[ 8 ]

ON THE PENETRATION OF INSECTICIDES THROUGH THEINSECT CUTICLE

BY J. E. WEBB, B.Sc, PH.D. (LOND.) AND R. A. GREEN, M.R.C.V.S.The Cooper Technical Bureau, Berkhamsted

{Received 6 May 1945)

(With Two Text-figures)

CONTENTS

PAGE8

810

11

13

IntroductionThe effect of organic solvents on the rate of pene-

tration of diphenylaminePenetration of the wax phase of the cuticlePenetration of the aqueous phase of the cuticleThe effect of volatility of the solvent

INTRODUCTIONDuring work on the control of human pediculosisCraufurd-Benson & Macleod found that body lice,which normally have a high resistance to derris, wererapidly killed when high-boiling tar acids wereadded to the insecticide. A similar effect wasobserved in experiments with the sheep ked,Melophagus ovinus, when mixtures of cresylic acidand diphenylamine were applied either as a dustingpowder with a china clay base, or as a liquidwash. It was suspected that a rapid penetrationof the insecticide through the insect cuticle, in-duced by the presence of the solvent, might beresponsible for this increase in efficiency. If thesynergy, already found to exist between derris ordiphenylamine on the one hand, and phenols on theother, could be extended to other insecticides andsolvents, then an understanding of the factorsgoverning the phenomenon would have considerablevalue in deciding the correct mode of presentationof a particular insecticide.

THE EFFECT OF ORGANIC SOLVENTSON THE RATE OF PENETRATION OF

DIPHENYLAMINE

As diphenylamine has a high solubility in themajority of organic solvents it was decided to use thisinsecticide in the investigation . of the problem.A number of these solvents with widely differingphysical properties were tested for insecticidalefficiency, both with and without the insecticide.

Powders containing solvent + insecticide were pre-pared in the following manner. A mixture of 10%ground diphenylamine and china clay was dilutedwith more china clay to give a concentration of

Factors governing synergy between diphenyl-amine and organic solvents

The carrier efficiency of mixed solventsExperiments with other insecticidesDiscussionSummaryReferences

PAGE14

161717202O

0-25 % diphenylamine in the powder. To this, 1 %by weight of solvent was added, and the powder thenshaken mechanically for half an hour. In these pro-portions the diphenylamine was completely solublein all the solvents tested. Control powders con-taining i % by weight of solvent alone were preparedin a similar manner.

The test insect used in these experiments was thesheep ked, M. ovinus. A batch of five keds wasplaced in a corked specimen tube, liberally dustedwith the powder under test and incubated at 30° C.Treated keds were observed under the binocularmicroscope at half-hourly or hourly intervals for amaximum period of 30 hr. or until they died. Thecriterion taken to indicate death was the final cessa-tion of heart beat: in the ked the beating heart isreadily visible through the dorsal integument of theabdomen. When making observations, each ked wasremoved temporarily from its tube and the powderlightly brushed from the dorsal surface of theabdomen, sufficient powder being placed initially inthe tube to ensure a surplus with which the kedbecame recoated after each examination.

It was shown by Webb (1945 a) that the suscepti-bility of keds to ground derris root increased in thepresence of 5 % carbon dioxide which, by stimu-lating respiration, caused greater quantities of theinsecticide to be taken in through the spiracles. Todetermine the extent to which powders on a chinaclay base penetrate the spiracles, keds were dustedwith a powder containing 0-25 % diphenylaminewithout solvent, in air, and in a mixture of air and5 % carbon dioxide. As no difference in time ofdeath of the keds was observed in the two series, itwas assumed that the amount of diphenylamineentering the spiracles was negligible and that pene-

On the penetration of insecticides through the insect cuticlewas taking place primarily through the ex-

termn cuticle. Over a number of experiments with0-25 % diphenylamine alone in china clay, keds diedin 25-30 hr.

In Table 1 the results of the experiments withpowders containing diphenylamine + solvent, andsolvent alone, are shown. The time of death of kedsfor each powder is based on the findings of a numberof trials. It is seen that the addition of a solvent todiphenylamine in some cases reduced the time ofdeath from 25-30 hr. to as low as 1-5 hr., whereas inothers little or no reduction was observed.

Some of the solvents when used alone were foundto exert a lethal effect. With the exception of the

treated with benzyl alcohol alone died in 21 hr.,whereas those treated with benzyl alcohol + di-phenylamine died in 2-5 hr. This shows that areduction in the time of death of keds of the sameorder as that obtained on adding cresols or xylenolto diphenylamine can be obtained with a com-paratively non-toxic solvent. It is not proposed,therefore, to consider the toxicity of the solvent inassessing the results obtained with mixtures of sol-vent and insecticide, except where the toxicity of thesolvent alone approaches that of the mixture. Anexample of this is given by o-dichlorobenzene, wherethe effect of the solvent alone is as great as that of themixture. Here, the solvent is not held to promote

Table 1. In vitro experiments with diphenylamine

Solvent

w-Cresolm-CresolXylenolBenzyl alcohol/>-CresolOctyl alcohol4-Methyl-cyclohexanolQuinolineCyclohexanoneDiacetone alcoholCyclohexanolAcetophenoneBenzonitrileAnilineCarbitolDimethyl-anilineMethyl benzoateCastor oilAnisoleo-Dichlorobenzene

Time of death of keds in hours

Solvent +insecticide

I -5J'Si ' 52-52'54568

! 3172 2

232424

272726 +27 +2 1

Solventcontrol

1 0

99

2 1

930302726—

2 128——

26+ .—27

—

19

Time in hours at whichKCUb UCL-tHIlC lll l l l lUUllC

after treatment withsolvent control

1 0I'O1 02 0

I'O2 02 02 02 - 0—

2 01 06 0I'O

2 0

I'O2 0

3 0

N.B. Keds dusted with 0-25 % diphenylamine die in 25—30 hr.

cresols and xylenol, however, this was so slight thatit may, for practical purposes, be neglected. Evenwith these solvents, although their toxicities arerelatively high, they are considerably lower thanthose observed when diphenylamine is present.Almost all the solvents used caused immobility ofthe ked in a comparatively short time (see Table 1),but in every case subsequent death, when it oc-curred, was protracted. In many instances keds againbecame active, even after a lapse of more than 5 hr.

It is doubtful to what extent toxicity of solventinfluences results obtained with mixtures. Althoughkeds already suffering from cresol or xylenol poison-ing would probably be slightly less resistant to di-phenylamine than unaffected keds, the reduction inthe lethal dose of the latter due to the toxicity of thesolvent is unlikely to be great. For instance, keds

the rate of entry of diphenylamine through thecuticle and it is, therefore, placed lowest in order ofefficacy in Table 1.

As certain organic solvents of diphenylamine,notably phenols and benzyl alcohol, when added tothat insecticide greatly increase its rate of action,while others, such as dimethyl-aniline, methyl ben-zoate and anisole, appear to have little or no effect,it is clear that solubility of diphenylamine in thesolvent is not the only factor involved in this process.

It is suggested that solvents may increase the rateof penetration of diphenylamine through the cuticlein one of two ways. First, the diphenylamine may betransported through the cuticle in solution in drop-lets or small molecular aggregates of the solvent, and,secondly, the presence of solvent may render thecuticle more permeable to the insecticide. As the

10 J. E. WEBB and R. A. GREEN

time taken for the solvent alone to penetrate thecuticle, i.e. the time in which the solvent causes im-mobility of the ked, bears no relationship whatsoeverto the time of death when diphenylamine is present(see Table i), transport of diphenylamine throughthe cuticle in solution in the solvent is thought to beunlikely. The degree to which a solvent inducesrapid penetration of an insecticide through the insectcuticle will be referred to in future as its carrierefficiency.

Wigglesworth (1944) has shown that when thewaxes covering the surface of the cuticle are removedby abrasion with hard dusts, evaporation of waterfrom the underlying layers of the cuticle readily takesplace and death of the insect from desiccation ensues.It is clear, therefore, that the wax must be in contactwith a free-water surface which is continuousthroughout the cuticle and has its source in the bodyfluids of the insect. Thus the structure of the insectcuticle, although highly complex, may be considered

pregnated with celluloid were dipped in a j ^of beeswax in carbon tetrachloride. In this m ^ alamina of beeswax strengthened by threads of muslinand conforming to an approximate standard of thick-ness was obtained. A circular glass cell was thencemented on to the card surrounding the membraneand filled with a measured volume of solvent. Thetime taken for the solvent to pass through the bees-wax at 300 C. was then observed. This was repeatedtwelve times for each solvent and the mean of theresults obtained was taken as a measure of the rate ofpenetration of the solvent through beeswax. Thesefigures are given in the first column of Table 2.

On comparing the rates of penetration of thesolvents through beeswax with their observed carrierefficiency with diphenylamine (see Table 2), littlecorrelation exists between them except that aniline,carbitol and castor oil, those solvents either pene-trating slowly or not at all, agree in showing little orno carrier efficiency. Dimethyl-aniline, anisole, and

powder coating insect

wax phase

aqueous phase- ~

epicutic/e

exocuticleand

endocutic/elipoid elements

hypodermis

Fig. 1. Diagram showing the insect cuticle considered as a two-phase system.

simply as a two-phase system comprising an outerphase of lipophilic elements associated with theepicuticle and an inner phase of hydrophilic ele-ments associated with the exocuticle and endo-cuticle traversed by lipophilic elements extendingfrom the hypodermis to the epicuticle (see Fig. 1).This conception of cuticle structure was adopted asa basis for the clearer understanding of the problem.

PENETRATION OF THE WAX PHASEOF THE CUTICLE

The differences observed in carrier efficiency of thevarious solvents may be due to their ability to dis-solve or penetrate the outer wax layer of the insectcuticle. As it has been shown by Beament (1945)that this wax is similar in composition to beeswax,the rate of penetration of the solvents through arti-ficial beeswax membranes was taken as an approxi-mate measure of their relative rates of penetrationthrough the wax on the cuticle. A large number ofthese membranes was prepared in the followingmanner. Pieces of muslin supported by card im-

o-dichlorobenzene, however, penetrate beeswaxrapidly and also show no carrier efficiency. Althoughthe ability of a solvent to penetrate the wax coveringthe insect cuticle appears to be essential for highcarrier efficiency, it is evidently not the only factorinvolved.

The passage of wax solvent from the external sur-face of the insect to the hypodermis may take placevia the lipophilic elements along which the waxsecretions of the hypodermal cells pass to the epi-cuticle. Where diphenylamine is dissolved in thesolvent, however, unless the solvent passes throughthe cuticle in bulk, it does not follow that theinsecticide will be carried through the cuticle with it.It has been shown by Hurst (1943), and Lauger,Martin & Miiller (1944), that insecticides whosemolecules contain both an oil-soluble group and awater-soluble group, by orientating themselves atan oil/water interface, pass rapidly through insectcuticle by diffusion along such an interface. Althoughthis mode of penetration may be open to those sol-vents possessing a molecule of this type, it is evidentthat it is a property inherent in the solvent alone and

On the penetration of insecticides through the insect cuticle I I

affect the passage through the cuticle of anin?Wl!icide lacking a water-soluble group, eventhough that insecticide is dissolved in the solventwhen on the external surface of the cuticle. The rapidpenetration of diphenylamine in the presence ofcresols, xylenol and benzyl alcohol does not appearto be due solely to its passage through lipophilic ele-ments since, if this were so, there seems no reasonwhy dimethyl-aniline, anisole, and o-dichloro-benzene, in view of their high rate of penetrationthrough beeswax, should not also increase the rateof penetration of the insecticide. Transport of di-phenylamine across the aqueous phase of the cuticleis, therefore, the next consideration. ,

of solution of the organic solvents and insecticides inthe wax and aqueous phases of the insect cuticle thanthose taken after equilibrium between solute andsolvent had been established. The solubility figuresgiven in the tables cannot, therefore, be taken asphysical constants.

The partition coefficients of the solvents betweenbeeswax and water were then determined in thefollowing manner. To 400 ml. of boiling distilledwater was added 4 ml. of a mixture of equal parts byweight of molten beeswax and solvent. This was thenshaken for 5 min., cooled to 300 C , and filtered toremove the solid beeswax. The volume of solventleaving the beeswax and entering the water was then

Table 2. Physical properties of solvents

Solvent

o-Cresol»n-CresolXylenolBenzyl alcoholp-CresolOctyl alcohol4-Methyl-cyclohexanolQuinolineCyclohexanoneDiacetone alcoholCyclohexanolAcetophenoneBenzonitrileAnilineCarbitolDimethyl-anilineMethyl benzoateCastor oilAnisoleo-Dichlorobenzene

Rate of penetrationthrough beeswax at

30° C.

IS min.3-5 min.15 min.

320 min.I2-O min.i'S min.40 min.

120 min.80 min.

45-0 min.200 min.200 min.200 min.2 hr. 30 min.

15 hr.i-o min.70 min.Nil10 min.O'l min.

Solubility inwater in

c.c./ioo c.c. at30° C.

2-42 10-44 - ii-5o - i

o-5°'57 100

3-50025O-I3 600

Nil0-15Nil0 1Nil

Partition coefficientbetween beeswax

and water at30° C

0-0093009500-013200950000330-0013O-OO2IOO2OOOOI5High

0-0950NilNil

0-0450HighNil

0-0017Nil

0-0009Nil

Boiling pointin°C.

1 9 12 0 0

216-2172 0 62 0 1

185175238

1551641 6 12 0 11 9 0184198

1932 0 0

>2OO1541 8 0

PENETRATION OF THE AQUEOUSPHASE OF THE CUTICLE

The partition coefficient of solvents betweenbeeswax and water

A factor governing the passage of a solvent from thewax phase of the cuticle into the underlying aqueousphase is the partition coefficient of the solvent be-tween wax and the aqueous medium permeating theexo- and endo-cuticles. As the solubility of thesolvents in water is one factor governing their distri-bution between these two phases, this was- deter-mined at 300 C. for each solvent. These solubilitieswhich are given in the second column of Table 2 andall solubility figures appearing in succeeding tableswere determined by a rough method in which thesolute was shaken with the solvent for a short periodof time only. It was considered that figures obtainedin this way would more nearly represent the degree

determined by measuring the amount of solvent stillrequired to saturate the solution, and subtractingthis figure from that already obtained for the solu-bility of the solvent in water at 30° C. The partitioncoefficient for each solvent wasthen expressed as theconcentration of solvent in water over the concentra-tion of solvent in beeswax. These figures are given inthe third column of Table 2.

When a solvent, having penetrated the wax phaseof the cuticle, passes into the aqueous phase beneath,it would be expected to increase the rate of penetra-tion of diphenylamine if it either transported theinsecticide across the aqueous phase, or renderedthat phase more permeable to it. Any solvent shownto be unable to leave the wax phase would seem to beincapable of influencing the rate of penetration ineither of these ways. In this connexion, therefore,it is significant that acetophenone, benzonitrile,dimethyl-aniline, castor oil, and o-dichlorobenzene,

12 J. E. WEBB and R. A. GREEN

those solvents in which the partition coefficient is nil,all show very low or negative carrier efficiency. Theremaining solvents, with the exception of aniline,carbitol and castor oil which are unable to penetratewax, all pass more or less readily from wax intowater. These include, however, quinoline, cyclo-hexanone and cyclohexanol, which, in addition to ahigh partition coefficient, also have a high rate ofpenetration through beeswax, and would, therefore,be expected to possess a carrier efficiency as high asthat of the cresols, xylenol and benzyl alcohol. Sincethe carrier efficiency of these two groups of solvents

presence of solvent dissolved in the ^might increase the solubility of diphenylarm^Pinthat phase. When a solvent is incapable of leaving thewax phase, diphenylamine will pass from the solventinto the aqueous phase at a rate governed by itspercentage saturation in the solvent. This will de-crease as the diphenylamine leaves the solvent andthus the rate at which the insecticide enters theaqueous phase will fall.

The solubility of diphenylamine in water and insolutions of the solvents in water at 80, 60, 40 and20% saturation was measured at 30° C. In the case

Solvent

o-Cresolm-CresolXylenolBenzyl alcoholp-CresolOctyl alcohol4-Methyl-cyclohexanolQuinolineCyclohexanoneDiacetone alcoholCyclohexanolAcetophenoneBenzonitrileAnilineCarbitolDimethyl-anilineMethyl benzoateCastor oilAnisoleo-Dichlorobenzene

Table

Solubility ing./ioo g. solvent

at 300 C.

1 9 1

15513714914048

105270

355230130280

29529S160

245225

25265175

|. Solubility of diphenylamine*

Solubility in mg./ioo c.c. unsaturated solutions ofsolvent in water at 30° C.

80 % sat.

7 070+9-2+90+7-S5-76-33-o+3-5

400-03-57-07-0

3'S2700

—2 0—

4 0

60% sat.

8-77 3

1 0 01 0 0

8 060+7-0

5'52 8

128-04ot7-06-84-5+

ioo-o—2 0 +—4 0

40% sat.

7-7+8-56-57-S6 05-o67+S-o2 0

2 3 0

3-S7-05 25 0

34'O

2 - 0

6-5+

20 % sat.

7'55-85 26 0

5-otS-o5-o5-82-ot

io-ot3-o7-SS-o6-5

i7-ot—2'5—8-o—

* Solubility of diphenylamine in water at 300 C. = s-s mg./ioo c.c.t Solubilities selected in Table 4.

has been shown to differ considerably, it is evidentthat factors other than those already considered areinvolved.

The solubility of diphenylamine insolutions of solvents in water

The passage of a solvent from the wax phase of thecuticle into the aqueous phase would be unlikely toincrease the permeability of that phase to diphenyl-amine, unless it tended to increase the rate of dif-fusion of the insecticide through the water perme-ating the cuticle. This might obtain in two ways.First, the continued passage of solvent from the waxphase into the relatively voluminous aqueous phase,by concentrating the solution of diphenylamine insolvent in the wax phase, would increase the per-centage saturation of the insecticide in the aqueousphase in accordance with the partition coefficients ofboth diphenylamine and the solvent. Secondly, the

of diacetone alcohol and carbitol, which are misciblewith water in all proportions, a mixture of equalparts of solvent and water was chosen to represent a100% saturated solution. The figures for these solu-bilities in solutions of the solvents used are given inTable 3.

Although, in the insect, it is impossible to estimatethe percentage saturation of solvent in the waterpermeating the cuticle, an approach to the problemcan be made by considering the factors governing it.As the water is in the substance of the cuticle and isfree from convection currents, the concentration ofsolvent in water in the vicinity of the wax/waterinterface will tend to be relatively high. As the sol-vent diffuses away from the interface, however, thepercentage saturation of solvent in the water willfall progressively as the hypodermis is approached.The degree of saturation of the aqueous phase withsolvent will depend, first, on the partition coefficient

On the penetration of insecticides through the insect cuticle

between wax and water; secondly, onits SWubility in water; and thirdly, on the quantity ofsolvent available. It would, therefore, be expectedthat a solvent with either a low partition coefficientor a high water solubility, and also a high volatilitywould give a solution in the aqueous phase at a lowerpercentage saturation than one with high partitioncoefficient or low water solubility, and with a lowervolatility. It is possible, therefore, to group the sol-vents according to the percentage saturation of theaqueous phase which might be expected to obtain inthe living insect. This has been done in Table 4 and

Table 4

Solvent

QuinolineXylenolm-CresolBenzyl alcoholCyclohexanolMethyl benzoateOctyl alcoholAnilineAnisole4-Methyl-cyclo-

hexanolo-Cresol^-CresolCyclohexanoneDiacetone alcoholCarbitolAcetophenoneBenzonitrileDimethyl-anilineCastor oilo-Dichlorobenzene

Expected %saturation of

solvent inaqueous phase

High

Medium

Low

Nil

Solubility of di-phenylamine inmg./ioo c.c. ofsolution of sol-vent in water

3 09 27-09-0

4 02 06 04-56-56-7

7 7

S-°2-O

10 p17-0

NilNilNilNilNil

a figure for the solubility of diphenylamine in a solu-tion of the solvent in water has been selected fromthe range given in Table 3. It is stressed, however,that no accuracy is claimed for these saturationfigures, except that those solvents in the first groupshould give a higher percentage saturation than thosein the succeeding group and, likewise, with thesecond and third groups.

The high carrier efficiency of the cresols, xylenoland benzyl alcohol compared with the lower carrierefficiency of quinoline, cyclohexanone and cyclo-hexanol has already been considered in the light oftheir rates of penetration through beeswax and theirpartition coefficients, and it was decided that dif-ferences in these two factors alone were insufficientto account for the disparity shown. With regard tothe solubility of diphenylamine in solutions of thesesolvents in water, however, it is seen that whereas the

13cresols, xylenol and benzyl alcohol increase the solu-bility of the insecticide in water, quinoline, cyclo-hexanone and cyclohexanol all depress its solubility.It is clear, therefore, that the solvents of the formergroup should induce a higher rate of diffusion ofdiphenylamine across the aqueous phase than thoseof the latter group.

Among the cresols, £-cresol shows a slightly lowercarrier efficiency than its two isomers and this, too,may be correlated with a lower solubility of di-phenylamine in solution of the solvent in water,though, in this case, the solvent also has a lowerpartition coefficient.

Methyl benzoate has been shown to have a nega-tive carrier efficiency, yet this solvent penetrates bees-wax more rapidly, and has a partition coefficient ofthe same order as £-cresol, where carrier efficiencyis high. The very low solubility of diphenylamine insolutions of methyl benzoate in water may well ex-plain this difference in carrier efficiency.

Thus, although diphenylamine may pass rapidlythrough the wax phase of the cuticle in the presenceof certain solvents, which, in addition, may also beable to cross the wax/water interface, unless thepresence of the solvent in the aqueous phase itself isfavourable to solubility of the insecticide in thatphase, high carrier efficiency will not be displayed bythe solvent in question. A correlation is thus seen toexist between the high carrier efficiency of a solventand those factors influencing diffusion of diphenyl-amine across the aqueous phase of the cuticle.Furthermore, it seems that this is achieved by a pro-gressive increase in the percentage saturation ofdiphenylamine in both wax and aqueous phases,together with an increase in its solubility in theaqueous phase in the presence of the solvent.

It has already been suggested that high volatilityof the solvent, by reducing the quantity of it availablefor passage into the aqueous phase, should play apart in determining carrier efficiency, and a furtherconsideration of the importance of this factor is givenin the next section.

THE EFFECT OF VOLATILITYOF THE SOLVENT

Any factor governing the percentage saturation ofdiphenylamine in the solvent in the wax phase of thecuticle will influence its distribution between the waxand aqueous phases, and will, therefore, affect itsrate of diffusion across the aqueous phase. Evapora-tion of solvent from the powder coating the insect,although concentrating the solution of diphenyl-amine, will not only precipitate the insecticide aftersaturation is reached, thus reducing the total amountin solution, but will also reduce the volume of solventavailable for solution in the aqueous phase. In theserespects, high volatility of the solvent would not beexpected to favour rapid diffusion of the insecticideacross the aqueous phase. This is further compli-

J. E. WEBB and R. A. GREEN

cated by the solubility of diphenylamine in the sol-vents, the figures for which are given in Table 3.Where solubility for diphenylamine is high in rela-tively non-volatile solvents, the low initial percentagesaturation of the insecticide cannot be increased bysubsequent evaporation of the solvent, even thoughit may be increased by diffusion of the solvent intothe aqueous phase. Such factors will limit the carrierefficiency of quinoline, acetophenone, benzonitrile,dimethyl-aniline, and methyl benzoate, all relativelynon-volatile solvents in which the solubility of di-phenylamine is very high. Neither will diffusion ofthe solvent into the aqueous phase be operative inthese cases in improving the initial low' percentagesaturation of diphenylamine, except in the case ofquinoline whose partition coefficient does not pre-clude this taking place. This agrees with the observedcarrier efficiency of these solvents which has beenshown to be either very low or nil.

The factors determining the carrier efficiency of asolvent so far discussed would seem to account formost of the differences in effect observed in the seriesof solvents tested. There are, however, two notableexceptions. There is a difference in carrier efficiency,first between/>-cresol and cyclohexanol, and secondly,between benzyl alcohol and diacetone alcohol, whichcannot be explained by any differences in their pene-tration rates through beeswax, partition coefficients,or solubility of diphenyiamine in solutions of solventin water. From a consideration of these factors alone,cyclohexanol and diacetone alcohol should possess acarrier efficiency approaching that of jp-cresol andbenzyl alcohol (see Table 2). From the list given inTable 2 it will be seen, however, that both cyclo-hexanol (161° C.) and diacetone alcohol (1640 C.)have boiling points considerably lower than thoseof p-cresol (201° C.) and benzyl alcohol (2060 C).The higher volatility of the former solvents maywell be the factor responsible for their relativelylow carrier efficiency.

To verify the effect on carrier efficiency of thevolatility of a solvent with a low boiling point,powders containing (a) i -5% of cyclohexanol and(b) 1-5 % of diacetone alcohol, each with and without0-25% diphenylamine, were tested against keds,maintained in atmospheres saturated with solventvapour. The initial percentage of solvent in thepowder was increased in this instance to allow forevaporation during mixing. It was found that kedstreated with cyclohexanol + insecticide died in 5 hr.and those treated with diacetone alcohol + insecticidein 3-5 hr., while keds dusted with the control powdersdied in 14 and 25 hr., respectively. These figures forsolvent + insecticide show a reduction in the time ofkill, over and above the times obtained in air, of12 hr. in the case of cyclohexanol and 9/5 hr. fordiacetone alcohol (see Table 1). It is evident fromthis experiment that the volatility of a solvent doesplay a major part in determining carrier efficiencywhere evaporation from the powder is rapid.

THE FACTORS GOVERNING SYNEIBETWEEN DIPHENYLAMINE AND

ORGANIC SOLVENTSThe influence of certain physical properties oforganic solvents of diphenylamine on the rate ofpenetration of that insecticide through the insectcuticle has now been studied. Five physical proper-ties of a solvent have been investigated in turn andtheir individual effects linked with observed carrierefficiency. The combined effect of these propertieswhen present to varying degrees in a single solvent,however, has not yet been examined.

In Fig. 2 the relative value of the figures obtainedfor the four major factors on which carrier efficiencyof solvents has been shown to depend is expressed indiagrammatic form. The figures for each factor havebeen grouped into four categories, namely, high,medium, low and nil. These cover as far as possiblenumerical groups. The key to the grouping for eachfactor is appended. The solvents, likewise, are di-vided into high, medium, low and nil. These valuesrefer to their carrier efficiency as already given inTable 1 and, again, divisions have been made wheregrouping of the figures most nearly occurs.

The fifth physical property shown to influencecarrier efficiency, namely, the solubility of diphenyl-amine in the solvent, has been omitted from thediagram. In those solvents with active carrierefficiency the differences in solubility of the insecti-cide, as shown in Table 3, are. seen to be com-paratively slight, with the exception of octyl alcohol,the only case in which this factor may exert a specificinfluence. Further reference to this solvent will bemade below.

The diagram shows that all those solvents withhigh or medium carrier efficiency exhibit either ahigh or medium rating for each of the four factors,with the exception of quinoline, cyclohexanone anddiacetone alcohol, which occur at the end of themedium group. Those solvents in which carrierefficiency is low or nil, on the other hand, show a lowor negative rating for at least one factor in each case.The three solvents with the highest carrier efficiency,o-cresol, m-cresol and xylenol, are alone in pos-sessing a high rating for all four factors. Thus, a highdegree of correlation exists between carrier efficiencyof the solvents and their physical properties. Thissupports the view that the properties examined areat least the principal factors involved.

As the carrier efficiency of a solvent represents thesum of effects of a number of distinct factors, it isessential to consider the importance of each in itsrelationship to the others, rather than as an entity tobe treated separately. Thus, carbitol, and to a lesserdegree aniline, through inability to penetrate theouter wax phase of the cuticle, show a low carrierefficiency in spite of high rating in all other factors.Quinoline, on the other hand, which is deficient onlyin reducing solubility of diphenylamine in water, has

On the penetration of insecticides through the insect cuticle

•c -g SOLVENT

o-Cresol

m-Cresol

Xylenol

Benzyl Alcohol

p-Cresol

Octyl Alcohol

4 - Methyl-cyclohexanol

Oulnoline

Cyclohexanone

Diacetone Alcohol

Cyclohexanol

Acetophenone

Benzonitrile

Aniline

Carbitol

Dimethyl-aniline

Methyl Benzoate

Castor Oil

Anlsole

o-Dichlorobenzene

BC

o •o o• o

ooo

oo

@. o •® ® ^o • o

HIGH MEDIUM LOW NIL

A Rate of penetra- <3om. 3O-6om. >6om. -cotion throughbeeswax

B Partition coeffi- >o.oos 0.005-0.001 <o.ooi nilcient between waxand water

C Solubility of di- >6.omg. 6.0-4.omg. <4-omg. nilphenylamine insolution of sol-vent in water

D Boiling point > I Q O ° C . IQO-I7O°C. <i7o°C. nil(volatility)

Fig. 2.

15a much higher carrier efficiency. Rapid penetrationthrough wax, therefore, is of greater importance indetermining the carrier efficiency of a solvent thanhigh solubility of diphenylamine in a solution of thatsolvent in water.

Anisole and cyclohexanone agree in possessing ahigh rate of penetration through wax and a lowboiling point, but differ in the remaining two factors,anisole having a low partition coefficient and a highsolubility for diphenylamine in solution of the sol-vent in water, while for cyclohexanone the values forthese two factors are reversed. In these two solvents,in which two of the factors have a high rating andtwo a low rating, it is evident that the lower carrierefficiency of anisole is determined by its low partitioncoefficient, which precludes its passage in any quan-tity from the wax phase to the aqueous phase whereits presence would increase the solubility of di-phenylamine in water. Other examples of the effectof negative partition coefficient in a solvent with highrate of penetration through wax are given by aceto-phenone, benzonitrile, dimethyl-aniline, and o-di-chlorobenzene, in all of which carrier efficiency isalso nil. Methyl benzoate is an example of a solventin which negative carrier efficiency is due to thecombined effect of a medium partition coefficientwith low solubility of diphenylamine in solutions ofthe solvent in water.

High volatility of a solvent, as shown by a lowboiling point, limits the effect of other factors. Thecarrier efficiency of diacetone alcohol and cyclo-hexanol is lower than would be anticipated from aconsideration of these other physical properties, but,as will be remembered, these two solvents showedimproved carrier efficiency when operating in atmo-spheres saturated with their vapour. Further,£-cresol and 4-methyl-cyclohexanol, apart from thefact that the latter increases the solubility of di-phenylamine in water to a slightly greater extent,otherwise differ only in volatility. The greatervolatility of 4-methyl-cyclohexanol is, likewise, con-sidered to limit carrier efficiency.

The comparatively high carrier efficiency of octylalcohol.is not readily understood in the light of amedium rating for partition coefficient, solubility ofdiphenylamine in a solution of the solvent, andboiling point. In this instance, the low solubility ofdiphenylamine in the solvent, by providing an initialpercentage saturation of the insecticide much higherthan that obtaining with the other solvents, in whichdiphenylamine is far- more soluble, may be re-sponsible for the high carrier efficiency observed.

It thus appears that certain organic solvents, suchas the cresols, benzyl alcohol, and xylenol, exert ahigh carrier efficiency with diphenylamine by greatlyincreasing its rate of diffusion across the insectcuticle. Diphenylamine alone, when applied to thecuticle as a solid, undoubtedly penetrates by diffusionboth through lipoid elements and, to a limited extent,through the aqueous region. The presence of certain

i 6 J. E. WEBB and R. A. GREEN

solvents appears to favour its diffusion across thecuticle, first, by transporting the insecticide to theinterface between the wax layer and the underlyingaqueous layer (exo- and endo-cuticles); secondly, byconcentrating it at the interface as the solvent ispassing into the aqueous phase and thereby in-creasing its diffusion gradient across the interface;and thirdly, by increasing the solubility of the in-secticide in the aqueous phase, and thus raising itspartition coefficient between the wax phase and theaqueous phase.

The factors which determine in any solvent theefficiency with which this process is carried out arethose influencing the percentage saturation of di-phenylamine in the solvent in the wax phase, therate at which penetration of the wax phase takesplace, the partition coefficient of the solvent betweenthe wax phase and the aqueous phase, and the extentto which the solvent increases the solubility of di-phenylamine in the aqueous phase.

Results show, with three exceptions, acarrier efficiency for each of the mixtures testeSpphisrepresents a reduction in time of kill of keds of atleast twelve hours over and above that of carbitol oraniline when used alone with diphenylamine (seeTable i). Mixtures of equal parts of two'solvents,one of which tends to remain in the wax phase,would not be expected to show a carrier efficiency ashigh as a pure solvent, such as o-cresol. Whereas,with a pure solvent, passage from the wax phase tothe aqueous phase results in the development of ahigh percentage saturation of diphenylamine at theinterface, in the mixture, only half of the totalvolume can pass into the aqueous phase, the otherhalf remaining in the wax and limiting the per-centage saturation of diphenylamine at the interface.Furthermore, in mixtures of solvents, the physicalproperties of each constituent will tend to be modi-fied and the carrier efficiency of the mixture will notbe entirely predictable from data already obtained

Table 5. Diphenylamine with mixed solvents

Solvents

Carbitol + acetophenoneCarbitol + benzonitrileCarbitol + dimethyl-anilineCarbitol + methyl benzoateCarbitol + o-dichlorobenzeneAniline + acetophenoneAniline + benzonitrileAniline + dimethyl-anilineAniline + methyl benzoateAniline + o-dichlorobenzene

Time of death of keds in hr.

Mixed solvents+ insecticide

20>3°

11

9111028

91012

Mixed solventscontrol

28

c. 30

3°28

Rate of penetrationthrough beeswax

at 300 C.

1 hr. 45 min.1 hr. 15 min.

6 min.9 min.5 min.

23 min.> 24 hr.

6 min.11 min.2 min.

THE CARRIER EFFICIENCY OFMIXED SOLVENTS

Carbitol and aniline have been shown to possess allthose physical properties necessary for high carrierefficiency with the exception of ability to penetratethe outer wax layer of the cuticle. Solvents, such asdimethyl-aniline and o-dichlorobenzene, on theother hand, show high rates of penetration throughwax and are deficient in almost all the remainingfactors. It should be possible, therefore, to employmixtures of solvents which would lack none of thefour factors and would display carrier efficiencyhigher than that of either of the constituents.

A series of powders was prepared employing 1 %of mixed solvent, comprising equal parts by volumeof the two solvents tested, and 0-25 % diphenyl-amine, on china clay. These powders, together witha control series containing 1 % of mixed solvent alonewere then tested against keds. The solvent mixturesused, the results obtained, and the rates of penetra-tion of the mixtures through beeswax membranes aregiven in Table 5.

for the pure solvent. It is certain, however, that thehighest value for each property in either constituentcannot be taken to represent that of the mixture.

The addition of acetophenone and benzonitrile tocarbitol, and of benzonitrile to aniline, producedlittle or no increase in carrier efficiency. This can becorrelated with the relatively slow rate of penetrationof these mixtures through beeswax (Table 5). Incontrast, the mixture of aniline and acetophenone,in spite of the presence of the latter solvent, pene-trated beeswax readily and, in consequence, agreeswith the other mixtures in showing an improvedcarrier efficiency. Thus, although others of thefactors, such as partition coefficient and solubility ofdiphenylamine in solution of solvent in water, mayhave become modified by reason of the mixing to-gether of two solvents, it is apparent that, in the caseof those three mixtures showing low or negativecarrier efficiency, the major limiting factor is pene-tration through the wax of the epicuticle.

The results of this experiment thus support thehypothesis that the rate at which a solvent increases

On the penetration of insecticides through the insect cuticledi^Bjon of diphenylamine across the cuticle is de-pernrent on its physical properties. The applicationof a mixture of two solvents, themselves showing nocarrier efficiency, but with complementary physicalproperties, could otherwise hardly be expected toincrease the rate of diffusion of diphenylamine.

EXPERIMENTS W I T H OTHERINSECTICIDES

If, as has been supposed, carrier efficiency is a dif-fusion phenomenon then there appears to be noreason why it should not extend to other insecticidespredominantly oil soluble and incapable of orienta-tion at, and rapid diffusion along, an oil/waterinterface.

Apart from the synergy found to exist betweenhigh boiling tar acids and derris, mentioned in theintroduction to this paper, all the work so far de-scribed on the carrier efficiency of solvents has beencarried out on a single insecticide, diphenylamine.To ascertain whether this synergy can be extendedto other chemically pure insecticides, dixanthogen,oi-nitrostyrene dibromide, and rotenone were chosenfor experiment. Solvents were then selected to coverthe range of carrier efficiency, from high to nil, ob-served with diphenylamine, and powders were pre-pared as before with i % of solvent and 0-25 % ofinsecticide in china clay. Results of trials with theseagainst keds, together with the solubility of theinsecticide both in the solvents and in solutions ofthe solvents in water, are given in Tables 6, 7 and 8.

With all the solvents used, excepting methylbenzoate with cu-nitrostyrene dibromide, and 4-methyl-cyclohexanol with rotenone, a carrier effici-ency similar to that observed with diphenylaminewas obtained for each of the three insecticides. Thetime of death of keds in 7 and 6 hr. when treatedwith rotenone in carbitol and in methyl benzoate,respectively, agrees with a time of death of 6 hr. forkeds treated with rotenone* alone and is, therefore,taken to indicate a negative carrier efficiency.

Although methyl benzoate displays no carrierefficiency with diphenylamine, dixanthogen or rote-none, it shows a medium carrier efficiency with tu-ni-trostyrene dibromide. This may be attributed to theproportionately higher solubility of w-nitrostyrenedibromide in solutions of methyl benzoate in water.

* It was shown by Webb (1945 a) that ground derrisroot dusted on to keds in which all the spiracles weresealed penetrated the cuticle slowly at 30° C. but not atall at 20° C. Derris resin extracted from the root andground was also used in later experiments (Webb, 1945 b),when it was found that penetration of the cuticle wasrapid at 300 C. but very slow at 20° C. The difference inbehaviour of the ground root and the ground resin, at300 C , was attributed to the higher concentration of thetoxic agent present in the latter preparation. It seemsthat this suggestion may be incorrect, since a powdercontaining only 0-25 % of rotenone, and unable to pene-trate the spiracles, caused death of keds in as short a timeas 6 hr. at 300 C. In spite of the high concentration of

With rotenone, 4-methyl-cyclohexanol displayedhigher carrier efficiency than with diphenylamine.This may, perhaps, be explained by the low solu-bility of rotenone in this solvent. For, in this in-stance, the partition coefficient of the insecticidebetween the solvent in the wax phase, on the onehand, and the aqueous phase, on the other, would beexpected to be considerably higher than that of di-phenylamine. This would result in more rapid dif-fusion of the insecticide across the aqueous phase.

In spite of their varying rates of action in theabsence of solvent, the times of kill obtained for allthe insecticides tested with o-cresol, xylenol andbenzyl alcohol are approximately the same. It isevident, therefore, that where insecticides are insolution in solvents it is the rate of penetrationof the solvent through the insect cuticle that largelydetermines the rate of action.

It appears from this experiment that the synergyobserved to exist between diphenylamine and certainorganic solvents can be extended to other insecticides.It does not follow, however, that the carrier efficiencyof a solvent is necessarily similar for all insecticides,since variations both in the solubility of insecticidein solvent and in its solubility in solutions of solventin water are contributory factors.

DISCUSSION

The ability of a solvent to facilitate the passage of aninsecticide across the insect cuticle has been shownto depend on its physical properties. Rapid penetra-tion of the waxes covering the epicuticle enables asolvent to reach the interface between this layer andthe underlying aqueous layer. Not all solvents ofbeeswax, however, increase the rate of diffusion ofthe insecticide across the cuticle; only those whichpossess a high partition coefficient between beeswaxand water. This indicates that the main channels ofdiffusion of the insecticide in the presence of solventare through the hydrophilic elements of the cuticleand not the lipophilic elements. Where a solventpasses readily from the wax phase into the aqueousphase, the concentration of the insecticide in the waxphase will tend to increase and will favour its rapiddiffusion across the interface. Rate of diffusion isfurther increased where the solvent raises the solu-bility of the insecticide in the aqueous phase.

derris resin in ground root, most of this is probably con-tained in cellular elements and is, therefore, unable tocome into contact with the surface of the insect. If theavailable resin is composed solely of particles liberatedfrom parenchyma cells of the root during the process ofgrinding, then, as these are carried freely by air currents,it would be expected that ground root would act ratheras • a respiratory than as a contact insecticide. Whereground resin or rotenone is used, every particle of theinsecticide is free to come into contact with the cuticleand, in a powder containing 0-25 % rotenone, this maybe far greater in amount than the free particles in asample of ground derris root, although in the latter theconcentration of resin may be very much higher.

i8 J. E. WEBB and R. A. GREEN

Table 6. Dixanthogen*

Solvent

o-CresolXylenolBenzyl alcohol4-Methyl-cyclohexanolCarbitolMethyl benzoate

Time of death in hr.of keds treated withsolvent + insecticide

3-o2-2345-o

3027

Solubility ing./ioo g.

solvent at 30°C.

00oo00oo00oo

Solubility in mg./ioo c.c. unsaturatedsolutions of solvent in water at 300 C.

80 % sat.

0-250-50'750-25

24-0c. o-i

60 % sat.

O-75i-oi-oO'5

15-0c. o-i

40 % sat.

o-5O-75i-o0-490

c. o-i

20 % sat.

O-2So-50-750-2530

c. 01

N.B. Keds dusted with 0-25 % dixanthogen remain active after 30 hr.

* Solubility of dixanthogen in water at 300 C. = 0-25 mg./ioo c.c.

Table 7. w-Nitrostyrene dibromide*

Solvent

o-CresolXylenolBenzyl alcohol4-Methyl-cyclohexanolCarbitolMethyl benzoate

Time of death in hr.of keds treated withsolvent + insecticide

30392-25-0

>2813

Solubility ing./ioo g.

solvent at 30° C.

8355551864

135

Solubility in mg./ioo c.c. unsaturatedsolutions of solvent in water at 300 C.

80 % sat.

910109

20030

60 % sat.

13191510804'5

40% sat.

918128

356-5

20 % sat.

811111120V5

N.B. Keds dusted with 025 % oj-nitrostyrene dibromide die in approximately 30 hr.

* Solubility of <u-nitrostyrene dibromide in water at 30° C. = 10.o mg./ioo c.c.

Table 8. Rotenone*

Solvent

o-CresolXylenolBenzyl alcohol4-Methyl-cyclohexanolCarbitolMethyl benzoate

Time, of death in hr.of keds treated withsolvent + insecticide

202-o2-03-07-06-0

Solubility ing./ioo g.

solvent at 30° C.

486221

102-7

18

Solubility in mg./ioo c.c. unsaturatedsolutions of solvent in water at 300 C.

80 % sat.

O-2o-75i"5O-2i'5

<o-i

60 % sat.

o-6i -o

2-50-251-2

< O I

40 % sat.

o-5i-o1 250-25i-o

<o-i

20 % sat.

0-4o-75o-502o-75

<o-i

N.B. Keds dusted with 0-25 % rotenone die in 60 hr.

* Solubility of rotenone in water at 30° C. — c. o-i mg./ioo c.c.

On the penetration of insecticides through the insect cuticle

solvents showing no carrier efficiency werefouncTto have temporary anaesthetic effects on keds,which may have been due to inhalation of the vapouror to rapid penetration of the solvent alone throughthe cuticle. Hurst (1943) has shown that solventsmay penetrate the cuticle by means other than dif-fusion, but it seems inevitable that some diffusionmust take place. Although this may be true for somepure solvents, it is held here that penetration byinsecticides, such as diphenylamine, is governedprimarily by diffusion phenomena. When both thesolvent and the insecticide exert toxic effects, thesemay readily be distinguished when the ked is usedas a test insect as the cessation of heart beat in thisinsect can conveniently be taken as an index ofdeath. Although the small quantities of solvent in-volved were, in some cases, sufficient to produceimmobility, death, when it occurred, took placeslowly. Thus, with benzyl alcohol, for instance, thereis rapid immobility of the ked, but death does notoccur until 21 hr. The addition of dixanthogen, how-ever, with which alone keds are still active after 30 hr.,causes cessation of heart beat in 3-4 hr. Although thetoxicity of the solvent alone may reduce the resistanceof the insect to the insecticide to a small degree, it isnot held that such a difference would be sufficientlylarge to affect the general trend of the results.

In examining the effect of homologous series ofalcohols and fatty acids on blowfly larvae, Hurst(1943) concluded that diffusion phenomena did notaccount for the variations in their rates of penetrationthrough the cuticle. He notes in the series of alcoholsas they are ascended from Ci to C8 that their physicalproperties change uniformly in the same direction aschain length increases. He found that the rate ofpenetration of the alcohols through the cuticle in-creased as the series progressed from Ci to C5,where maximum rate of penetration was observed,and then decreased from C5 to C8. It seems thatthis may well be explained as a diffusion pheno-menon, as it would be expected that only thosealcohols capable not only of penetrating the waxcovering the epicuticle but also of leaving the waxand entering the underlying aqueous medium wouldpenetrate insect cuticle rapidly. This would mostreadily be achieved by the C5 alcohol where thepartition coefficient between wax and water ap-proaches unity. The behaviour of the fatty acids, onthe other hand, does not conform to this arrange-ment and, here, Hurst's suggestion of a strong polarinteraction of fatty acids with the protein com-ponents of the cuticle provides an explanation.

The carrier efficiency shown by mixtures of twosolvents which alone have no carrier efficiency butwhich possess complementary physical propertiessupports a theory based on diffusion rates and parti-tion coefficients. It may also provide an explanationof the effects observed by Hurst on blowfly larvae ofmixtures of ethyl alcohol and kerosene. The failureof ethyl alcohol alone to penetrate the cuticle may bedue ^H> inability to pass through the epicuticle. In

kerosene/ethyl alcohol mixtures, the kerosene dis-solves the outer wax layer of the epicuticle and allowsthe mixture to pass through to the exocuticle, wherediffusion of alcohol through the water permeatingthe cuticle may then readily take place. When thelarva is returned to pure ethyl alcohol, although theouter layers of the wax will have been destroyed, thelipophilic elements comprising the remainder of theepicuticle, together with any wax precipitated fromthe kerosene coating the insect at the time of im-mersion in pure ethyl alcohol, probably form abarrier to continued penetration.

Increase in rate of penetration through the insectcuticle of a contact insecticide in the presence ofsolvents such as the cresols, xylenol and benzylalcohol, is perhaps more readily appreciated whenthose factors influencing penetration of a pure in-secticide, such as rotenone, are considered.

With solid insecticides of low water solubility,penetration will be limited initially by the hardnessof the waxes covering the epicuticle. Wigglesworth(1945) has shown that these waxes vary in hardnessin different forms of insect and this may be a primaryfactor contributing to differences in susceptibility tocontact insecticides. Rhodnius, which is normallycompletely resistant to rotenone, has a wax of aboutthe same degree of hardness as beeswax with a thin"cement layer" on the outer surface. Wigglesworth(1944) has shown, however, that when the cuticle ofRhodnius nymphs is first rubbed with abrasive dusts,death follows the subsequent application of rotenonein under 24 hr. It is possible, therefore, that contactinsecticides penetrate the cuticle principally in thoseregions where there is damage to the epicuticle andsecretion of liquid wax for its repair is actively takingplace. After penetrating the epicuticle the insecticidemay pass by diffusion through the bulk of the lipo-philic elements traversing the cuticle from the hypo-dermis and also, to a limited extent, through thoseelements permeated by water. Here, the pore canals,passing from the hypodermis to the base of the epi-cuticle, may facilitate diffusion.*

Where the insecticide possesses both a lipophilicgroup and a water soluble group in its molecule,Hurst (1943) has suggested that orientation of themolecule between lipophilic and hydrophilic ele-ments at the interface takes place, and that a rapidpenetration of the cuticle occurs by two dimensionaldiffusion along such an interface. With an insecticidenot possessing a molecule of this type diffusion

* Since this paper went to Press it has been shownby Wigglesworth in Rhodnius prolixus ^Hemiptera) andby one of us (J. E. W.) in Eomenacanthus stramineus(Mallophaga) that the pore canals do not end at thebase of the epicuticle, but pass through the epicuticleas far as the outer wax layer. Thus after an insecticidehas penetrated the wax layer and has entered the aqueouscytoplasmic contents of the pore canals its passage intothe body should be facilitated by streaming of the pro-toplasm within the canals.

Further details of this work on the structure of insectcuticle will be published elsewhere.

2 0 J. E. WEBB and R. A. GREEN

through lipophilic and hydrophilic elements musttake place independently and far more slowly.

Where an insect is dusted with pure insecticide agradient of diffusion is set up across the cuticlegoverned by the quantity of insecticide in contactwith the epicuticle. As an excess of dust must beapplied to ensure a maximum area of contact withthe epicuticle, a quantity of insecticide far higherthan that necessary to kill the insect must be used.When, however, the insecticide is in solution in asuitable solvent in the powder, the solvent, by dis-solving the hard outer wax layer and rendering theepicuticle relatively fluid, enables the insecticide topass rapidly by diffusion to the interface between theepicuticle and the exocuticle. Here, a high concen-tration of insecticide is built up by the passage of thesolvent into the water permeating the exocuticle.This continuous layer of insecticide at the interfacerepresents an area of contact far in excess of thatobtaining when the insecticide is present as solidparticles on the surface of the cuticle. The presenceof a high concentration of insecticide at the interfacealso results in a sharp diffusion gradient across theinterface, and insecticide passes into the aqueousphase in quantities governed by its partition co-efficient between the two phases. In addition, wherethe presence of solvent dissolved in the aqueousphase increases the solubility of the insecticide, thenthis partition coefficient rises and diffusion of theinsecticide both across the interface and across theaqueous phase is, in consequence, more rapid.Wigglesworth (1941) has stressed the importance ofpartition coefficients in determining the rate atwhich an insecticide will leave an oily base and enterthe tissues of an insect, and it is clear that this phe-nomenon will be of equal importance in determiningthe rate of penetration through the cuticle itself.

Apart from the effect of a solvent on the diffusionrate of the insecticide across the aqueous phase,diffusion through the lipophilic elements traversingthe cuticle continues and may or may not be in-creased slightly by the presence of the solvent. Theuse of a solvent in an insecticidal dust thus enablesa far higher proportion of the insecticide to reach thetissues of the insect than is possible when the solidinsecticide is used alone, and the concentration ofinsecticide can, therefore, be reduced considerablywithout loss of efficiency.

SUMMARY

1. Using Melophagus ovinus, the sheep ked, astest insect, it was found that certain organic solventsof diphenylamine, such as the cresols, benzyl alcohol

and 4-methyl-cyclohexanol, greatly i ^rate of action of this insecticide. Others, sTrcn ascarbitol and methyl benzoate, gave little or no im-provement in the time of kill. The degree to whicha solvent induces rapid penetration of an insecticideis referred to as its 'carrier efficiency'.

2. The influence of the physical properties of thesolvents on carrier efficiency was investigated. Itwas found that a high carrier efficiency could becorrelated with a high rate of penetration throughbeeswax, a high partition coefficient of the solventbetween beeswax and water and a high solubility ofinsecticide in a solution of the solvent in water. Thevolatility of the solvent and the solubility of insecti-cide in solvent were also contributory factors.

3. Mixtures of two solvents, each showing nocarrier efficiency but together possessing all theessential physical properties, were tested and showeda carrier efficiency considerably higher than that ofeither constituent. This is taken as supporting evi-dence that carrier efficiency depends on certainphysical properties of a solvent.

4. Using a range of solvents shown to exhibitvarious degrees of carrier efficiency with diphenyl-amine, comparable results were obtained withdixanthogen, <u-nitrostyrene dibromide and rotenoneand showed that the synergy could be extended toother insecticides.

5. It is suggested that certain solvents increase therate of penetration of contact insecticides through theinsect cuticle:

(a) By transporting the insecticide through thelipoid elements of the epicuticle to the interface be-tween this layer and the water permeating theexocuticle.

(b) By concentrating the insecticide at the inter-face between the epicuticle and the exocuticle, as the.solvent passes into the exocuticle, and thus increasingthe diffusion gradient of the insecticide across thatinterface.

(c) By increasing the solubility of the insecticidein the water permeating exo- and endo-cuticles andthus, by raising its partition coefficient betweensolvent in the epicuticle and water in the exocuticle,further increasing its rate of diffusion, not onlyacross this interface, but also through exo- andendo-cuticles to the hypodermis.

We wish to express our thanks to Dr V. B.Wigglesworth and Mr J. W. L. Beament for theirhelpful criticism and for permission to quote fromtheir unpublished work; also to Mr D. A. Lambiewho determined the solubilities of the insecticides inthe solvents.

REFERENCESBEAMENT, J. W. L. (1945). J. exp. Biol. 21, 115.CRAUFURD-BENSON, H. J. & MACLEOD, J. (1945). J-Hyg.

Camb. (in the Press).HURST, H. (1943). Trans. Faraday Soc. 39, 390-412.LXUGER, P., MARTIN, H. & MOLLER, P. (1944). Helv.

Mm. Ada, 27, 892—928.

WEBB, J. E. (1945a). Bull. ent. Res. 36, 15.WEBB, J. E. (19456). Proc. zool. Soc. Lond. 115, 218.WIGGLESWORTH, V. B. (1941). Nature, Lond., 147, 116.WIGGLESWORTH, V. B. (1944). Nature, Lond., 153, 493.WIGGLESWORTH, V. B. (1945). J. exp. Biol. 21,