It is now widely accepted, following the work of Alexander, Kitchener and Briscoe (1944 ac), Wigglesworth (1944, 1945) and Beament (1945), that the thin layer of lipoid material which is present on or near the outer surface of insect cuticle serves to restrict the evaporation of water from the animal. Existing knowledge of the nature of this waterproofing layer has been built up from observations on the effect of dusts, heat, detergents and organic solvents on the rate of evaporation from intact cuticle and also from a study of the changes in permeability of natural and artificial membranes when coated with thin layers of lipoid material.

This paper describes experiments undertaken to obtain information on the nature of the waterproofing layer in the worker honey-bee.

Alexander et al. (1944 a) demonstrated that certain dusts were lethal to insects by virtue of their physical properties, and that this was due to their action in promoting water-loss. Wigglesworth (1944, 1945) showed that dusts can abrade the surface lipoid layer of the insect’s cuticle thereby increasing the rate of water-loss through the cuticle. Beament (1945) and Alexander et al. (1944c) found that certain dusts could adsorb artificial wax films deposited on such substances as tanned gelatin which exerted little orienting force : no evidence was obtained that dusts could adsorb wax films deposited on insect epicuticle unless abrasive action augmented surface adsorption. The above-mentioned workers all indicated that dusts had no effect on dead insects. Hurst (1943, 1948) considered that the action of the dusts was due to their adsorptive properties only, and suggested that the reported ‘abrasive’ action resulted from the presentation of fresh surfaces of the dust to the epicuticular lipoid. He also claimed that dusts were effective in promoting cuticular water-loss in dead insects.

Preliminary experiments with bees showed that Almicide, an alumina dust, when deposited on the cuticle of both live and dead insects, effected a marked increase in the rate of water-loss through the cuticle. Although Wigglesworth (1945) showed that the rate of diffusion of water through the cuticle was a passive process not altered by the death of the insect, the action of the dusts on the dead bees was considered to be a surprising result as no mechanical abrasion of the cuticle appeared to have occurred. The following experiments were undertaken to determine the mechanism by which dusts such as Almicide effect an increase in the permeability to water of the worker bee’s cuticle.

Experimental procedure

Unless stated otherwise, all the experiments were carried out on uniform groups of worker honey-bees, at least 5 days old, which had been recently killed with hydrogen sulphide. (This chemical was chosen as it is highly toxic to bees, has a rapid ‘knock-down ‘effect, and is not expected to have any effect on lipoid material.) Twenty bees were used in each test with a similar number acting as controls. After treatment, each group of insects was placed on a small perforated zinc tray lined with filter-paper; this was weighed and put in a desiccator over calcium chloride. The trays were reweighed at intervals up to 24 hr. For convenience the results are summarized by expressing the decrease in the weight of treated bees as a multiple of the loss of the controls. It was assumed that all the additional weight lost by the treated bees was due to the evaporation of moisture ; the reasons for this assumption are fully discussed by Alexander et al. (1944 a).

The dusting was carried out in Exps. I—III by placing 0· 5 g. of the dust (together with the bees) in a 200 ml. round-bottomed flask ; the flask was slowly rotated until the insects appeared to be thoroughly coated and excess dust was then blown out of the flask with a jet of compressed air.

Experiment I. The action of Almicide on living and dead bees

Forty living bees were anaesthetized with carbon dioxide and dusted with Almicide which had been kept in a desiccator for 48 hr. Twenty of the bees were killed at once with hydrogen sulphide, the remainder being permitted to recover and live for 40 min. so that the dust might penetrate between and abrade the articulating surfaces of the cuticle. They were then killed. A further group of twenty bees was killed and used as controls. The subsequent rate of water-loss of the bees killed immediately after dusting was 3·07, and of those which had lived 4 ·34 times greater, than the loss in the controls. These results confirm the action of Almicide suggested by the preliminary experiments, and indicate that although the rate of water-loss was greater when abrasion of the cuticle could occur, the dust did disrupt the waterproofing layer when it came into contact with the dead insects. (During the process of dusting the hairs on the bee would act as a buffer between the cuticle and glass and prevent abrasion occurring.) A microscopic examination of both groups of dusted bees revealed that a large proportion of the dust adhering to the bee was trapped in the plumose hairs, but more dust appeared to have reached the surface of the cuticle of the bees which had lived after dusting. This alone could serve to explain the increased rate of water-loss in this group of bees.

Experiment II. The action of different dusts

Groups of bees were dusted with eleven dusts widely differing in physical properties. (Before use, each material was maintained for 24 hr. in a desiccator over calcium chloride.) The dusts varied in their adherence to the bees, but this factor was not considered to be important in a qualitative test. The results (Table 1) show that all the dusts effected an increase in the rate of water-loss of the dead bees ; that the three most effective dusts were silica gel, Almicide and activated charcoal, the common physical property of these materials being their capacity to act as powerful absorbents ; that Bentonite and activated charcoal, both soft materials, were more effective than carborundum, which is hard and highly abrasive. These facts again indicate that abrasion is not an important factor in the action of these dusts in disrupting the waterproofing layers, and suggest furthermore that they may act by adsorbing the lipoid material.

Table 1.

The action of different dusts on the rate of water-loss from worker bees

The action of different dusts on the rate of water-loss from worker bees
The action of different dusts on the rate of water-loss from worker bees

Experiment III. The relationship between particle size and effectiveness

Workers with the ‘inert’ dust insecticides have shown that particle size was generally inversely related to effectiveness (Alexander et al. 1944b; David & Gardiner, 1950; Bartlett, 1951). To obtain data on the effect of variation in particle size on the action of dusts on the cuticle of the worker bee, two minerals were used —a range of china monodisperse fractions prepared by sedimentation and three alumina dusts of different physical characteristics. The same technique was used as in the previous experiments. Although different fractions varied in their power to adhere to the bees, the weight of dust adhering to each group did not vary more than ± 10 %. Unfortunately no measure could be obtained of the amount adhering to the cuticle itself. The results (Table 2) show that all the fractions increased cuticular water-loss. The progressive increase in effectiveness of the china clay as the particle size fell from 50 to 5 μ, confirmed the results of other workers mentioned previously, but a further decrease in size below 5 μ resulted in a decrease in effectiveness. This anomaly was explained when it was noticed on microscopic examination of the dusts that particles in the fractions 2·5 μ and less also existed as aggregates. The existence of these aggregates has been clearly demonstrated by Gregg & Hill (1953), using a similar range of kaolin fractions. In addition, it is of interest to record that whilst the fractions from 5 to 20 μ felt smooth to the touch, those under 1 μ were distinctly abrasive. Of the aluminas, the gamma grade with the smallest particles was the most effective and caused a particularly high rate of water loss. A visual inspection of the three groups of dusted bees showed that the gamma grade had the least dense deposit, but examination under the microscope revealed that more of it had reached the cuticular surface; the other two dusts were largely trapped by the plumose hairs.

Table 2.

The relationship between particle size and effectiveness of different grades of china clay and alumina

The relationship between particle size and effectiveness of different grades of china clay and alumina
The relationship between particle size and effectiveness of different grades of china clay and alumina

The surface of the worker bee’s cuticle

In order to proceed with an examination of the action of the dust in disrupting the waterproofing mechanism, it is necessary to visualize the dust in contact with the surface of the cuticle. The presence of hairs and the apparent ridging of the cuticle itself had been noticed, and as these factors are likely to influence the deposition of dusts, their significance was considered.

The cuticle bears three types of hairs: long plumose 0·2–0·4 mm., short plumose < 0·2 mm., and simple setae <0·15 mm. in length. The proportion of each type of hair varies on different regions of the insect. The long plumose type predominates on the thorax with an average density of 590 per mm.2. Measurements from enlargements of photographs of these hairs in situ, and calculation based on the average hair density, suggested that dust particles greater than 2 μ in diameter will be prevented from reaching the cuticle. Tests with the china clay fractions used in Exp. III confirmed that this figure was approximately true. The abdomen has an average density of 126 long plumose hairs per mm.2. Although the short simple hairs are more numerous here than on the thorax, they do not form an efficient barrier to dust particles less than 30μ. Many areas measuring approximately 30×100μ. were found to be devoid of hairs.

A study of the surface relief of the cuticle (Glynne Jones, Connell & Nixon, unpublished work) has shown that in the abdomen it is raised into a series of folds. Fig. 1 shows a diagrammatic section of abdominal cuticle drawn approximately to scale from sections and a phase-contrast examination of the surface (Fig. 2). In addition, the surface of each fold appears to be heavily pitted (Fig. 3).* In the thorax the folds are replaced by more distinct ridges which trace a hexagonal pattern.

Fig. 1.

Semi-diagrammatic sketch drawn to scale to show surface relief of abdominal cuticle of worker bee.

Fig. 1.

Semi-diagrammatic sketch drawn to scale to show surface relief of abdominal cuticle of worker bee.

Fig. 2.

Phase-contrast surface view of abdominal cuticle of worker bee. ×2400.

Fig. 2.

Phase-contrast surface view of abdominal cuticle of worker bee. ×2400.

Fig. 3.

Electron micrograph of transverse section of abdominal cuticle of bee showing folds, × 3680.

Fig. 3.

Electron micrograph of transverse section of abdominal cuticle of bee showing folds, × 3680.

If the cuticular surface was smooth, the worker bee would have a surface area of approximately 2·2 cm.2, but if the folds and pitting are taken into account the true surface area may be at least 10 times this value.

It would appear, therefore, that when a bee is dusted, as in Exps. I–III, the following considerations arise :

  • (a) A proportion of the dust applied will be held back by the plumose hairs, particularly in the thorax.

  • (b) Dust particles or aggregates greater than 5 p. will tend to rest on the uppermost parts of the folds and only come into contact with a relatively small proportion of the total surface area of the cuticle.

  • (c) The agitation of an insect after dusting will tend to promote the setting free of some of the particles trapped by the hairs, the settling of particles between the folds, and the breaking up of aggregates.

  • (d) When a bee is heavily dusted, the particles trapped in the plumose hairs and also resting on the folds of the cuticle may form a barrier sufficiently dense to restrict water-loss.

It thus appeared, that if the dusting technique in Exps. I—III was modified so as to prevent aggregates of large dust particles settling on the tips of the folds, thereby restricting the access of further particles to the underlying areas of the cuticle, the effectiveness of the dust would be increased as it would be brought into contact with a much larger surface area of the cuticle. Two such techniques were developed, one using an aqueous suspension of a dust, and the other, a dust cloud of very fine particles. These will now be described.

Experiment IV. Dipping bees in an aqueous suspension of alumina dust

A 1 % (w/v) suspension of the gamma grade alumina was prepared using 0·05 % (w/v) Bentonite as a suspending agent, and 0·01 % Lissapol as a wetter, This suspension was violently agitated to aid the dispersion of aggregates. Twenty recently killed bees were immersed in the suspension for 2 min. and another twenty bees were placed in water containing the same amounts of Bentonite and Lissapol for the same time, to act as controls. The superficial moisture was dried off in an air oven at 30° C. and the two groups of bees placed in a desiccator. The subsequent rate of water-loss of the treated bees was 12· 2 times greater than in the controls. It is appreciated that the alumina itself is hygroscopic and that this introduces an error into the comparison of rates of water-loss. However, subsequent examination showed that the amount of alumina adhering to the cuticle was very small, probably less than 10 mg./bee, and it was estimated that the error introduced could not affect the results by more than 10 %.

The increased effectiveness of this dust when applied as an aqueous suspension suggests that the technique brings the particles into contact with a larger surface area of the cuticle than is possible with flask dusting. Furthermore, this experiment clearly shows that the dust need not abrade the cuticle to be effective. Wet dusts have been shown to be inactive (Hurst, 1948; Bartlett, 1951) and if this is the case the alumina could not exert any effect until after the bees had been placed in the desiccator and there no movement of the dust particles against the cuticle was possible.

Experiment V. Exposure of bees to a dust cloud of activated charcoal

Approximately 5 g. of finely ground activated charcoal which had been dried at 100° C. for 24 hr. and passed through a 300-mesh BSS sieve was placed in a glass tube i m. in length and 4 cm. in diameter which had a single hole stopper at each end. The tube was inclined at an angle of 45°; the lower orifice was connected to a supply of compressed air and the other to a 500 ml. flask by rubber tubing. The compressed air stream set up turbulence within the tube and the continual impacting of the charcoal against the sides of the tube tended to break down aggregates. By varying the air pressure and slope of the tube the size of the dust particles entering the flask could be altered. It was thus found possible to produce in the flask a dust cloud of particles less than 1 μ in diameter. Twenty recently killed bees were exposed to such a cloud for 10 min., the bees being rolled over at intervals to expose fresh areas of cuticle. The treated bees, together with controls, were placed in a desiccator and the rate of water-loss from the two groups compared. In the first 2 hr., the dusted bees lost water at a rate 30 times greater than the controls, and in 24 hr. the average rate was 18 times greater. This greatly enhanced increase in the rate of water-loss using activated charcoal as compared with its effect in Exp. Ill was considered to be due to a much larger area of the cuticle being brought into contact with the dust. A detailed microscopic examination showed that approximately 40 % of the abdominal and 20 % of the thoracic cuticle appeared to be in actual contact with the dust particles. (These estimates do not include the areas of cuticle covered by the overlapping ridges of the cuticle.) The result of this experiment again seems to rule out abrasion as the main factor in the action of dusts in disrupting the waterproofing mechanism.

Discussion

The results of Exps. I-V clearly show that various dusts can disrupt the cuticular waterproofing mechanism of the worker honey-bee ; that the dusts are effective on both living and dead insects, and that they apparently do not need to abrade the cuticle to promote increased water-loss. The importance of considering the surface relief of the cuticle when developing dusting techniques is clearly demonstrated. All the most effective dusts possess powerful adsorbent properties, and when this fact is considered alongside the evidence that the rate of increased water-loss is proportional to the area of cuticle in contact with the dust, it would appear that the dusts act by adsorbing at least one component of the waterproofing layer. This is presumed to be of a lipoid nature.

Alexander et al. (1944 b), when considering the mechanism of dust action, suggested that the epicuticular lipoid film is preferentially attracted to the crystalline forces at the surface of a solid dust particle; it adheres and orients itself on the crystal rather than on the relatively structureless surface of the cuticle, which then ceases to be waterproof. Further work by Wigglesworth (1945) and Beament (1945) (with a variety of insect species which did not include adult hymenoptera showed that this only occurred when the lipoid layer was in the nature of a mobile grease as found in Blattids. When a wax was present, Beament showed that (at the surface of the epicuticle) there was an innermost layer of wax molecules which was specifically oriented and capable of resisting the forces of adsorption exerted by the dusts. Abrasion was held to be necessary to disrupt this orientation ; afterwards adsorption might occur on the dust. Hurst (1948), as indicated previously, disagreed with the main conclusions of Wigglesworth and Beament concerning the mechanism of dust action. Recently, Helvey (1952), using the Mexican Bean beetle, found that dust particles with sharp edges had little or no insecticidal effect whilst others, e.g. Attaclay with no obvious abrasive properties, were highly toxic.

The effect of ‘inert’ dusts on adult Hymenoptera was studied by Bartlett (1951). He was uncertain whether the dusts acted by abrading the cuticle and suggested the disrupting of the wax film through sorption by dust particles as a possible alternative explanation. Anderson & Tuft (1952) reported that an attapulgite clay was toxic to worker bees kept at a low humidity; this mineral has a high adsorptive capacity and is non-abrasive. It would seem likely, therefore, that in the worker bee and possibly other adult Hymenoptera the waterproofing layer is different from that in the insects examined by Wigglesworth and Beament, and far more susceptible to adsorption by dusts.

Experiment VI. The action of chloroform on the epicuticular lipoid of the honey-bee

Wigglesworth (1945) showed that insects placed in an atmosphere of chloroform vapour became less waterproof. He considered that this vapour disorganized the orientation of the wax molecules in the lipoid layer and showed that the effectiveness of the chloroform vapour was directly proportional to the hardness of wax.

The effect of chloroform vapour on the waterproofing mechanism of the worker bee was determined, using the same technique as in previous experiments for measuring the rate of cuticular water-loss. Groups of freshly killed bees were treated as follows : (a) group kept at 28° C. in saturated atmosphere of chloroform vapour for 1 hr. ; (b) treatment as in (a) but kept in vapour for 2 hr. Controls were kept at the same temperature. The subsequent increased rates of water-losses over controls were: (a) 1·36, (b) 1·64.

Further groups of bees were immersed in liquid chloroform at 28°C. for varying periods of up to 1 hr. The bees were gently shaken at intervals, and were then subjected to desiccation after the solvent had evaporated. The increased rates of water loss over controls during the first 2 hr. in the desiccator were for 1 min. exposure, 5 ·2; 12 min., 11·4; and 1 hr., 37·5. Further tests showed that the extent to which the bees were shaken in the solvent was important; bees kept in chloroform at 28° C. for 20 min. without shaking lost water at a rate 8 times faster than controls ; with agitation, the loss was 20 times faster. Treatment in boiling chloroform for 2 min. resulted in a greatly increased rate of water-loss (approximately × 60), and prolonged treatment did not increase this effect.

The results obtained with chloroform vapour indicate that its effect was small. When compared with the results obtained for Nematus, Blatta and Rhodnius by Wigglesworth (1945) it would seem that there is a hard wax present on the epicuticle of bees similar to that on Rhodnius and dissimilar to the mobile grease found in Blatta. On the basis of his results with Rhodnius, Wigglesworth (1945) suggested that as hot chloroform was required to effect an appreciable wax extraction, and as the wax so obtained was readily soluble in cold chloroform (Beament, 1945), it was possibly protected by another layer termed cement which was only attacked by hot chloroform. In a later publication (Wigglesworth, 1947) the presence of such a layer was confirmed. The present results, whilst following the same trend as those obtained by Wigglesworth with Rhodnius, do not appear to suggest that the chloroform ever removes any substance other than wax. The deep ridges or folds on the bee’s cuticle will tend to restrict the flowing of the solvent over its surface, particularly at the bottom of the ridge. It has been clearly shown that the effectiveness of the cold chloroform is greatly increased if the insects are shaken in the solvent ; such an action would increase the rate of flow of solvent against the cuticle, i.e. its ‘washing action’. It is considered, therefore, that the pronounced effect of boiling chloroform is merely due to an extension of this action, the increased temperature having an additional effect on the speed at which the chloroform frees the wax from the underlying protein. When this evidence is considered alongside the fact that the wax layer is readily adsorbed by a dust without mechanical abrasion taking place, it would appear that in the worker bee there is no continuous layer of cement protecting the underlying wax.

Experiment VII. The effect of temperature on permeability

Wigglesworth (1945) using the cuticle of intact insects, and Beament (1945) using films of wax isolated from insect cuticle, showed that the cuticular waxes have fairly definite melting-points, and a ‘critical temperature ‘for the passage of water occurs about 5–10° C. below the melting-point. It was suggested that at this ‘critical temperature’ the orientation of the wax molecules changes so as to permit the passage of water molecules. The soft waxes showed improved waterproofing after they were heated above their critical temperatures and allowed to cool.

To determine the effect of increases in temperature on the permeability of the bee’s cuticle, groups of freshly killed bees were subjected to a range of different temperatures for half hourly periods. The bees were placed in a small tube which was immersed in a thermostatically controlled water-bath. After treatment, each group was placed in a desiccator to cool together with a control group which had been kept at room temperature. The two groups were then weighed and returned to the desiccator and re-weighed after 24 hr. Table 3 shows the average loss in weight for one bee in each group compared with that of the corresponding control group. The difference between these two figures, indicating the effect of the different temperatures, shows that heating the bees up to 58° C. causes a slight increase in waterproofing. Further increases in temperature reverse this effect and above 63·2° C. the decrease in waterproofing becomes very pronounced. Presumably this coincides with the melting of the wax. It would appear then that the critical temperature for the lipoid layer on the epicuticle of the honey-bee is near 59° C., a temperature significantly close to its melting-point, viz. 63–65° C. This further strengthens the hypothesis that the lipoid consists of a hard wax, and its melting-point suggests that it might be similar to beeswax (average m.p. 63° C.).

Table 3.

The effect of temperature on cuticular water-loss

The effect of temperature on cuticular water-loss
The effect of temperature on cuticular water-loss

Experiment VIII. The effect of increased temperature on the susceptibility of the lipoid layer to disruption by dusts

A group of dead bees was kept at 57° C. (just below the critical temperature) for 6 hr., the air in the flask containing the bees being kept saturated with water vapour. A similar group was kept at 20° C. for the same period, also in a water-saturated atmosphere. After treatment, half the bees from each group were dusted with Almicide using the technique as in Exps. I—III and the remainder kept as controls. The rates of water-loss of the various groups were determined as in previous experiments. Special care was taken to ensure that both groups were evenly dusted.

The results in Table 4 indicate that the increased temperature has made the lipoid layer less susceptible to disruption by dust. It is suggested that the increased temperature improved the orientation of the wax molecules in contact with the underlying protein (cf. Beament, 1945).

Table 4.

The effect of temperature on the action of Almicide

The effect of temperature on the action of Almicide
The effect of temperature on the action of Almicide

Wigglesworth (1945) demonstrated that living insects would slowly recover their impermeability after dusting. In Rhodnius the degree of recovery depended on whether the dust was removed after application, and it was considered that once the protective wax film had been disrupted by abrasion, recovery could be impeded by the adsorptive action of dust particles on the cuticle.

Experiment IX. Demonstration of recovery after dusting by worker bees

Forty living workers were exposed to a dust cloud of activated charcoal for 1 min. using the technique described in Exp. V; very little dust was visible on the bees. Ten dusted bees were killed and their rate of water-loss compared with controls.

The remainder were caged with a supply of sugar syrup at 30° C. and a relative humidity of 65 %. After 24 hr., ten bees were removed and their rate of water-loss compared with undusted controls which had also been caged, and 48 hr. later the twenty remaining bees were killed and ten of them redusted. The rate of water-loss of each group was then measured and compared with controls.

The results (Table 5) indicate that the dusted bees, when kept alive, completely recovered their impermeability in 24 hr. and that further dusting again produced an increase in water loss. Further tests suggested that bees exposed to the same dust cloud for longer periods did not recover their impermeability, but a high rate of mortality of the test insects occurred and the results were probably not significant.

Table 5.

Variation in water-loss when dusted bees are kept alive for different periods of time

Variation in water-loss when dusted bees are kept alive for different periods of time
Variation in water-loss when dusted bees are kept alive for different periods of time

A preliminary investigation was made into the effect of dusts and solvents on insects allied to the honey-bee (other tests not quoted showed no appreciable differences among the three castes of the honey-bee).

Ten workers of three species of Bombus and of the wasp (Vespa germanico) were killed with hydrogen sulphide and placed in separate flasks. Each group was dusted with Almicide and the excess blown off with a jet of compressed air. The subsequent ratios of water-loss when compared with controls (Table 6) show that the dust had effected an increase in water loss with all species.

Table 6.

Effect of dusting Bombus and Vespa workers with Almicide

Effect of dusting Bombus and Vespa workers with Almicide
Effect of dusting Bombus and Vespa workers with Almicide

One group of worker wasps (V. germanica) was immersed for 2 min. in boiling chloroform and another in cold chloroform (20° C.) for the same period. Subsequent rates of water-losses were respectively 23·2 and 8·1 times greater than the controls. These figures, though at a different level, are comparable to those obtained with the worker honey-bee.

There seems no reason therefore to suggest that the properties of the epicuticular lipids of the honey-bee are in any way peculiar to that species, but are probably applicable to other Aculeates.

In the present paper evidence has been obtained which strongly suggests that the waterproofing layer on the cuticle of the worker bee embodies a hard wax probably similar to beeswax. It has been clearly shown that the permeability of the cuticle to water vapour is increased when a variety of dusts are brought into intimate contact with the cuticle, and the evidence suggests that the dusts act by adsorbing wax. The loss of waterproofing brought about by activated charcoal in Exp. V is half that which occurs when the bees are immersed in boiling chloroform (Exp. VI) which should effect the complete extraction of any lipoid material on the surface of the cuticle. As in Exp. V the dust appeared to be in contact with less than 40 % of the total surface of the cuticle ; it seems probable, therefore, that the dust and solvent are acting on the same wax layer which cannot be effectively protected by a superficial layer of cement. Beament (1945) found that the thickness of the wax layer on the epicuticle of a range of insect species averaged from 0·2 to 0·3 μ. These figures were obtained by relating the total wax extracted from the surface to the apparent surface area indicated by camera lucida drawings. No account was taken of any folds in the cuticle, and the method could not show whether the wax thickness varied in different regions of the same cuticle. Assuming an average wax molecule to be 100 Å. in length (Muller, 1930) and that all the molecules are oriented vertically, an even wax layer 0·2 μ thick would correspond to at least 20 molecules in depth. The work of Alexander et al. (1944a, c), Wigglesworth (1945) and Beament (1945) clearly shows that it is the innermost compact monolayer of oriented wax molecules which is the main waterproofing barrier.

If in the worker bee the wax layer was approximately 20 molecules thick, then it is difficult to conceive how a dust particle settling on the wax could exert any effect on the innermost layer of molecules. The dust particles would always preferentially attract to their surface the wax molecules not bound and oriented to the underlying protein. It is suggested, therefore, that in the worker honey-bee the wax layer approaches a monolayer in thickness, at least on some areas of the cuticle. If this is the case, we can envisage dust particles settling on the film and when they gravitate to their final resting position, will attract to their surface wax molecules present at the point of contact, thus producing minute gaps in the monolayer. It is probable that the Brown & Escombe ‘pinhole ‘effect (1900) operates in such circumstances, and the production of a large number of such gaps would greatly increase the permeability of the cuticle to water vapour.

In living insects the presence of dust between moving surfaces of the cuticle will tend both to abrade the wax layer and, as Hurst (1948) suggested, bring fresh surfaces of the dust into contact with the wax. It may well be that the main difference between the waterproofing mechanisms in the worker bee and Rhodnius is the presence of a continuous cement layer in the latter insect only. The need for abrasion by dusts in Rhodnius as demonstrated by Wigglesworth (1945) might then be explained in terms of penetrating the cement layer and not the wax.

  1. Experiments are described which show that the rate of water-loss from living and dead worker bees is increased when a variety of dusts are brought into intimate contact with the surface of the cuticle. The common property of the more effective dusts is their capacity to act as adsorbents. Considerable evidence has been accumulated to suggest that the dusts need not abrade the surface of the cuticle in order to effect an increased water-loss and that the dusts act by adsorbing the epicuticular lipoid.

  2. The surface relief of the cuticle of the worker honey-bee is described and the importance of considering this feature of the insect in any experiments dealing with the action of dusts is demonstrated.

  3. An evaluation of the physical properties of the epicuticular lipoid has indicated that it contains, or possibly entirely consists of, a hard wax similar to beeswax.

  4. The action of the dusts and of chloroform suggests the absence of a continuous cement layer, and it is suggested that the wax approaches a monolayer in thickness, at least on some areas of the cuticle.

  5. Living worker bees were shown to be capable of recovering their impermeability after dusting.

  6. The type of waterproofing mechanism described in the honey-bee is not thought to be peculiar to that species. It is probably present in other Aculeates.

I should like to thank Drs Potter and McIntosh of Rothamsted Experimental Station and Dr S. J. Gregg, University College, Exeter, for their criticisms of this manuscript, and also Dr J. W. L. Beament and Mr M. Holdgate, Department of Zoology, Cambridge, for their interest and encouragement in this work. I am particularly grateful to Miss J. U. Connell for valuable technical assistance and to Mr H. L. Nixon of Rothamsted Experimental Station for his help with the study of the cuticle and the preparation of Figs. 2 and 3. The research laboratories of Messrs English Clays, Lovering Pochin and Co., Ltd., kindly supplied the china clay dust fractions, and Messrs G.E.C. Ltd., the Almicide. This work was commenced during the tenure of a grant from the Agricultural Research Council.

Alexander
,
P.
,
Kitchener
,
J. A.
&
Briscoe
,
H. V. A.
(
1944a
).
Inert dust insecticides. Part I. Mechanism of action
.
Ann. Appl. Biol
.
31
,
143
9
.
Alexander
,
P.
,
Kitchener
,
J. A.
&
Briscoe
,
H. V. A.
(
1944b
).
Inert dust insecticides. Part II. The nature of effective dusts
.
Ann. Appl. Biol
.
31
,
150
6
.
Alexander
,
P.
,
Kitchener
,
J. A
&
Briscoe
,
H. V. A.
(
1944c
).
The effect of waxes and inorganic powders on the transmission of water through celluloid membranes
.
Trans. Faraday Soc
.
40
,
10
19
.
Anderson
,
L. D.
&
Tuft
,
T. O.
(
1952
).
Toxicity of several new insecticides to honey bees
.
J. Econ. Ent
.
45
,
466
9
.
Bartlett
,
B. R.
(
1951
).
The action of certain ‘inert’ dust materials on parasitic hymenoptera
.
J. Econ. Ent
.
44
,
891
7
.
Beament
,
J. W. L.
(
1945
).
The cuticular lipoids of insects
.
J. Exp. Biol
,
21
,
115
31
.
Brown
,
H. T.
&
Escombe
,
F.
(
1900
).
Static diffusion of gases and liquids in relation to the assimilation of carbon and translocation in plants
.
Phil. Trans. B
,
193
,
223
91
.
David
,
W. A. L.
&
Gardiner
,
B. O. C.
(
1950
).
Factors influencing the action of dust insecticides
.
Bull. Ent. Res
.
41
,
1
61
.
Gregg
,
S. J.
&
Hill
,
K. J.
(
1953
).
The aggregation of kaolinite
.
J. Appl. Chem
.
3
,
169
73
.
Helvey
,
T. C.
(
1952
).
Insecticidal effect of inert solid diluents
.
Science
,
116
,
631
2
.
Hurst
,
H.
(
1943
).
Principles of insecticidal action as a guide to drug reactivity-phase distribution relationships
.
Trans. Faraday Soc
.
39
,
390
411
.
Hurst
,
H.
(
1948
).
Asymmetrical behaviour of insect cuticle in relation to water permeability
.
Dite. Faraday. Soc. no
.
3
,
193
210
.
Muller
,
A.
(
1930
).
The crystal structure of the normal paraffins at temperatures ranging from that of liquid air to the melting point
.
Proc. Roy. Soc. A
,
127
,
417
30
.
Wigglesworth
,
V. B.
(
1944
).
Action of inert dusts on insects
.
Nature, Lond
.,
153
,
493
.
Wigglesworth
,
V. B.
(
1945
).
Transpiration through the cuticle of insects
.
J. Exp. Biol
.
21
,
97
114
.
Wigglesworth
,
V. B.
(
1947
).
The epicuticle of an insect Rhodniutprolixut
.
Proc. R. Ent. Soc. Lond. B
,
134
,
163
81
.
*

Note added m proof. Further work has suggested that the large pits shown in Fig. 3 are artefacts.