1. The average thickness of the epicuticular wax layers on the wing membranes and elytra of a number of different insects has been measured by relating the volume of extracted wax to the area of cuticle over which it was spread.

  2. The true surface area of the epicuticle was measured by krypton adsorption.

  3. The ratio of absorption area to apparent projected area was found to be 1·6 for the wing membranes, and 6·7–8·2 for the elytra, with an average value of 4·1.

  4. The wax layers were found to be remarkably similar in thickness on the wing membranes, ranging from 0·11 to 0·13 μ but to vary from 0·11 to 1 ·26 μ in the case of elytra, where the wax thickness appears to be related to extent of modification.

It is now generally accepted, following the work done principally by Ramsay (1935), Alexander, Kitchener & Briscoe (1944a, b), Wigglesworth (1945, 1947, 1948), Kramer & Wigglesworth (1950) and Beament (1945, 1955, 1958, 1959) that the impermeability of an insect’s cuticle to water is due to a thin superficial layer of wax in the epicuticle.

Beament (1945), as part of an extensive investigation of insect waxes, measured the thickness of this layer on the exuvia and puparia of a variety of insects by assuming the wax to form a complete and uniformly thick layer over the surface of the cuticle and so relating the volume of wax to the area of cuticle over which it was spread. This he did by extracting the cuticle with chloroform, weighing the extracted wax and relating this quantity, expressed as a volume, to the total surface area of the sample. The inherent limitations of this method will be discussed later, but Beament found the wax layers to be from 0·2 to 0·3 μ. thick.

Beament measured cuticular areas in three ways, namely by (a) camera lucida, (b) use of the formula S=KW23 and (c) ‘geometrical considerations’. These methods, however, will only record the apparent surface area of the cuticle, whilst the highly irregular nature of the epicuticle itself will be largely ignored. These irregularities in the surface (see Holdgate & Seal, 1956 ; Locke, 1959), which usually take the form of hairs and folds, although on a microscopic scale, will increase the area many times. Indeed, Glynne Jones (1955), who investigated the cuticular surface of the worker honey-bee, estimated the true area of the epicuticle of this insect to be at least ten times greater than the measured apparent area, the discrepancy being due to the additional area of the hairs and microfolds of the epicuticle.

The accurate measurement of surface area is, generally speaking, very difficult, but a technique has been developed by surface chemists which exploits the phenomenon of gaseous adsorption by solids and which enables one to measure the surface area of certain inorganic materials with some confidence. Up to the present time, however, this technique has not been tried to any great extent on biological materials. It was therefore considered that a re-examination of the question of wax thicknesses, using Beament’s method in principle, but measuring the true surface area of the cuticle by gaseous adsorption, would provide an opportunity not only for testing the suitability of the technique for biological materials, but for enabling one to determine to what extent a true measure of the epicuticular surface alters the final value of the wax thickness.

The gas-adsorption technique for measuring the surface area of a solid makes use of the fact that when an evacuated solid comes into contact with a gas, some of the gas molecules become physically adsorbed on the surface of the solid so as to form a complete monolayer at low gas pressures and many layers at high gas pressures. Thus, provided one can determine that point at which the monolayer is complete (the monolayer capacity of the solid), calculation of the surface area is a question of measuring the total volume of gas adsorbed by the solid at that point and expressing this quantity in terms of area by calculating the number of tightly packed molecules in the adsorbed film and assuming the area occupied by the gas molecule to be equivalent to its cross-sectional area.

This is of course expressing the principles of the gas-adsorption technique in very simple terms, and for a complete account Brunauer (1945) and Gregg (1951) are recommended.

In practice one usually measures the volumes of gas taken up by a known weight of outgassed solid (maintained at a constant temperature) at increasing gas pressures and plots the volumes of gas adsorbed per gramme of solid against the equilibrium gas pressures. A plot of this kind is referred to as an adsorption isotherm and in all five types are recognized (Brunauer, Emmett & Teller, 1938; Brunauer, Deming, Deming & Teller, 1940).

The theory of multimolecular adsorption of Brunauer et al. (1938), later modified by Brunauer et al. (1940) allows one to determine from the adsorption data the monolayer capacity of a solid and hence its specific surface. The specific surface of a solid is defined as the surface area per gramme of material.

The Brunauer, Emmett and Teller (BET) equation is usually expressed in the form
formula
where
formula
Thus, a plot of P/[V(P0— P)] against P/P0 will give a straight line, of slope (C–1)/(XmC) and of intercept I/(XmC), from which the monolayer capacity may be calculated. The monolayer capacity of a solid, which by definition is that quantity of gas which 1 g. of solid has adsorbed at the completion of the monolayer, is related to the specific surface by the equation :
formula
formula
The BET theory is imperfect in many ways (Gregg & Jacobs, 1948). One major limitation is that it can only be applied with confidence to adsorption data within the relative vapour-pressure range of P/P0 = 0·05–0·35. In spite of this and other limitations, the theory does provide a practical method of determining, with ease and repeatable accuracy, the specific surface of a solid from its adsorption data.

The main problems in applying the gas-adsorption technique to measure the surface area of insect cuticle can best be summed up in three questions. First, what is a suitable source of insect cuticle; secondly, can insect cuticle be successfully outgassed at room temperature; and thirdly, which gas should be used?

Material

Wing membranes and elytra were chosen as a suitable source of insect cuticle for the following reasons. First, both wings and elytra are isobilateral structures with an epicuticle forming the outermost layer of both surfaces. Such a structure is ideal for the present work where the gas molecules must be confined to the epicuticular surface alone. Secondly, in this method of determining the average thickness of the wax layer, one must assume that the wax forms a complete and uniformly thick layer over the surface of the epicuticle. These assumptions, which will be considered later, were thought more likely to be true of small and discrete areas of cuticle, such as wings, rather than exuvia and puparia with their heterogeneous structure; though the use of whole wings as a source of insect cuticle, as opposed to exuvia, suffers from the disadvantage that one can never be certain that the extracted wax was necessarily present in the cuticle alone.

Cuticle samples were prepared in the following way. The insects were killed with hydrogen sulphide gas and undamaged wings were removed from the thorax by pulling at their bases with a fine pair of forceps. The wings were washed in distilled water, dried between sheets of filter-paper and finally transferred to a weighed sample bulb. The tegmina of Periplaneta americana were not washed because of the mobile nature of the lipoid (Ramsay, 1935 ; Beament, 1955, 1958).

The outgassing of insect cuticle

Before the surface area of a substance can be measured by gas adsorption, the surfaces must be cleared of contaminating adsorbed molecules. This procedure is termed ‘outgassing’ and with inorganic materials is usually accomplished by subjecting them to a vacuum greater than 10–4 mm. Hg at about 100° C. for an hour or more. It was clearly undesirable that insect cuticle should be heated to such temperatures. But outgassing is an endothermic process favoured by high temperatures, and it was therefore quite likely that if carried out at room temperatures the surfaces would not be sufficiently cleared. This problem was investigated by leaving evacuated samples of cuticle under vacuum overnight attached to the adsorption apparatus and testing the vacuum the following morning. If complete outgassing is not achieved, the gas pressure in the apparatus rises owing to adsorbed molecules leaving the sample. As the result of a series of such tests, in which different samples of cuticle were subjected to progressively longer periods of evacuation, adequate outgassing at room temperature was found to be possible provided the cuticle was subjected continuously to a vacuum greater than 10-4 mm. Hg for a minimum period of 100 hr.

Krypton adsorption

In the volumetric method of determining areas, measured volumes of gas of known pressure and temperature are allowed to expand into a bulb of known volume containing the evacuated sample. The resultant reductions in gas pressures are recorded and the gas laws applied to determine the volume of gas adsorbed by the sample at each addition of gas. It is usual for the sample to be kept at a constant temperature throughout the experiment, and as physical adsorption is an exothermic process favoured by low temperatures, the bulb containing the sample is usually immersed in either liquid nitrogen (b.p. 77·6° K.) or liquid oxygen (b.p. 90·5° K.).

The gas most frequently used in surface measurements is nitrogen and this was tried in some exploratory experiments using a slightly modified version of Tomkins & Young’s apparatus (1953). But the results from these experiments were not entirely satisfactory, as the relatively low area-weight ratio of insect cuticle combined with the high saturated vapour pressure of nitrogen frequently led to the volume correction for unadsorbed gas exceeding the volume of adsorbed gas, indicating that the areas being measured were below the range of the nitrogen method. Now Beebe, Beckwith & Honig (1945) recognized the limitations in using nitrogen for measuring small surface areas (less than 1 m.2/g.), and suggested that a gas such as krypton would be more suitable because of its low saturated vapour pressure. (Viz. 1· 753 mm. Hg at 77· 6° K. (Kingston & Holmes, 1953) compared with 760 mm. Hg for nitrogen.)

The krypton-adsorption method was accordingly investigated and an apparatus was built designed by Dr S. J. Gregg of the University of Exeter. The method proved to be satisfactory and was used for all subsequent determinations.

Areas determined by krypton adsorption usually agree to within ± 5% of the known areas of standard materials, and in the present work agreement between individual surface measurements of a particular sample of cuticle was found to be within 2 %.

The krypton-adsorption apparatus

The apparatus (Fig. 1), which is built of ‘Pyrex’ glass, is enclosed in a constant temperature cabinet (not shown in the diagram) and is composed of the following parts :

Fig. 1.

Krypton-adsorption apparatus.

Fig. 1.

Krypton-adsorption apparatus.

  1. The krypton gas system, consisting of a cylinder of krypton gas (99–100% pure krypton, remainder xenon) and a small side-arm F, terminating in bulb D, in which the krypton may be frozen. This reduces the gas pressure in the cylinder to just under 2 mm. Hg and facilitates a more delicate control of the quantities of gas entering the doser. The gas system communicates, via high-vacuum taps T1 and T2 with:

  2. The doser (Vd), calibrated for volume with krypton and consisting of that part of the apparatus delimited by high-vacuum taps T2, T3 and T5, including an accurately calibrated Macleod gauge (M), designed to measure gas pressures ranging from 5 × 10–6 to 7·0 mm. Hg.

  3. The dead-space system, consisting of Va, the dead-space volume above the liquid nitrogen level Y, and Vc, that volume of the sample bulb immersed in liquid nitrogen. Both volumes were calibrated by weighing with mercury.

    Before the volume of the doser can be calibrated, or indeed any determinations made, the apparatus has to be outgassed in order to free the glass of adsorbates. This was done by evacuating the apparatus and flaming the glass with a gas-air hand-torch to a temperature of about 200° C., this operation being repeated until the apparatus was capable of maintaining a vacuum greater than 10–4mm. Hg, when left overnight with high-vacuum tap T3 shut.

Experimental technique

A bulb containing a weighed sample of cuticle is attached to the apparatus at ground-glass joint J, and the whole apparatus evacuated via tap T3, by means of a mercury distillation pump backed by a single-stage rotary pump, so that a vacuum greater than 10–4 mm. Hg is maintained. The cuticle is outgassed for 100 hr. Bulb D and the sample bulb S are immersed in liquid nitrogen, the latter up to calibration mark Y, where it is maintained throughout the experiment. Taps T3 and T5 are closed and a dose of krypton is let into the doser ; the pressure is measured on the Macleod gauge, the mercury levels being read with a cathetometer. T1, the temperature of the cabinet, is recorded and tap T5 is opened. A period of 15 min. (previously determined) is allowed for equilibrium to become established and the equilibrium pressure P2 and the cabinet temperature T2 are recorded.

This procedure is repeated with successive doses of gas and may be continued until the saturated vapour pressure of krypton is reached, if a complete isotherm is required, or simply limited to the BET region of the isotherm (viz. P/P0 = 0·05–0·35) if a measure of the surface alone is required.

By application of the gas laws the volumes of gas adsorbed by the sample (Vad ml.) may be calculated from the equation
formula
where

Data so obtained were plotted first as an adsorption isotherm and then as a BET plot. The monolayer capacity of the sample was then calculated by means of the BET equation (equation 1). Before the specific surface of the sample can be calculated from the monolayer capacity (equation 2), the cross-sectional area of the krypton molecule must be known. Although several different values are given in the literature, depending upon which property of the molecule has been used in the determination (Davis, De Witt & Emmett, 1947; Livingstone, 1949; Davis, Shuler & Weaver, 1950; Kingston & Holmes, 1953), that of 19·5 sq. Å., suggested by Beebe et al. (1945) has been used throughout the present work.

Some typical krypton adsorption isotherms of insect wings and elytra, together with the corresponding BET plots of the adsorption data, are given in Fig. 2. In all cases, the adsorption isotherms were found to conform to type 2 of the BET classification (Brunauer et al. 1940). This type of isotherm, which indicates the absence of very narrow pores ( < 10 Å. in diameter) in the surface, occurs frequently in surface chemical work and is one of the more fully understood isotherms to which the BET procedure can be applied with some confidence.

Fig. 2.

The krypton adsorption isotherms of insect cuticle and the BET plots of the adsorption data.

Fig. 2.

The krypton adsorption isotherms of insect cuticle and the BET plots of the adsorption data.

The krypton adsorption areas (hereafter referred to as the BET areas) of those wings and elytra which were investigated are given in Table 1, column 4. A comparison of these values with the corresponding apparent areas (Table 1, column 8b) gives a ratio of 1·6 for the wing membranes and 6·7, 7 ·3 and 8·2 respectively for the elytra, with an average ratio of 4·1. In the present work, the apparent area of a particular wing was determined by projecting the outlines of 10 of them on to a graduated screen and measuring the area covered by the enlarged images. This projection method was considered to be at least comparable to the three methods used by Beament (1945). In addition, microscopical examinations of the wing surfaces were made in an attempt to account for the area of the hairs which frequently occur in large numbers on the surface and which were likely to provide an important correction to the final apparent area. It is possible to calculate the total apparent area of these hairs by measuring their size and density and assuming their shape to be that of a cone. The relative importance of this additional area can be judged by comparing the values given in Table 1, columns 8 a and 8b, where to take two extreme examples, it may range from as little as 6 % of the apparent area, as in the case of the forewings of the worker honey-bee, Apis mellifera, to as much as 50 %, as in the case of the wings of the tsetse fly, Glossina morsitans.

Table 1.

The BET and apparent area of insect wings and the average thickness of the epicuticular wax layer

The BET and apparent area of insect wings and the average thickness of the epicuticular wax layer
The BET and apparent area of insect wings and the average thickness of the epicuticular wax layer

It is of some interest to note that the experimentally determined BET areas of the wing membranes (Table 1, column 4) are in the same relative order as the corresponding apparent areas (Table 1, column 8), and that the ratio of BET area to apparent area varies from one insect to another and from one part of the body to another.

On microscopical examination, the surfaces of the elytra, although showing considerable ridging, were found to be free of hairs and the relevant values in Table 1, column 8c, are the projected areas only.

The main problem in interpreting the BET areas is to decide whether the gas molecules have been confined to the outer surface of the epicuticle, for that is where the wax layer is situated, or whether the gas has penetrated into the substance of the wing, so recording the general internal area as well as the outer, superficial area. Although no direct evidence can be cited on this question, we can deduce indirectly from the data that the gas is confined to the outer epicuticular surface. Perhaps the most striking single piece of evidence is provided by the areas of those wings and elytra which were extracted with petroleum ether. In Table 1, column 4, the BET areas of extracted wings of G. morsitans and the extracted elytra of Tenebrio motitro are given. Comparing these areas with the corresponding values for the untreated wings, we can see that as the result of extraction, the area is more than doubled in the case of Glossina and increased more than tenfold in the case of Tenebrio. It is difficult to explain these differences in area solely in terms of an increase in the external surface of the epicuticle. Removal of the superficial wax layer will almost certainly open up the pore canal system of the cuticle to the krypton molecules, which in turn will probably enable the gas to reach the exocuticle and endocuticle. It is much more likely that with an extracted wing, krypton is penetrating into the substance of the cuticle, being adsorbed on to all available internal surfaces and so recording the internal area as well as the external, superficial area. This does not appear to happen with a normal wing. Furthermore, if krypton is penetrating into the cuticle it will only do so by diffusion, which at liquid nitrogen temperatures will be very slow. This point was carefully checked early in the work by determining experimentally the times taken by the krypton-cuticle system to reach equilibrium at increasing gas pressures. In all cases, equilibrium was found to be reached quickly and within 15 min. Finally, supporting evidence is provided from some experiments in which the BET area of a sample of normal wings was determined, followed by a similar determination in which the wings were cut into small pieces. The average area of the cut wings was always considerably greater than that of normal wings. In conclusion, one is struck by what we may term the ‘reasonableness ‘of the BET areas when compared with the corresponding apparent areas. The greatest difference that we find between the two is just under tenfold, in the case of the tegmina of Periplaneta americana, and in the majority of cases the difference is much lower than this. If the BET area included both the superficial and the general internal area, one would expect the difference to be much greater.

So far we have only examined the possible errors in the area measurements when considering the wing as a whole structure, but clearly the physical condition of the outer layers of the cuticle will also affect the surface area as measured by gas adsorption. Holdgate & Seal (1956) demonstrated the presence of a ‘bloom’ of wax over the surface of the pupal cuticle of Tenebrio molitor and such a layer if present generally throughout the cuticle of insects will greatly increase the area. The presence of a ‘bloom ‘of wax may be one explanation for the relatively large areas of the elytra, though the low area ratio of the wing membranes suggests that such a layer is unlikely to be present in this case. Due consideration must also be given to the effects which the rather rigorous experimental conditions may have on the cuticle. The prolonged period of outgassing at room temperature is likely to cause distortion and shrinkage of the cuticle and possibly the formation of fissures in the surface layers. In addition, high vacuum is likely to remove solvent from the wax so producing a more porous layer ; this latter effect may well be operating in the case of the grease of Periplaneta americana (Beament, 1955). Considering the wing membranes first, once again the low area ratio of 1.6 does suggest that the effects on the surface are comparatively slight, though with the elytra they may be a factor contributing to the high values.

Another possible source of error in the BET method which must be considered is the possibility of krypton entering the vein system of the wings via the tom wing bases. This point was checked by measuring the area of a sample of wings in which the base of each wing was sealed with high-vacuum ‘Apiezon’ wax. The results of this experiment indicated no appreciable difference in area between normal and sealed wings. This is certainly difficult to explain, but it may be due to the blood in the veins blocking the broken ends on clotting.

Dennell (1958) showed that atmospheric oxygen was unable to penetrate through the larval epicuticle of Calliphora vomitoria, and in the present work all the evidence does seem to indicate that the krypton is being confined to the outer surface of the epicuticle by the wax layer. As a result we may conclude that the BET method gives a reasonably accurate measure of the true surface area of the epicuticle.

After the BET area of a sample of wings was determined the wax was removed by refluxing with two successive lots of 50 ml. of petroleum ether (b.p 60–80° C.) for 15 min. at a time. The solution was filtered through filter-paper washed in hot petroleum ether and the wings were rinsed several times with fresh solvent to remove any solution remaining on the surface. The wax solution was concentrated, transferred to a weighing bottle and allowed to evaporate to dryness. The resulting wax was then weighed and the average thickness of the wax layer calculated by dividing the volume of wax (relative density assumed to be 0 ·96 g./c.c.) by the total BET area of the sample.

The average thickness of the wax layers of those wings and elytra which were investigated are given in Table 1, column 6, whilst in column 10, for comparison, the thicknesses of the same layers are given, calculated from the total apparent areas.

We can see from Table 1, column 6, that the wax layers of the wing membranes are remarkably similar in thickness, varying from 0·11 to 0·13 μ. Those of the elytra, however, differ considerably in thickness and show a progressive increase from Periplaneta americana (0·11μ) to Tenebrio molitor (1·26μ). This variation may be explained if we remember that elytra are essentially modified forewings in which the original function of flight has been superseded, to a varying degree by that of protection. Now the three insects Periplaneta, Dysdercus and Tenebrio provide a series in which the degree of modification of the forewings progressively increases and we can see that as the degree of modification increases so does the thickness of the epicuticular wax layer. In the cockroach, Periplaneta, the forewings are only sli ghtly modified and the wax layer is similar in thickness to that of an unmodified wing membrane (0·11μ). The hemelytra of the adult cotton-stainer, Dysdercus, provide an intermediary condition in which the anterior half of the wing is elytrum and the posterior half is membrane. Here the average wax thickness is 0·4μ. Finally, in the adult meal-worm, Tenebrio, true elytra are present, the principal function of which is the protection of the hindwings and the excessively thin dorsal tergites. In these true elytra the wax thickness is 1·26μ. The presence of such a thick layer is not difficult to explain, as Wigglesworth (1948) found that there is a continuous secretion of wax throughout the adult life of Tenebrio. This lipoid on examination was found to be a soft wax which could easily be removed from the surface of the cuticle, and it may well be that, as Tenebrio is a burrowing insect in which the elytra provide the main protection of the body, a continual supply of wax is necessary to replace that which is being continually removed by abrasion. Also, a thick layer of soft wax such as this has the advantage of providing an immediately available reserve, which can spread rapidly over the abraded areas of cuticle, so completing the all-important monolayer, before freshly secreted wax reaches the area.

In this method of measuring the thickness of the epicuticular wax layer we must assume first that the wax forms a complete layer over the surface of the cuticle and secondly that the wax is uniform in thickness. We can infer from the BET areas that the epicuticular wax forms a complete covering to the surface of the wings and indeed it is difficult to see how the wax layer can function as efficiently as it does if it is discontinuous. The first of these assumptions then may be taken as being reasonably correct. In the case of the second assumption, however, Shaw (1955) found that the wax layers of the young stages of Locusta and Sialis lutaria varied in thickness in different parts of the body cuticle. Unfortunately, Shaw does not describe how he measured cuticular areas so that a precise comparison of his results with those of the present work is not possible, although it should be noted that the variations in thickness which he found could equally be due to a variation in the ratio of true area to apparent area at the different parts of the body. As the present work has shown (Table 1, columns 4 and 8) this ratio varies from one insect to another and from one part of the body to another. We can only conclude that the wax layer may vary in thickness throughout the body cuticle but, as was mentioned earlier, wing membranes and elytra were used as sources of insect cuticle in an attempt to minimize this possible inaccuracy.

The average value of 0·25 μ suggested by Beament (1945) for the thickness of an insect’s wax layer can be seen to be more than double that obtained in the present work for wing membranes, and from this point of view it would be more correct to speak of an average thickness of 0·12μ. This difference in the two average values is probably due almost entirely to a more accurate measurement of surface area by the gas adsorption method. For comparison the wax thicknesses calculated from the total apparent areas of the cuticle samples have also been included in Table 1, column 10, and these values are also less than Beament’s average figure, although if the uncorrected apparent areas are used in this calculation (Table 1, column 8 a), the values average out to 0·25 μ. Mention has already been made of the variation in thickness of the elytral wax layers and clearly it is meaningless to speak of an average thickness in this case. Indeed, this argument may well be extended to the wax layers of insects in general. The number of molecular layers represented by the various wax layers are given in Table 1, columns 7 and 11, and have been calculated by assuming the wax molecules to be vertically orientated throughout the layer and their chain-length to be 100 Å. (Beament, 1945).

In conclusion, brief mention may be made of more recent work by Beament (1958). By accurately measuring the loss of water from large nymphs of the cockroach, Periplaneta americana, and taking the wax thickness as 0·25 μ and the total surface area to be 7·5 cm.2, Beament was able to calculate the absolute water permeability of the cuticle. It is of some interest to repeat this calculation, substituting the appropriate values for wax thickness and surface area obtained in the present work, for those used by Beament. If we accept the wax thickness to be 0·11 μ and assume the total true area of the nymph to be eight times greater than Beament’s figure for the total apparent area (Table 1) then the absolute permeability of the cuticle of the cockroach comes to 0·13 × 10–8absolute units compared with 1·65 × 10–8 absolute units given by Beament. This extremely low water permeability reflects the highly organized nature of the epicuticular wax and emphasizes its efficiency as a water-proofing layer.

The work was done during the tenure of a research studentship awarded by the Agricultural Research Council and forms a part of a thesis for the degree of Ph.D. of London University. I wish to thank Dr A. B. P. Page, Imperial College, London, for his help and encouragement throughout the work and Dr S. J. Gregg of Exeter University, for his instruction in the principles of surface chemistry.

Alexander
,
P.
,
Kitchener
,
J. A.
&
Briscoe
,
H. B. A.
(
1944a
).
Inert dust insecticides
.
Ann. Appt. Biol
.
31
,
43
.
Alexander
,
P.
,
Kitchener
,
J. A.
&
Briscoe
,
H. V. A.
(
1944b
).
The effect of waxes and inorganic powders on the transpiration of water through celluloid membranes
.
Trans. Faraday Soc
.
40
,
10
.
Beament
,
J. W. L.
(
1945
).
The cuticular lipoids of insects
.
J. Exp. Biol
.
21
,
115
.
Beament
,
J. W. L.
(
1955
).
Wax secretion in the cockroach
.
J. Exp. Biol
.
32
,
514
.
Beament
,
J. W. L.
(
1958
).
The effect of temperature on the waterproofing mechanism of an insect
.
J. Exp. Biol
.
34
,
494
.
Beament
,
J. W. L.
(
1959
).
The waterproofing mechanism of Arthropods. 1. The effect of temperature on cuticle permeability in terrestrial insects and ticks
.
J. Exp. Biol
.
36
,
391
.
Beebe
,
R.
,
Beckwith
,
J. B.
&
Honig
,
J. M.
(
1945
).
The determination of low surface areas by krypton adsorption at low temperatures
.
J. Amer. Chem. Soc
.
67
,
1554
.
Brunaeur
,
S.
(
1945
).
The Adsorption of Gases and Vapours
. Vol.
1
.
Physical Adsorption
.
Oxford University Press
.
Brunaeur
,
S.
,
Emmett
,
P. H.
&
Teller
,
E.
(
1938
).
Adsorption of gases in multimolecular layers
.
J. Amer. Chem. Soc
.
60
,
309
.
Brunaeur
,
S.
,
Deming
,
L. S.
,
Deming
,
W. S.
&
Teller
,
E.
(
1940
).
On a theory of the van der Waals adsorption of gases
.
J. Amer. Chem. Soc
.
62
,
1723
.
Davis
,
R. T.
,
De Witt
,
T. H.
&
Emmett
,
P. H.
(
1947
).
Adsorption of gases on surfaces of powders and metal foil
.
J. Phys. Chem
.
51
,
1232
.
Davis
,
R. T
,
Shuler
,
L.
Weaver
,
V.
(
1950
).
The adsorption of gases by solids
.
Proc. Pa. Acad. Sci
.
24
,
133
.
Dennell
,
R.
(
1958
).
Puparium formation in Calliphora vomitoria: a hitherto unsuspected function of the epicubcle
.
Ent. Exp. Appl
.
1
,
268
.
Glynne Jones
,
G. D.
(
1955
).
The cuticular waterproofing mechanism of the worker honey-bee
.
J. Exp. BioL
.
32
,
95
.
Gregg
,
S. J.
(
1951
).
The Surface Chemistry of Solids
.
London
:
Chapman and Hall
.
Gregg
,
S. J.
&
Jacobs
,
J.
(
1948
).
An examination of the adsorption theory of Brunaeur, Emmett & Teller and Brunaeur, Deining, Deming 8c Teller
.
Trans. Faraday Soc
.
44
,
575
.
Holdgate
,
M. W.
Seal
,
M.
(
1956
).
The epicuticular wax layers of the pupa of Tenebrio molitor L
.
J. Exp. Biol
.
33
,
82
.
Kingston
,
G. L.
Holmes
,
J. M.
(
1953
).
Adsorption by evaporated copper films at 78° K
.
Trans. Faraday Soc
.
49
,
417
.
Kramer
,
S.
Wigglesworth
,
V. B.
(
1950
).
The outer layers of the cockroach Periplaneta americana and the function of the oenocytes
.
Quart, J. Micr. Sci
.
91
,
63
.
Livingstone
,
H. K.
(
1949
).
The cross-sectional areas of molecules adsorbed on solid surfaces
.
J. Coll. Sci
.
4
,
447
.
Locke
,
M.
(
1959
).
Cuticular pattern in an insect. Rhodnius prohxus Stâl
.
J. Exp. Biol
.
36
,
459
.
Ramsay
,
J. A.
(
1935
).
The evaporation of water from the cockroach
,
J. Exp. Biol
.
13
,
373
.
Shaw
,
J.
(
1955
).
The permeability and structure of the cuticle of the aquatic larva of Stalis lutaria
.
J. Exp. Biol
.
32
,
330
.
Tomkins
,
F. C.
Young
,
D. M.
(
1951
).
The adsorption of gases on caesium oxide
.
Trans. Faraday Soc
.
47
,
77
.
Wigglesworth
,
V. B.
(
1945
).
Transpiration through the cuticle of insects
,
J. Exp. Biol
,
21
,
97
.
Wigglesworth
,
V. B.
(
1947
).
The epicuticle of an insect Rhodnius prolixus (Hemiptera)
.
Proc. Roy. Soc. B
,
134
,
163
.
Wigglesworth
,
V. B.
(
1948
).
The structure and deposition of the cuticle in the adult mealworm, Tenebrio molitor L
.
Quart. J. Micr. Sci
.
89
,
197
.