ALBERT

Notes: Summary 1. In order to analyse the mechanism of accommodation in anurans, drugs (miotic or atropine) were applied to the cornea of anaesthetized animals to change the refractive state of their eyes. During such changes, the lens and cornea were photographed and the refractive state of the eye was measured using laser speckle refractometry. Measurements taken from the photographs confirmed suggestions by Beer (1898) that accommodation is achieved by moving the lens and not by changing the shape of the lens or cornea. The change in refractive state induced by pharmacological manipulation was about 10 diopters with an accompanying shift in lens position of about 150 μm. Calculations based on a schematic eye suggest a disparity between the amount of lens movement theoretically needed to produce a 10 D shift in refractive state and the amount actually observed. 2. The lens is probably moved by two protractor lentis muscles which are positioned so as to pull the lens towards the cornea (Tretjakoff 1906, 1913). Dissection and HRP preparations revealed that these muscles are innervated by fibres of the oculomotor nerve which relay in the ciliary ganglion. InR. esculenta andR. pipiens, the ciliary ganglion consists of only 8 to 12 nerve cells. 3. MS222 anaesthesia and lymphatic injection of curare cause the lens to move away from the cornea, presumably because they destroy the resting tonus of the protractor lentis muscles. We discuss this finding in relation to the frog's ‘resting’ accommodative state, and conclude that unparalysed frogs are likely to be myopic, and not emmetropic as previous work suggests. 4. Prey capture was analysed inR. pipiens after the disruption of accommodation by bilateral section of the oculomotor nerve. Estimates of prey distance remained accurate when vision was binocular. However, during monocular vision, when the oculomotor nerve was sectioned on one side and the other eye was either occluded or had its optic nerve cut, frogs consistently underestimated the distance of their prey. This result suggests, in agreement with earlier evidence, that accommodation is used for judging depth when vision is limited to one eye, but that binocular information predominates when it is available. 5. Atropine applied to the cornea of monocular frogs also causes distance to be underestimated. It is argued from this that frogs assess distance by monitoring the motor commands sent to their accommodative muscles, rather than by using sensory information from the muscles themselves.

Notes: Summary 1. To discover whether bees learn the colours of landmarks, individually marked foragers were trained to collect sucrose from a small reservoir on the floor of a room. The reservoir was placed at one of two sites each defined by its position relative to one of two different arrays of cylindrical landmarks. On each foraging trip, a bee encountered one of the two arrays. Once a bee was trained to both arrays, its pattern of search was occasionally recorded on videotape during test trials in which one array of landmarks was present and the sucrose absent. 2. Both training arrays were composed of two dark blue and two light yellow landmarks placed at the corners of a square. The arrays differed only in the arrangement of coloured landmarks. When bees were tested separately with each array, they searched close to the reward-site defined by that array (Figs. 1 and 2). They behaved similarly on tests in which dark yellow and light blue landmarks replaced the dark blue and light yellow landmarks respectively (Fig. 3). To distinguish between the two arrays, the bees must have used the arrangement of colours.

Notes: Summary 1. We have filmed maleEristalis sp. in their normal environment performing a sequence of behaviour which demonstrates that they have a memory of their position in space relative to visual landmarks in their surroundings. They hover stably in mid-air, periodically leaving their hovering station to chase passing insects. After the chase is over they return to approximately the same position in space. 2. The following findings indicate that they use visual cues to guide their return. The size of the “home” they return to depends on the proximity of landmarks, being more circumscribed the closer the landmarks are to the home (Figs. 3 and 4). The fly on its return decelerates as it approaches home (Figs. 7–10) and generally stops short of its previous hovering position (Fig. 6). The accuracy of the return is unaffected by the length of the previous chase (Fig. 5). 3. The fly seems able to return directly to its home from any direction (Fig. 10), though we have not been able to film the complete return after long chases. During the return the fly normally faces in the direction it is flying, though in the last stages it is not uncommon for it to be oriented at an angle of 40–60° to its line of flight. When hovering or flying slowly the fly changes orientation by rapid saccade-like movements and between saccades the orientation is kept very constant (Figs. 1, 2 and 10). 4. Flies will congregate and hover near to a single mobile marker placed in the middle of an empty lawn and, after a chase, return to a position close to it. If the marker is moved while flies are hovering near-by they follow the movement of the board if it approaches them (Fig. 12), but not if it retreats (Fig. 13). Thus by moving the board away it is possible to change the distance between the fly and the board. In these cases (Fig. 13) the fly remains stationary in its old position until it chases an insect, after which it returns not to this position, but to its original position with respect to the board before this was moved. This shows (1) the board must provide a cue guiding the return, and (2) the homing reflex does not operate constantly when the fly is hovering to correct for small involuntary displacements from a preferred position, but it is only turned on at certain times, for instance after a chase. 5. Blowflies perform a variant of the same behaviour: they bask on sunny walls waiting for passing insects which they chase, afterwards returning to approximately the same position on the wall (Fig. 16). Their resting site tends to shift slightly after each chase (Fig. 17), and an analysis of this drift indicates that they up-date their memory of home on each return. 6. In the Discussion we propose a model of how flies might use their memory of the position and form of landmarks as seen from their home to guide their return.

Notes: Abstract Grass frogs, Rana pipiens, will detour around a barrier to reach prey on the other side. However, if the distance between prey and barrier is short, frogs attempt to push through the barrier and reach the prey directly. The relationship between the probability of detouring and the distance between prey and barrier is the same whether the frog's starting position is 4 cm or 8 cm from the barrier. This suggests that frogs measure the absolute separation between the two objects. To discover whether the retinal elevation of the bottom of the barrier contributes to measuring this distance, the relationship between the frequency of detouring and barrier-prey distance was examined in several experiments in which the retinal position of the bottom of the barrier was manipulated. No evidence was obtained that the barrier's retinal elevation helps in gauging distance. On the other hand, retinal elevation influences strongly how far a frog lunges to reach its prey. It is suggested that different cues to distance are applied to the two classes of object because, under natural circumstances, it is difficult to judge where a barrier emerges from the ground. A barrier may be hard to detect below the horizon because of the low contrast between it and the ground, or because vegetation and ground litter mask where the barrier meets the ground. In contrast, the prey's movements make it easily detectable against a stationary background and the prey's short height means that partial occlusion will have little effect on its apparent vertical position in the visual field.

Notes: Abstract We investigated the ability of bees to associate a motor parameter with a sensory one. Foragers were trained to fly along a prescribed route through a large box which was partitioned into compartments. Access from one compartment to the next was through a hole in each partition. In two of the compartments, the back wall was covered with a grating of black and white stripes. Stripe orientations and the required trajectories differed in the two compartments so giving bees the opportunity to learn that one stripe orientation signalled the need to fly leftwards and the other rightwards. We videotaped the bees' trajectories through one of these compartments in tests with the grating on the back wall in one of four possible orientations. Flight trajectories to stripes in the training orientations were appropriately to the left or to the right implying that bees had linked a given flight direction to a given stripe orientation. With gratings oriented between the training values, flight directions were, under some conditions, intermediate between the training directions. This interpolation indicates that the training regime had induced a continuous mapping between stripe orientation and trajectory direction and thus suggests that trajectory direction is a motor parameter which is encoded explicitly within the brain. We describe a simple network that interpolates much like bees and we consider how interpolation may contribute to the ability of bees to navigate flexibly within a familiar environment.

Notes: Abstract The flights of individual wasps (Vespula) were recorded as they approached a small feeder on the ground that was marked by a black cylinder ca 15 cm away. Two navigational strategies are used in these approaches. Initially, the wasp aims at the cylinder, treating it as a beacon and fixating it with frontal retina. In the last stage of the flight, the wasp assumes a preferred orientation so that the cylinder takes up a constant, more peripheral retinal position as the wasp nears the feeder. Path guidance by image-matching is likely to be limited to this final segment of the return. Wasps could gain the information needed for these distinct navigational strategies during the learning flights that they perform on their initial departures from the feeder. They fly away from the feeder in a series of arcs while turning at a mean angular velocity of 226°/s. The cylinder tends to be viewed with frontal retina during the arcs suggesting that the information required for aiming at the cylinder is acquired then. For image matching, the appearance of the cylinder needs to be learnt when the wasp is in the orientation that it adopts close to the feeder on its return flight. Wasps tend to assume this orientation during learning flights while they face the feeder. Such inspections of the feeder occur at the ends of arcs when a wasp's turning velocity is low.

Notes: Summary We ask whether desert ants (Cataglyphis fortis) perform path integration on their homeward as well as on their outward journey. If path integration does occur on the return journey, then, after an enforced detour, the ant's trajectory should point directly at its nest. To test whether this is so, ants were trained to forage at a spot 25 m from their nest. As an ant began its return journey to the nest, it was caught and transported to a test area where it was released either 2 m or 12 m from a wide barrier which obstructed its homeward path. The direction of the ants' trajectory after detouring around the barrier corresponded closely to that predicted on the assumption that the home vector is accurately updated during the detour.

Notes: Abstract In order to analyse how landmarks guide the last stages of an insect's approach to a goal, we recorded many flights of individual wasps and honeybees as they flew to an inconspicuous feeder on the ground that was marked by one or by two nearby landmarks. An individual tends to approach the feeder from a constant direction, flying close to the ground. Its body is oriented in roughly the same horizontal direction during the approach so that the feeder and landmarks are viewed over a narrow range of directions. Consequently, when the insect arrives at the feeder, the landmarks take up a standard position on the retina. Three navigational strategies govern the final approach. The insect first aims at a landmark, treating it as a beacon. Secondly, bees learn the appearance of a landmark with frontal retina and they associate with this stored view a motor trajectory which brings them from the landmark sufficiently close to the goal that it can be reached by image matching. Insects then move so as to put the landmark in its standard retinal position. Image matching is shown to be accomplished by a control system which has as set points the standard retinal position of the landmark and some parameter related to its retinal size.

Notes: Abstract To investigate the priming of memories by contextual cues, bees were trained to negotiate two mazes in different places 25 m apart. In the first maze, bees flew leftwards when the inner wall of the maze was covered with 45° stripes or rightwards when the inner wall was coloured yellow. In the second maze, bees flew rightwards on viewing 135° diagonal stripes or leftwards on viewing blue. The trajectories evoked by 45° or 135° stripes were similar in both mazes. However, vertical stripes were treated like 45° stripes in maze 1 and like 135° stripes in maze 2. Contextual cues prime the response to stripes that are oriented in the training condition for that site so influencing responses to stripes in closely neighbouring orientations. What objects in a bee's surroundings determine its sense of place? Bees were trained to different visual patterns at two sites 40 m apart (A+ versus A– at site A, and E+ versus E– at site E). A+ was preferred over A– and E+ was preferred over E– at both training sites. A preference for A+ over E+ exhibited at site A dropped gradually with distance to suggest that spatial context includes both close and distant objects.

Notes: Summary 1. MaleSyritta pipiens have been filmed tracking other flies within a rotating drum in order to analyse the interaction between the optomotor and smooth tracking systems. Males have a forwardly directed region of enlarged facets with enhanced spatial resolution, and smooth angular tracking operates to keep targets on this fovea. The male's angular velocity (φ) is driven by the angular position of the target fly on its retina (θe 10 to 20 ms earlier (i.e. φ =k= κθe, where κ is about 30–40s−1). The optomotor system is used for controlling a fly's course. The rotational component of this reflex causes a fly to turn in the same direction as externally generated retinal image motion. When placed within a rotating drum, the angular velocity of freely flyingSyritta follows that of the drum (i.e. φ =Gφdrum, whereG = 0.8 for φdrum〈200 °.s−1). Thus, whenSyritta turns to pursue another fly, the image of its static surroundings will sweep across its retina in the other direction, tending to generate optomotor torque in opposition to that caused by the tracking reflex. Unless arrangements are made to cope with this predictable optomotor input, tracking will be slowed. 2. The angular velocity of a male tracking in a rotating drum has two components: one caused by the drum velocity; the other by the retinal position of the target image (i.e. φ =Gφdrum +kκθe). During tracking, then, the optomotor system is fully active and will minimiseunwanted image motion caused by a disturbance. There are at least three forms of interaction between the two reflexes which in stationary surroundings will allow tracking to operate without interference from an active optomotor system. It is shown that all of them are compatible with the above result. 3. One of these schemes is the follow-up servo in which the tracking system works by injecting a command into the optomotor loop, so changing the latter's set point, thereby inducing an angular velocity greater than zero. This scheme in its simplest form requires that tracking has the same delays and frequency behaviour as the optomotor response. Since the two reflexes probably have different frequency responses, the follow-up servo can be rejected. The gain of the optomotor response (G) measured in an oscillating drum falls steeply between 0.5 Hz and 5 Hz, as it should, if the optomotor response is to be stable. However, there are indications that the constant κ, which relates ϕ to φe during tracking has the same value, if θe oscillates at more than 6 Hz, as it has for very low input frequencies. 4. That the value ofk seems to be limited by stability requirements and is to a first approximation independent of input frequency suggests that the tracking system has indeed been tailored to cope with the optomotor input that must oppose tracking at low frequencies. A simple additive model in which the tracking and optomotor commands first come together at a final common pathway will give the required performance provided that the tracking gain (x in Fig. 9) is enhanced atlow frequencies to allow for the opposing optomotor input. However, a simpler way of eliminating the unwanted optomotor torque is to use the efference copy scheme of von Holst and Mittelstaedt (1950). In this scheme a copy of the tracking command signal is sent to the optomotor input in order to cancel the expected visual consequences of tracking. In general, efference copy is a useful way of mixing two reflexes with different temporal properties. 5. FemaleMusca exhibit behaviour that is somewhat analogous to the smooth tracking of male flies (Reichardt and Poggio, 1976), but is probably concerned, not with chasing small targets, but with maintaining the fixation of large stationary objects. When maintaining fixation of stationary objects, the tracking and optomotor systems will complement each other, rather than acting in opposition, so that the form of interaction between tracking and the optomotor system may be different in the two sexes. Thus the female tracking system need make no provision for the optomotor response, even though the latter is active during tracking (Virsik and Reichardt, 1976).