both compound eyes and subjecting flies with only the ocelli functional to a single light source. The erratic paths made by walking flies released facing in varied directions and at varied distances from the light source are shown in Fig. 7. From this experiment it may be concluded that phototactic orientation by means of the ocelli alone is impossible.

These experiments have shown that under laboratory conditions the ocelli of Calliphora are incapable of mediating a directional response and that with the ocelli occluded there is a decrease in the efficiency of phototactic movement and reorientation to a change in direction of incident light. These results are in agreement with observations of previous workers. That insects with the compound eyes blackened behave as if blind has been shown by Brandt (1937), with Drosophila by Bozler (1926), with bees by Müller (1931) and with ants by Homann (1924). Wellington (1953), however, showed that with low light intensity in the laboratory, Sarcophaga with the ocelli alone functional exhibited undirectional movement, but outdoors under conditions of high light intensity such insects produced movements as directed as any made by flies with the compound eyes functioning. Further, he showed that the ocelli of adult Sarcophagid flies are sensitive to changes in polarized light from the sky. Von Buddenbrock (1937) and Kalmus (1945) suggested that the ocelli aid in orientation during flight, and that they assist in the phototactic orientation of bees and ants has been suggested by Müller (1931). At low light intensities Bozler (1926) established that ocellar occlusion resulted in a slight reduction of the ‘normal phototactic orientation potential’.

In bright illumination, the decrease in the efficiency of the directional response of Calliphora with occluded ocelli suggests that there is a correlation of function between the ocelli and compound eyes. In dim illumination there was no marked decrease in the efficiency of response which suggests that in Calliphora there is a threshold of light intensity below which the ocelli are not sensitive. In bright illumination, the progression of normal flies was such that the longitudinal axis of the insect was alined with the steepest hght gradient and corresponding parts of both compound eyes were equally stimulated. With the ocelli occluded, two types of progression were exhibited during which the axis of the insect was not alined with the steepest light gradient, but at an angle to it, and there was a failure for the corresponding parts of the two eyes to be stimulated. The curved path shown in Fig. 5 (a) appears to be the result of a turning reflex evoked by stimulation of the antero-lateral region of one compound eye. In Fig. 5(b) it appears that as the result of too great a turn the antero-lateral region of the other eye becomes stimulated and this eye continues to mediate turning reflexes. Continued progression of this type results in the formation of a zigzag path, and movement towards the light source may result from either one or both these modes of progression. When a fly is subjected to a change in direction of incident light and the lateral region of one compound eye is stimulated, a turning movement is mediated by the lateral ommatidia followed by a direct path to the source resulting from equal stimulation of the anterior regions of both compound eyes. When the ocelli are occluded, however, orientation appears to be the result only of a reflex turning movement mediated by the lateral ommatidia, producing a less efficient path to the source.

First-instar nymphs were subjected to a single light source and paths made by walking hoppers are represented in Fig. 8. There appears to be no uniformity in the direction of walking, and from these observations it was impossible to define the normal phototactic response of Locusta. As similar results were obtained with all nymphal stages the following experiment was designed.

A grid of ninety-six 2 in. squares was drawn on a piece of black paper (Fig. 9), and a light source and filter placed at one end provided the light intensities as shown. One hundred and fifty hoppers (seventy-five first-instar and seventy-five second-instar nymphs) were released one at a time at the opposite end of the grid directly facing the source, and paths were recorded by noting the squares through which the hoppers walked. The table of frequency obtained has been represented as a density diagram which shows that there is a tendency for the hoppers to walk either towards the source or at right angles to it. The latter type of orientation is a ‘light compass reaction’ as was observed by Kennedy (1945).

The next experiment was conducted to explain the existence of two types of light reaction in the same insect. Three fifth-stage nymphs were subjected to bright illumination at a temperature of 36–40 ° C. for a period of half an hour. Repeated trials were made with each specimen using the light source as described in the above experiment, and the paths shown in Fig. 10 indicate that in all cases hoppers progressed towards the source. Detailed examination, however, shows that in many cases the paths are at a constant acute angle with the direction of incident light, and in repeated trials with the same specimen this angle is approximately preserved. The same three specimens were then placed in complete darkness for half an hour at the same temperature, and repeated trials were again made. The paths shown in Fig. 11 indicate that under these conditions hoppers exhibit a light compass reaction. Locusts released facing the source progress slightly towards it, turn, and continue at right angles to it. Those released facing at right angles to the light continue to walk approximately in this direction.

It was concluded from the results of this experiment that the direction of path made by a walking hopper depends on its state of photic adaptation. Dark-adapted specimens progressed approximately transversely to a light source with fixation of the stimulus by a set of ommatidia in the lateral region of one eye, and light-adapted hoppers progressed towards the illuminating source with fixation of the stimulus in the anterior regions of both eyes.

The next experiment was designed to ascertain the effect of occluding the ocelli on the directional response. Two sources (with water filters) emitting light of the same intensity were placed at opposite sides of a circle. The circumference was divided into seventy-two equal portions each corresponding to an angle of five degrees (see Fig. 12). Four fifth-stage nymphs were dark-adapted for a period of 1 hr. and each was subjected to the illumination of the two light sources in the following manner. With one source illuminated and the other extinguished, a hopper was placed at the centre of the circle, first facing the source (i.e. at o degrees), then at right angles to the source (i.e. at 90 degrees), thirdly facing away from the source (i.e. 180 degrees) and lastly again at right angles (i.e. 270 degrees). The angle of the point at which the insect crossed the circumference of the circle was recorded, and this procedure was repeated 5 times with one source illuminated and then 5 times with the second source illuminated. Four presentation angles were used because of the difference in response in the two conditions of adaptation. Two light sources were used to eliminate any undesirable lighting effects inherent in the apparatus. The same four hoppers were light-adapted for 1 hr. between four 40 W. bulbs situated so as to ensure equal adaptation of both eyes, and they were then subjected to the procedure described above for the dark-adapted state. Following this, the ocelli were occluded, the insects were again dark-adapted for an hour and further observations were recorded. Finally, with the ocelli occluded, the specimens were again light-adapted and a final series of observations was made.

The results of this experiment are summarized in Tables 4 and 5, and analyses of variance were made to ascertain whether photic adaptation or ocellar occlusion influences (a) the ‘positiveness of phototaxis’ and (b) ‘deviation from parallelism’. The values of ‘positiveness of phototaxis’ were obtained by ascribing a value of i to each positive observation and a value of to each observation of 90 or 270 degrees. The mean angle of ‘deviation from parallelism’ was calculated after converting the observed angles to their equivalents lying between the limits of o and 90 degrees. Paths towards the source at angles of 30 or 330 degrees, or away from the source at 150 or 210 degrees are thus both equivalent to 30 degrees when considering ‘deviation from parallelism’.

The analyses of variance showed that there was a significant effect of adaptation (P<0·001) on the ‘positiveness of phototaxis’ and angular direction of path. Furthermore, the significance of the interaction between the state of adaptation and condition of ocelli (P<0·001 for ‘positiveness of phototaxis’, P<0·01 for ‘deviation from parallelism’) showed that hoppers with their ocelli normal or occluded behaved differently in the two states of adaptation. Separate analyses of variance for each state of adaptation showed that occlusion of the ocelli produced a significant effect only in the state of light-adaptation, decreasing the positiveness of the reaction and increasing the angular ‘deviation from parallelism’.

In a repetition of this experiment using five third-instar nymphs each hopper was subjected to the stimulation of one light source 10 times at each of the presentation angles. As ocellar occlusion was found to have an effect only after lightadaptation no observations were made for the dark-adapted state. The sequence of dark- and light-adaptation and of ocellar occlusion as described in the previous experiment was however retained. Statistical analysis of the results shown in Table 6 indicated that occlusion of the ocelli increased the angular ‘deviation from parallelism’ (P<0·01), but appeared to have no significant effect on the ‘positiveness of phototaxis’. These results thus only partly confirmed the results of the previous experiment using fifth-stage nymphs.

Final experiments were made to investigate the part played by the ocelli in phototaxis when the compound eyes are occluded. Hoppers with the ocelli alone functional were placed at varied distances from, and at different angles to, a source of illumination. Their paths shown in Fig. 13 were found to have a strong resemblance to those made by completely blind hoppers. To confirm this observation the following experiment was designed.

The compound eyes of four fifth-stage nymphs were occluded and the specimens were dark-adapted for 1 hr. With the apparatus arranged as described previously, each hopper was placed at the centre of the circle seven times at each of the four presentation angles and subjected to the illumination of a single light source. The angle of response was recorded, the hoppers were then light-adapted for 1 hr. and further trials were recorded. Lastly, the ocelli were occluded and the hoppers, now completely blind, were again subjected to stimulation. Analysis of variance of the results shown in Table 7 indicated that there was no significant difference between the reactions of blind hoppers and of those with the ocelli alone functional in the light- or dark-adapted states. When the only form of stimulation is a light source, blind specimens will exhibit no directional response. Movement will be random and the mean angle of ‘deviation from parallelism’ will be 45 degrees. A final statistical test was made to ascertain whether the experimental values obtained with the ocelli alone functional differed significantly from this expected mean. A non-significant probability was obtained, indicating that the hoppers behaved as if blind.

These experiments have shown that the phototactic response of walking hoppers depends on their state of photic adaptation. Dark-adapted hoppers exhibit a light compass reaction and walk approximately at right angles to a light source, with fixation of the stimulus by a set of ommatidia in the lateral region of one compound eye. Light-adapted hoppers progress towards the source with fixation of the stimulus in the anterior regions of both compound eyes. It is difficult to explain the mechanism responsible for these two types of orientation. It has been shown (Roeder, 1953) that light-adaptation may be described as a loss in sensitivity of the eye and that an increase in light intensity is required to cause recurrence of a response. Presumably light-adaptation in the locust results in a decrease in the sensitivity of the compound eyes, the lateral ommatidia are then unable to fix the source and a response may be mediated only by the more sensitive ommatidia of the anterior regions of both compound eyes. Occlusion of the ocelli was shown to decrease the efficiency of orientation only when the sensitivity of the eye is decreased by light-adaptation and when a response is evoked by the stimulation of-the ommatidia of the anterior regions of the compound eyes. When the compound eyes are occluded, the ocelli alone are not capable of mediating a directional response under the light conditions of the laboratory, and the insects behave as if blind.

Investigations into the functions of the ocelli have been made by many previous workers. Parry (1947) remarks on the irregular distribution of ocelli among insects, whilst Kalmus (1945) notes a correlation between the presence or absence of wings and ocelli in the systematic insect groups. From a consideration of the optical powers Kolbe (1893), Hesse (1908) and Link (1909 a, b) suggest that the ocelli are used for the perception of distant objects, whilst Müller (1826), Lubbock (1889) and Lowne (1870) conclude that the ocelli are capable of perceiving only near objects. Von Buddenbrock (1937) supports both the above views. It has been claimed by Demoll & Scheuring (1912) that the ocelli function as distance assessors, whilst Lowne attributes to the ocelli the capacity of stereoscopic vision. These views assume that the ocelli are capable of perceiving form, a property which has been disproved by Homann (1924), Wolsky (1930) and later by Parry (1947). Other workers, viz. Bozler (1926), Parry (1947) and Von Buddenbrock (1937), have concluded that the ocelli stimulate the compound eyes thereby serving a photokinetic function. That the ocelli play a part in phototaxis has been suggested by Müller (1931) and Friederichs (1931), Kalmus and Von Buddenbrock, but that the ocelli appear to play no part in immobile orientation (photo-akinesis) has been demonstrated by Volkonsky (1939). Lowne (1878), Homann (1924), Wolsky (1933), Parry (1947) and Lubbock (1889) suggest that the ocelli possibly perceive light intensity or quick changes in light intensity.

The present work has shown that the efficiency of phototactic reactions mediated by the compound eyes is under certain conditions significantly reduced when the ocelli are occluded. In Calliphora their influence is manifest only under conditions of high light intensity when the less efficient response appears to be mediated by the antero-lateral ommatidia. In Locusta their influence is manifest only when the sensitivity of the compound eyes is reduced by light-adaptation, and when orientation depends on the stimulation of the anterior part of the eye. In both species the experimentally observed behaviour may be explained by the following theory.

Orientation of normal Calliphora is such that there is equal stimulation of the anterior regions of both compound eyes, but on the occlusion of the ocelli this arrangement is upset. It is suggested that ocellar occlusion results in a decrease in the sensitivity of the ommatidia of the anterior region of the compound eye, that they are then unable to function together and that orientation is mediated by the antero-lateral ommatidia. By this process the decrease in sensitivity of the eye is made good by less direct responses in which a greater number of ommatidia become stimulated. This theory is substantiated by the results obtained with Locusta, for a decrease in the efficiency of response of this species is evident only when normal orientation is mediated by the anterior regions of the compound eyes, and not when the response is effected by the lateral ommatidia of one eye. It is possible that the ocelli function in co-ordinating the responses mediated by the contra- and ipsi-lateral ommatidia, described by Mast (1923). It must be pointed out that the effect of occluding the ocelli of Locusta is manifest only when the sensitivity of the compound eye is reduced by light-adaptation. Under these conditions the effect of occlusion may be more evident than when orientation is mediated by highly sensitive compound eyes. Also, in Calliphora the influence of the ocelli is evident only when stimulated by a high light intensity. It is suggested that the degree to which the ocelli influence the response depends on the intensity of stimulation and the state of sensitivity of the compound eyes.

To conclude this discussion on the contribution of the ocelli to phototaxis, one may consider the efficient phototactic orientation of normal insects to be the result of directional responses mediated by the compound eyes, the efficiency of movement being maintained by the ocelli which correct directional deviations. The suggestion that the absence of ocellar stimulation resulting from occlusion decreases the sensitivity of the anterior regions of the compound eyes does not imply that a physical change occurs in these facets, but that there is a decrease in the excitation of the nerve elements from these receptors. Such a decrease in nervous excitation will not only be manifest in a decrease in the efficiency of directional response, but also in the speed of reaction or photokinetic response to stimulation. In immobile orientation or photo-akinesis, however, where efficiency or speed of movement is not the essential criterion of response, the stimulatory function of the ocelli to the nervous system will not be evident.

The writer wishes to offer his sincere thanks to Prof. O. W. Richards, Dr N. Waloff and Mr R. G. Davies for their interest in the work and helpful comments on the draft of this paper. Acknowledgement is also made to the Anti-Locust Research Centre for the supply of locust hoppers.