Studies on Plastron Respiration: III. The Orientation Responses of Aphelocheirus [Hemiptera, Aphelocheiridae (Naucoridae)] in Relation to Plastron Respiration; Together With an Account of Specialized Pressure Receptors in Aquatic Insects

In the first paper of this series (Thorpe & Crisp, 1947, later referred to as Part I) it was shown that the active predatory bug Aphelocheirus aestivalis, which inhabits the moderately swiftly flowing parts of rivers and streams, provides one of the most perfect instances of respiration by means of a gaseous plastron. This method enables the animal to obtain all its oxygen from that dissolved in the water by means of an extremely thin film of gas, of constant but negligible volume, into which specially modified spiracles open. This film of gas acts as a gill, and is held in position by a pile of exceedingly minute hairs extending over the greater part of the body surface. The structure, arrangement and size of these hairs is such that they are unwettable by any surface forces and any degree of hydrostatic pressure, either separately or combined, to which they are likely to be subjected in the normal environment of the animal. Consequently there is no danger of the air film being lost and similarly no need for the animal ever to come to the surface. Since Aphelocheirus carries no air store or other appreciable volume of air with it, in contradistinction to other adult aquatic insects it is heavier than water. This, in combination with its powers of swimming and its ability to cling tenaciously to stones and vegetation, has enabled it to become a true bottom-living form.

Although Aphelocheirus is mechanically well adapted to resist increase of pressure up to 3 atm., it must, on account of its dependence on the oxygen content of the water, be safeguarded against sudden or uncontrolled descent into the deeper parts of the river where it might be trapped in a region of low oxygen tension. It therefore requires some mechanism to prevent its being swept by the current into stretches of the river where, for this or other reasons, the water is unsuitable. It is the purpose of this paper to consider the orientation mechanisms of Aphelocheirus in relation to its respiratory needs—that is, in effect, the main responses which ensure its ability to maintain itself in regions where the medium is suitable for its life and development.

The well-developed compound eyes are presumably capable of the light compass reaction and almost certainly of rudimentary form sight. This sense will therefore enable the insect to maintain position and direction in daylight when clinging to the substratum or when swimming in close proximity to objects.

That Aphelocheirus has a well-developed ‘dorsal light response’ can be shown by placing the insects in a dark room in a glass tank which can be lit from beneath by a 40 W. bulb. Directly the bugs are exposed to bottom lighting there is a marked tendency for them, if at all active, to swim on their backs. If they are stationary on the bottom or if they finally settle down they tend to remain the right way up, or at any rate soon to right themselves in spite of bottom lighting. This response to reversed lighting tends to fade as soon as they grow more accustomed to the tank, and consequently less active. Also, after some period of conflict between the dorsal light response and the thigmotactic response to the bottom, the latter tends to get the upper hand and the light is disregarded to some extent. The same response is found in fifth instar nymphs, provided again that they are in an active state. Aphelocheirus then has the ability to direct its swimming in the vertical axis by reference to the direction of incident light.

A hitherto unrecognized pressure receptor is to be found in the form of a pair of oval sensory plates situated on the sterna of abdominal segment 2 (Fig. 1). They show a well-developed nerve supply. Comparison of their appearance in section (Fig. 2) with the somewhat similar sense organs found in Nepa and which Baunacke (1912) had decided were pressure receptors, gives one good reason to assume at the outset that this is also their function in Aphelocheirus. The organs in Aphelocheirus consist of shallow depressions in the cuticle, these depressions being devoid of the typical plastron hairs of the body surface. Instead of plastron hairs the depressed surface is covered by a large number (approximately 60,000 per sq.mm.) of very much larger hydrofuge hairs, all pointing backwards and inclined at an angle of approximately 30° to the surface. In among these hairs are dispersed at intervals somewhat more thinly walled hairs, each supplied with a sense cell, the proximal processes of the cells which penetrate these sensory hairs together making up the nerve. The structure of these sensory hairs is in most respects characteristic of the tactile type of sensillum and the mode of action of the organs is assumed to be as follows : When the pressure is increased, the air film on the sense organ will be compressed, and since the hairs are strongly hydrofuge they themselves will be pressed down, thus moving the sensory hairs at the same time. When the pressure is decreased the air film will expand and the hydrofuge hairswill be enabled by their elastic tension to resume their former position.

The proximal margin of the sense organ runs right across and thus truncates the spiracular ‘rosette’, so that the air film on the sense organ is in intimate contact with the tracheal system via the minute spiracular openings. The two spiracles concerned differ from all the other spiracles in that they each open into a small air sac which itself gives off a fine tracheal branch to the main tracheal trunk as well as finer branches supplying nearby tissues. The air sac is (Figs. 1, 3) of remarkable appearance, its walls being folded and ribbed in a peculiar concertina-like manner such as might allow it a possibility of undergoing great changes in volume with changing hydrostatic pressure. Under a high power of the microscope the air sac resembles to some extent the air bladders of the larvae of Corethra and, just as in that insect, we have been unable to find any muscles or contractile fibres. If they were indeed present, it should be easy to detect them with the polarizing microscope. Dissection failed to reveal any nerve supply to the structure. It has not been possible to secure any direct evidence throwing light on the possible function of this air sac. Its resemblance to the air sac of Corethra larva is remarkable, but its volume is far too small—not more than one-fiftieth of the total tracheal volume—for it to have any function as a hydrostatic organ; nor is a hydrostatic organ required. The possibility that it acts as a volume control, maintaining at a constant level the position of the gas-water interface on the special organ of pressure sense with which it is in such close relation, seems ruled out by the fact that it is in open connexion with the main tracheal trunk via a fine trachea without any intermediate valve. It does, however, provide a very small adjustable region in a tracheal system which as a whole must be highly resistant to compression, and its probable function will be considered below when the exact mode of action of the sense organ is being discussed. Before leaving this subject it should be recalled that a small air sac, but without the peculiar folded structure of the present example, is found in other aquatic bugs (e.g. Nepa and Naucoris) in much the same position relative to the second abdominal spiracle but not associated with any specific sense organ.

The types of response to pressure and to low-frequency pressure changes in any aquatic animal may conveniently be classified in one of three categories :

• Response to local low-frequency pressure changes due to currents, turbulence and eddies in the surrounding medium.

• Response to changes in total pressure caused by changes in the depth of the animal below the surface.

• Response to pressure differences on the body surface resulting from the position of the animal in relation to the direction of gravity. This type of response is analogous to that given by a statocyst but is brought about through the small variation in pressure of the medium with depth over the body surface or between specialized receptors, and not a direct response to gravity.

The ability of Aphelocheirus to respond to each of these in turn will now be discussed.

• The considerable equipment of trichoid tactile sensilla possessed by both nymph and adult, together with the antennae, are likely to provide sense organs indicating the presence and direction of local eddies and currents.

• Aphelocheirus adults have a characteristic behaviour response to changes in pressure which will tend to keep them from going too deep. When placed in a pressure jar they show a characteristic movement with each increase in pressure. This movement consists of a sudden jerky swim upwards, often reaching to the surface. Even if not fully manifested owing to the bugs being in a sluggish condition, it is rare for the pressure to be increased without the animals giving at least a single kick with the hind legs. That this response is due directly to pressure and not to rapid sensitivity to solution of oxygen from the tracheal system consequent upon the water being less than at saturation at the imposed additional pressure, is shown by the fact that the response remains both when the water is saturated with oxygen and when the insects are made comatose by being kept in freshly boiled water. A sudden increase in pressure will at once produce momentary signs of activity even under these conditions. A reduction of pressure, on the other hand, produces no response of any kind under any conditions. The nymphs, on the other hand, give no clear and definite response to increases in pressure unless these are very sharp and sudden. The most which is obtained with increases of up to 70 cm. mercury is a slight general increase in activity.

  It was also found that adults gave the same response to pressure when exposed to a diffuse red light in the dark room as they did under ordinary conditions of light, showing that the upward swimming was not dependent for guidance on the dorsal light reflex.

  It can readily be shown, as described below, that the specialized receptors on the abdomen, which have been described above (§3 (i)), are in part responsible for the sensitivity to pressure increment. Various methods were tried for eliminating these sense organs, such as wetting them with a drop of ‘Nujol’ or sealing them with paraffin wax. Finally, it was found that the best method was to touch them with a platinum micro-cautery heated to a point just sufficient to destroy them rapidly. It was curious that this treatment seemed to result in far less shock than did the treatment with paraffin wax or ‘Nujol’. Bugs treated in this way were observed in a pressure control tank. It was found that bugs with both sense organs cauterized showed a general diminution in sensitivity to vibration of the substratum, but that the characteristic jerk of the swimming legs in response to abrupt pressure increase, while less evident, was by no means eliminated. This suggests that major pressure increases can be appreciated by some other organs—probably by the proprioceptor sense organs of the cuticle. It must be admitted that elimination of the sense organs would probably make directed upward swimming more difficult, and this might account for some lack of response even if the pressure increment were appreciated. Since, however, the response is diminished appreciably even in the light which has been shown to be effective in directing movement, it is reasonable to assume that the apparent loss in sensitivity is real and is due to the extirpation of the sense organs.

• It is readily observed that adults are capable of directed swimming on an even keel in dull red illumination to which they are insensitive. This involves the appreciation of position in relation to gravity in order to maintain stability during swimming, and it was thought probable that the paired abdominal sense organs might be responsible for this. Intact bugs (controls) and others from which the sense organs had been eliminated were tested for their ability to make direct downward swims (as distinct from passive uncontrolled sinking) in a large tank containing water 10 in. deep, at atmospheric pressure and a temperature of 16° C. Experiments were carried out in the dark room with a diffuse Wratten safelight on one side of the tank. When light was required from above, a 40 W. lamp was used 7 in. above the water surface.

The results of these experiments are set forth in Tables 1-4. Table 1 gives the times taken by a number of bugs to reach the bottom of the tank after being placed just below the surface. In order to overcome individual variations seven operated and ten control individuals were used, and a mean taken from the times of twenty trials. Applying the t test of significance to the means it is found that in the dark the cauterized individuals take much longer than the controls with a high level of significance (between 0·01 and 0·001). The ability to swim in the light is also affected, and here again the significance level lies between 0·01 and 0·001. It is also likely that even normal bugs will swim more efficiently in the light than in the dark (level 0·05). This points to the conclusion that in the light a directed swim downwards or upwards can be accomplished by means of a directional light response. In the dark only those bugs with the sense organ intact are sensitive to small pressure changes and so swim directly to the bottom; those with the sense organ cauterized are badly disorientated. In order to assess this factor, experiments were repeated under the same conditions, except that this time the number of turns taken by the bugs during the swim to the bottom was counted, all experiments being done in a diffuse red light. It will be seen that here again there is a marked difference as between the experimental animals and the controls in the number of turns taken.

The significance of the results in Exps. I and II in Table 2 may be shown as follows : The time taken for a turn is at most 1 and at least ⁠. We may divide up the total time (adding both sets of results) into periods of 1 sec. or into ⁠. intervals. Each interval may be considered as a trial during which a turn may or may not occur. In this way 2×2 contingency tables are obtained representing the two extremes (Table 2 A).

The value of χ² for the 0·05 level of significance for 1 degree of freedom is 3-84. It therefore follows that even after allowing for the longer time taken to reach the bottom by the operated bugs, the number of turns in unit time (rate of turning, Table 2) is slightly but significantly higher in the operated animals. It appears that the difference between the time taken by the cauterized and the control bugs to reach the bottom is not much, if at all, due to different speeds of swimming, but is an expression of the fact that whereas intact bugs usually swim straight downwards in a purposive manner, more turns and gyrations are taken by the cauterized bugs:

From Table 2 (Exp. Ill) it will be seen that fifth instar nymphs behaved rather like cauterized adults; although they turn less often they have little or no power of directed swimming in the absence of light. They tend to sink passively in both light and darkness; the proportion of ‘sinkers’, however, appears higher in the dark, but the difference is scarcely significant (o·1-0·05 level).

Some experiments were then performed in which the sense organ of one side only was treated. These experiments were set up as previously in diffuse red light. The general impression from preliminary observations with bugs with the right sense organ only destroyed was that they dived with a right-handed spiral, and those with the left destroyed with a left-handed spiral, but it was extremely difficult to follow accurately. Careful observations were therefore undertaken using a more precise classification of behaviour, and these are summarized in Tables 3 and 4. The connotation of the letters used in the tables is as follows: F, falling passively to bottom; S, swims straight to bottom; D, dorsal surface uppermost; U, swims upwards; R, right-hand spiral; L, left-hand spiral; RL, right- and left-handed spirals alternating; T, turns over in tumbling fashion; H, horizontal swimming; V, ventral surface uppermost.

From Table 3, which gives the result of a typical experiment, it will be seen that with both sense organs cauterized the direction of turning is random. In contrast to this, the turning of insects in which the left sense organ only is cauterized and the right sense organ only is cauterized is markedly directional.

The results in Table 4 were analysed statistically and significant differences are given in heavy type.

It will be seen that the result of unilateral cautery of the sense organ is to produce a spin towards the same side, to reduce the ability to swim straight to the bottom (S), and to increase the total number of turns (R and L) taken. When both organs are cauterized the total number of turns is reduced, but instead of swimming directly to the bottom, all sense of depth appears to have been lost, the number of horizontal (H) and upward (U) swims being possibly higher, and the tendency to sink passively (F) is significantly greater. Aphelocheirus adults therefore can control the direction of their swimming in relation to the vertical axis either by the light compass response or by very sensitive specific pressure receptors. The nymphs, on the other hand, lack the specific receptors and are therefore dependent solely upon the light compass response.

As recorded in an earlier paper (Part I), Aphelocheirus nymphs and adults show a rapid response to lack of oxygen in the surrounding medium. If placed in boiled distilled water they exhibit spasmodic but powerful upward swims to the surface, then display a typical attitude characteristic of asphyxia, and finally become comatose. The initial behaviour is similar to that described above as a result of sudden externally applied pressure, and it appeared possible that the organs of pressure sense might be responsible for both reactions.

In any organ of pressure sense, the movable sensory member (drum, membrane, hydrofuge hairs, etc.) should ideally have a definite position corresponding to a given external pressure. Moreover, it is essential that the medium internal to this member should be compressible. In its simplest form, therefore, the organ of pressure sense will be open to the environment on one side and to an internal gasfilled space on the other. The correct functioning of the organ will now depend upon the constancy of pressure which can be maintained within the internal space. As this space must be filled with gas in equilibrium with or actively secreted by the surrounding tissues, its pressure will only remain constant within the limits either of the constancy o£ gas tension in the tissues or the regulatory powers of the secretory system. It therefore follows that to achieve any absolute pressure sehse acting over long periods will be a very difficult problem for the animal.

On the other hand, a pressure sense acting over a short period of time will require only that the pressure changes in the internal space should be slow. In the extreme case of organs of hearing the pressure changes are so rapid that the internal space may also open into the atmosphere so that the pressure on both sides of the moving member is equal except for high-frequency vibration.

In Aphelocheirus the internal space of the organ of pressure sense is continuous with the plastron, and this in turn with the air in the tracheal system. It is essential therefore that the air in the tracheal system should remain at a constant pressure, otherwise the organs of pressure sense will become stimulated.

Putting A = 1 ·0 sq.cm., V=0 ·9 × 10^-3 c.c.,^*i=4 ·8 × 10^-4,

1/λ = 0 ·9X 10^-3/1 ·0 × 4 ·8X IO^-4=1 ·9 sec.

Any change, therefore, in the gas tension of the medium will very rapidly influence the pressure in the tracheal system and cause stimulation of the organ of pressure sense.

However, there seems to be no doubt that this insect is able to distinguish between oxygen lack and change in external pressure. If it is placed in water saturated with nitrogen the response is identical with that displayed in gas-free water, although the organs of pressure sense will not be appreciably stimulated by the latter treatment as the plastron will exchange oxygen for nitrogen, both at atmospheric tension. Even more conclusive evidence is to be obtained by experiments carried out on operated animals. The removal of the organs of pressure sense does not interfere with their reaction to oxygen deficiency. Hence the response to oxygen lack is not located in the specialized pressure receptors. As a critical test of the ability of the bugs to distinguish an actual pressure increase from loss of gas by the plastron due to unsaturation of the medium, a number of bugs were placed in water deficient in nitrogen but containing 20% saturation in oxygen, while controls were placed in water saturated in oxygen and nitrogen at atmospheric tension. The first group gave persistent ‘pressure-increment responses’, swimming powerfully to break the surface at intervals.

It is clear, therefore, that while the adult bug can respond independently both to oxygen lack and to increase in pressure, it cannot fully distinguish between the two stimuli, since unsaturation in oxygen must, like unsaturation in nitrogen, stimulate the pressure receptors as demonstrated above. The sense organs which are specific to oxygen want will of course be stimulated at the same time. There is apparently no nervous correlation between the organs of pressure sense and the organs of (presumably) chemical sense, whereby apparent pressure increments due to deficiency in the gas content of the water are disregarded; the reaction to these and to real pressure increments is the same, so that when nitrogen is deficient the animal behaves exactly as it would if it were exposed to water lacking in oxygen or to an increase in external pressure.

It should be emphasized that the dangers of oxygen deficiency are only likely to be met in deep, slow-moving or stationary water, and it is not surprising that the response both to pressure increase and oxygen deficiency is the same, viz. powerful upward swimming.

The possibility of deficiency of some gas other than oxygen, e.g. nitrogen, as in the above experiment, is entirely unnatural, and it is scarcely to be anticipated that Aphelocheirus would be adapted to meet this exigency.

The possibility that stimulation of the pressure receptors, caused by changes in the gas tension of the medium, might interfere with their natural efficiency as organs of balance must now be considered. In nature the changes in the oxygen content of the water will not be as sudden or as great as under the very artificial conditions of experiment, especially in flowing streams. But changes in oxygen content which occur, notably in the morning and evening, may be comparable in their effect with the small pressure changes involved in the mechanical orientation of Aphelocheirus by these sense organs. It may be in this connexion that the concertina-like air sac, placed close to each pressure receptor, is of value. Small fluctuations in the oxygen tension of the water will cause pressure fluctuations in the tracheal system which would tend to be damped by a small and very compressible sac in this position.

Movements of body and muscle might also cause fluctuations in the pressure of the tracheal system which would require to be damped in a similar manner. The range in which damping can occur is very limited, since resistance to external pressure by the tracheal system is the sine qua non of plastron respiration. This organ must be capable of withstanding a high pressure, if necessary, but under ordinary conditions it should be capable, within narrow limits, of taking up small pressure changes of short duration. In this respect it would function very similarly to the hydrostatic organs of Corethra.

When there is sufficient illumination for the animal to recognize objects, or when it is clinging to the substratum, both nymphs and adults will be capable of maintaining position and appreciating the direction and force of water movements. In the absence of contact with the bottom and even when there are no objects in view, the dorsal light response will give the animal an approximate indication of its relation to the surface and the bed of the stream, and, together with its tendency to negative phototaxy, will guide it back to the substratum—even where the incident light is not vertical. Once in contact with a rough object the thigmotactic response takes control. Thus, in accordance with observations in captivity, Aphelocheirus spends the greater part of the time during the day crawling on the bottom over stones, probably in search of small chironomids which appear to be its natural food.

When swimming horizontally away from visible objects, both adults and nymphs will be maintained in the correct position with respect to the vertical by the dorsal light response. The adult, however, has a further check on its position through the paired pressure receptors on the first abdominal segment, and will therefore perhaps be more expert in the very turbulent and swift-flowing parts of the stream.

Under conditions of low illumination in deep parts of the stream, or in the dark, the nymph will be unable to maintain position, but the adult relying on the organs of pressure sense will be capable of controlled swimming and will be able to appreciate its depth, rising towards the surface if the pressure is too great or the water deficient in oxygen. The nymph, on the other hand, does not respond to pressure increase unless this is considerable, and the response is not likely to bring it to the surface. It will respond to conditions of low oxygen tension, however, but in the dark its swimming will be uncontrolled. It is not surprising therefore that while the adults are commonly found in the stony, deeper and more turbulent regions towards mid-stream, there is a tendency for the nymphs which are less active to be found in the more sluggish parts clinging to strands of Potamogetón pectinatus.

Very few aquatic insects are heavier than water, and it is only in those which can sink passively to the bottom that organs of pressure sense are important. It is very interesting therefore to find that Nepa, a stagnant-water form which is heavier than water, has organs of pressure sense superficially similar to those of Aphelocheirus.

Although, in general, the sense organs of Aphelocheirus are somewhat similar to those of Nepa, they are by no means the same in the two insects. Nepa has three pairs of organs which are described by Hamilton (1931) as ‘three pairs of oval scars at the side of the third, fourth and fifth abdominal sterna’ (Fig. 4). Externally these show as oval membranes apparently perforated by small holes and surrounded by a sclerotized ring covered with fine hairs. A section of these organs in Nepa (Fig. 5) shows a rich nerve supply running to sensilla of two main types, the first terminating in a tiny papilla, and the second bearing a large flattened scale, recalling in shape toadstools of the genus Paxillus or Clitocybe. These sensory papillae of the first type are not exposed to the medium but are enclosed in an air space which is covered by an outer, membrane which may stand well away from the sensilla. The outer membrane is made up of the overlapping margins of the second or scale type of sensillum. Baunacke concluded that the organs function as hydrostats and that by their means the animal is enabled to keep itself orientated in the right direction by the pressure of water. The lower organ is, of course, under greater pressure than those nearer the surface. He states that if a normal Nepa is blinded so that visual orientation is eliminated and allowed to crawl on a small see-saw submerged beneath the water surface it turns round at once when the inclination is reversed, whereas the insect with all the organs put out of action will walk up and down indiscriminately, giving no response to reversal of the apparatus. Oevermann (1936) has confirmed Baunacke’s work but has shown that in addition Nepa possesses a general sense of position resident in the appendages and persisting after section of the abdominal nerves. By means of this sense the animal, even with its sense organs destroyed, can still give correct responses on the see-saw provided that its body is given additional weight with a small piece of lead. Weber (1930, p. 108), quotes one, Moller, as stating that Lethocerus (Belostomatidae) has a pair of more simple static organs composed of the deeply sunk abdominal spiracles with overlying tactile hairs. He, however, does not give the reference, and it has not yet been possible to trace the original paper. On the face of it there does not appear any reason why a bubblecarrying insect such as the Lethocerus should need a special organ for the detection of pressure changes, for one would think that the long sensory hairs among the bubble-holding hairs would suffice.

Baunacke’s theory that the lower organs, being under greater pressure than the rest, would provide the animal with a means of appreciating its position relative to the horizontal implies a remarkable sensitivity. Yet apparently not in fact greater than that of the pressure organs of Aphelocheirus. The organs in Nepa are about 2-3 mm. apart; thus at the most only about 5 mm. separates those on the third sternum from those on the fifth. Suppose, for the sake of argument, that the animal is on an incline of 45°; this will mean that the uppermost sense organ is 1·5 mm. above the lowest. This involves a pressure difference of 0·00015 atm. = 0·1 14 mm. of mercury.

For Baunacke’s theory of the mode of action of the sense organs to be plausible two conditions must obtain. First, all parts of the tracheal system must be effectively protected against the entry of water even under pressure. This condition certainly appears to be fulfilled. Secondly, the gas-containing cavity of the sense organ must be in communication with its fellows via the adjacent spiracles and the tracheal system. This point has hitherto seemed uncertain; though stated by Baunacke to be the case and implied in his argument, it is not clearly evident from his figures and is directly contrary to the observations of Hamilton, who states categorically that the membrane does not cover the site of the spiracle, which lies close beside the organ ‘but entirely separate from it In view of this contradiction we have made a careful study of the sense organ of Nepa by sections, injection and otherwise. As a result we have come to the conclusion that while even excellent sections often appear to support Hamilton’s interpretation they are apt to be misleading where the exact point of attachment of such a thin membrane is concerned, and we are convinced that Baunacke’s interpretation is essentially the correct one. This being so the organs of Nepa must be considered as constituting a single system connected by the trachea.

Since the membranes, particularly because of their laminated structure, must have considerable elastic properties, and since the sensilla of the first type are short, vertical and rigid, we must look upon a single organ not as being sensitive to a wide range of pressures but as constructed primarily to register whether the force acting on the membrane is tending to pull it out or push it in and whether or not the membrane is touching the small sensory papillae. Thus the system of organs must be regarded as acting according to the scheme shown in Fig. 6, the membrane of the uppermost organ being convex and that of the lowest concave. Such a system would respond with great sensitivity to relative changes as between different organs in the series, but not necessarily to changes of pressure affecting the whole animal uniformly.

Since Nepa is a soft-bodied insect, and its tracheal system does not require to be highly resistant to external pressure, the air in the tracheal system will be effectively exposed to the external pressure of the atmosphere and the head of water above it. Hence the pressure in the tracheal system which is applied at the internal surface of the membranes will always approximate to the pressure outside the membranes, thus allowing a sensitive all-or-none response to differences of pressure due to the animal’s position relative to the vertical. Nepa is unlikely to be troubled by eddypressure effects as it lives in still water.

We have carried out a number of experiments in which intact Nepa were exposed to increases of pressure up to 1 atm., but in no case have we been able to detect any response. This goes some way to confirming the above conclusion as to the mode of action of the sense organs. Moreover, it must be remembered that Nepa is very sluggish, is incapable of vigorous swimming, and is essentially a shallow-water animal dependent for its oxygen supply to a small extent on the oxygen content of the water, but mainly on the atmosphere. In the former method it raises the wing cases and exposes a shallow groove into which open the spiracles, and over which long hydrofuge hairs are extended to form a roof; in the latter method the posterior siphon is pushed out of the water, after gaining access to the surface by climbing up a plant or other object. As long as it can maintain its orientation in the vertical axis absolute pressure changes, within the limits it is likely to encounter, are probably immaterial to it. We have repeated Baunacke’s experiments using a somewhat smaller apparatus in which the movable plank of the see-saw (20 × 4*3 cm.) is enclosed in a glass tube 4-8 cm. in diameter to prevent the animal from falling off the sides. The tube is closed at each end by a perforated cork, and the whole, filled with water, immersed in a tank of water 30 cm. deep. Instead of blinding our Nepa we carried out the experiments in the dark room, the tank being illuminated from one side by a very dim and diffuse Wratten Safelight no. 3. Under these conditions visual orientation seemed to be eliminated, and we were able satisfactorily to confirm Baunacke’s results. Both Baunacke and Oevermann, however, had been content only to investigate the effect of eliminating all the pressure receptors, whereas we, in order to secure more information about the mode of action of the system, wished to study the consequences of extirpating some and leaving others intact. Accordingly, we divided a collection of thirty bugs into six lots and extirpated certain of their pressure sense organs according to the following scheme :

These groups were then tested on the see-saw under uniform conditions in the dark room and the number of correct turns immediately following a tipping of the see-saw was recorded in order to obtain a sample estimate of the probability P of a successful response for each bug. It was found that the control lot A gave a total of 68 ‘correct’ responses in 92 tests, the individual values of P being 0 · 69, 0 · 94, 0 · 835, 0 · 72, 0 · 61 and 0 · 5, giving a mean value of ⁠. Lot B was not significantly different, 43/56 and ⁠. Lots C and D were so seriously affected that the numerical significance of the tests was doubtful; but while two individuals in lot C gave occasional positive results (7 in 40) all those in D were negative. In general, all the insects in these lots were clumsy, sluggish and disorientated, lot C, in particular, being apt to turn aimlessly in circles while lot D showed in fact just as complete loss of power to respond to tilting of the see-saw as do insects with all the sense organs eliminated. Lot E were rather sluggish and slow so that experiments were difficult to perform, but, nevertheless, ability to respond was surprisingly little impaired, figures being ⁠. Finally, lot F, in which activity was greater than in E, gave results—49/115, —showing that the power to respond, though still present, was at a yet lower level. The significance of the results was determined as follows. The total variance for each group of individuals referred to as ‘lot’ was derived from the sum of the squared deviation from the mean due to the individual differences of each animal and of the calculated random variance due to sampling for each set of trials. The t test was then carried out between the various pairs of lots A, B, C, etc., using this total variance. Using this test it was found that the difference between A and B is negligible, and the effect of operation E is also outside the level of significance (P= 0 · 2). The difference between F and the control is, however, fully significant (P= 0·01), as are also, of course, the quite overwhelming effects on lots C and D.

The results as a whole indicate that for the response to persist it is necessary to have at least two of the organs symmetrically intact on each side. They also suggest, though they do not prove, that the anterior and posterior pair together make a slightly more efficient system than any two pairs of adjacent organs, and they show quite conclusively that the destruction of one organ only is more disorientating than the destruction of two symmetrically. All these conclusions seem to give satisfactory further support to the theory of action put forward by Baunacke—though it is a little surprising that elimination of the middle pair has generally no observable effect whatever. Thus the principle of action of the pressure-organ system of Nepa is essentially comparative, the relative pressure differences at each end of the system being registered by distensible membranes. Finally, there is a curious and as yet unexplained result of extirpation of the abdominal sense organs which seems worth putting on record. The operated and the control stocks of Nepa were kept in large shallow dishes of water, each containing a stone which projected above the surface. From time to time the operated individuals would emerge from the water and stand motionless on the stone for hours at a stretch, their front and middle legs flexed and their hind legs usually fully extended so as to raise the ‘siphon’ and hinder part of the body to its fullest extent, the inclination of the body often being more than 45°so that the insect appeared to be standing on its head. A remarkable feature of the performance was the tolerance to desiccation when in this state. Normally, Nepa are very sensitive to drying, and 10 min. or so in a dry atmosphere is usually enough to break the stubborn cataleptic state which is such a well-known characteristic of this animal’s behaviour (see Brocher, 1916; Steiniger, 1936). This abnormal stance and tolerance to desiccation were observed occasionally in all the groups B-F inclusive, but never in the controls. The periods in between the acrobatic bouts are evidently considerable, for at any one time there were usually one or two insects out of the 25 in this state but seldom more than three or four.

We have already noted the fact that it is vital for Aphelocheirus, in view of the risks to which it is exposed, to have some type of sense organ which will be rapidly sensitive to a wide range of pressure changes affecting the whole animal as it ascends or descends. For this animal therefore the type of organ found in Nepa would be unsuitable, since it appears to give an all-or-none response rather than a finely graduated one, and in Aphelocheirus, whose tracheal system is maintained at a pressure independent of the surroundings, it would function only at one level. Indeed, the mode of action of the plastron precludes, and its structure is designed to prevent, the equalization of pressure between the tracheal system and surrounding water. The type of organ actually found in Aphelocheirus is capable of graded stimulation as the hairs are depressed and the sensory end-organs successively touched. It would be unnecessary for Aphelocheirus to possess more than one pair of such organs, since it very rapidly traverses a distance equivalent to its own length.

It has been shown that the two sense organs of Aphelocheirus, interconnected as they are via the tracheal system, do function as a ‘differential manometer’ enabling the insect to keep on an even keel in the transverse plane while swimming rapidly forwards, as well as giving the animal some absolute pressure sense. The function of balance might be achieved as a result of the comparison of sensory input from the two organs were their cavities not in communication; but it seems clear that such an arrangement would be inefficient compared to the existing one.

We are very grateful to Mr W. E. China for his efforts (unfortunately unsuccessful) to trace the Moller reference for us.