NEW esters of chrysanthemic acid have been prepared with highly specific insecticidal action. Two of the new compounds are more toxic to some insect species than the natural pyrethrins or allethrin, and are more stable than the natural pyrethrins. Several non-ketonic esters of chrysanthemic acid (I), such as dimethrin (from 2,4-dimethylbenzyl alcohol, Ila) and barthrin (from 6-chloropiperonyl alcohol), have useful insecticidal activity1 but are less toxic to many insect species than are esters from pyrethrolone (Ilia), cinerolone (III6), and allethrolone (IIIc). In 2,4-dimethylbenzyl alcohol (Ila), the 4-methyl group is held by the benzene ring in a steric relation to the H?O bond of the alcohol similar to that between the side-chain CH2X and the HO link in the cycZopentenolones (III). Unsaturated side-chains CHgX in the chrysanthemates of the ketols (III) are important for insecticidal activity2'3, so 4-alkenylbenzyl chrysanthemates were examined for their activity.To adult female Musca domestica L. (house-flies), treated topically with measured drops of solutions in acetone4, 4-allylbenzyl ()-rans-chrysanthemate (from (116)) was approximately four times, and the ()-cis-trans-chrysanthemate twice, as toxic as the mixture of esters of the natural pyrethrins. By this method the mixed pyrethrins, allethrin (from IIIc), and pure pyrethrin I (ref. 5) were equitoxic and 4-allyl-benzyl ()-cis-trans-chrysanthemate (ABC) was seven times more toxic than dimethrin. (()-Cis-trans esters were prepared from the commercially available ethyl ester, which contained approximately 30 per cent cis-isomers; ()-trans-chrysanthemic acid was naturally derived.) But to adult Phaedon cochleariae Fab (mustard beetles), treated similarly, allethrin was twice as toxic as dimethrin and six times more toxic than ABC. (Relative toxicities are expressed as inverse ratios of median lethal doses, LD50.) Methyl groups on the benzene ring of ABC increased tox-icity to mustard beetles considerably. Thus, 4-allyl-2,6-dimethylbenzyl ()-c^s-rans-chrysanthemate (DMABC) (from lie) was twice as toxic as allethrin and half as toxic as the natural pyrethrins mixture to mustard beetles and just significantly more toxic than ABC to house-flies. (New compounds are protected by British Patent applications 10701/63, 10702/63 and 42715/63, and corresponding foreign applications assigned to the National Research Development Corporation.) 4-Allyl-2,6-dimethylbenzyl and 4-allylbenzyl ()-ra?is-chrysanthemates were more toxic to house-flies than was any other pyrethrin-like ester, natural or synthetic; also with other insect species (for example, Tribolium castaneum (Herbst.), Periplaneta americana (L.), Aphis fabae, Scop., and the red spider mite Tetranychus telarius (L.), both in laboratory tests and conditions nearer to practical use, the 2,6-dimethyl compound (DMABC) was as toxic or more so than allethrin. The insecticidal activity of a film of DMABC, irradiated with ultra-violet light, persisted longer than that of comparable deposits of allethrin or the natural pyrethrins. The relationship between toxicity to house-flies and mustard beetles and chemical structure in esters related to the allyl compounds already mentioned was examined in detail. To both species, 4-allylbenzyl esters were more toxic than 4-n-propyl compounds; similarly, allethrin was more active than its n-propyl analogue2. 4-Allylbenzyl esters were about ten times more active than 3-allyl compounds and 2-allylbenzyl esters gave no kill at the largest doses. This showed that it was important to maintain the appropriate steric relation between the unsaturated side-chain of the alcohol and the linkage to the acid portion of the molecule. Esters in which the side-chain double bond was conjugated with the benzene ring (from 4-vinyl and 4-propenylbenzyl alcohols, Ud and lie, respectively) did not kill mustard beetles and were only slightly toxic to house-flies. The ester from the enyne alcohol (IV) (ref. 6) in which an acetylenic link maintained a steric relation between the hydroxyl and allyl groups similar to that in 4-allylbenzyl alcohol was not toxic either to mustard beetles or house-flies, so this steric relation on its own was insufficient to give insecti-cidal activity.Most of the alcohols formed more toxic chrysanthemates when the hydrogen atoms ortho to the CH2OH group (II, E1 and B5) were replaced with methyl groups. The 4-allyl-2,6-dimethyl ester was the most toxic compound to mustard beetles, but to both species the ester from 4-allyltetramethylbenzyl alcohol (II/) was less toxic than the 2,6-dimethyl compound. The 4-Zrans-crotylbenzyl-alcohol (Ilg) gave esters less toxic to both species than esters of 4-allylbenzyl alcohol (116) but, to mustard beetles only, the ester from 4-(2'-methallyl) benzyl alcohol (Ilh) was better than that from 4-allylbenzyl alcohol. The ester from the readily accessible 4-rans-sorbylbenzyl alcohol (Hi) was only one-twentieth as toxic as the 4-allyl compound to house-flies and one-tenth as toxic to mustard beetles. 2-, 3- or 4-Allyloxybenzyl chrysanthemates had little insecticidal activity, less than that of the corresponding methoxy compounds. Free hydroxyl groups suppressed toxicity; the ester from the alcohol (V), easily made from O-allyl kojic acid7, was not insecticidal. Methoxyl groups at any position on the ring decreased the toxicity of otherwise active compounds. Thus, Barthel et al.8 made 3-allyl-4-methoxy- and 4-ethoxybenzyl chrysanthemates, and we made 3-allyl-4-methoxy- and 4-allyl-3-methoxy-benzyl chrysanthemates; but all were much less effective than the related unmethoxylated allyl benzyl esters.The high toxicity of the 4-alkenylbenzyl chrysanthemates led to an investigation of methylbenzyl compounds. The 2,4-, 3,4- and 2,5-dimethyl esters had already been tested1. 2,4,6-Trimethylbenzyl ()-cis-rcms-chrysanthe-mate (from II?') was approximately one-third as toxic as DMABC to house-flies and about twice as toxic as dimethrin to house-flies and mustard beetles. Penta-methylbenzyl ()-CM-2rans-chrysanthemate (from Ilk) was significantly more toxic than the 2,4,6-trimethyl ester to mustard beetles, but less active to house-flies. 2,4,6-Triethyl benzyl alcohol (III) gave a non-toxic ester. Other dimethyl esters were less toxic than the 2,4- and 3,4-dimethyl compounds. 2,4- and 3,4-Dichlorobenzyl chrysanthemates were less toxic than the corresponding dimethyl compounds. Thus, those benzyl alcohols the structures of which relate them most closely to pyrethrolone, cinerolone, etc. (Ill), give the most toxic chrysanthemates. 4-Allylphenyl and 4-allyl-2,6-dimethylphenyl chrysanthemates, which have some structural features in common with the benzyl esters but differ stereochemically, had no insecticidal activity; this indicated that there must be a specific and appropriate stereochemical relationship between the acidic and alcoholic parts of the esters. The 4-allyl-2,6-dimethylanilide of ()-c*'s-Zrans-chrysanthemic acid differs little from DMABC stereochemically but obviously lacks the requisite chemical properties, for it was not toxic.Molecular models show that, when fche allyl groups in allethrin and in 4-allyl-2,6-dimethylbenzyl chrysanthemates are orientated similarly relative to the chry-santhemic acid parts of the molecules, one of the 2,6-dimethyl groups in the benzyl ester occupies a position equivalent to that of the methyl group on the cyclo-pentenolone ring; a possible explanation for the lack of toxicity of 2,4,6-triethylbenzyl chrysanthemate is that here there is no room for an ethyl group to be similarly accommodated. Additional evidence that the high insecticidal activity of the 4-allyl and 4-allyl-2,6-dimethyl-benzyl esters is associated with their similarity in molecular shape to the natural esters and allethrin is that ()-trans-chrysanthemic acid gives esters approximately twice as toxic to house-flies as those from the ()-cis-trans-acid, and that, to house-flies and mustard beetles, the 2,4,6-trimethylbenzyl ester of ()-rans-chrysanthemic acid is much more active than that from (? )-nms-chrysanthemic acid. Analogous results were obtained with c^/cfopenteno-lone esters of (), ( ?), and ()-cis-trans acids2'3. The results support an earlier suggestion9 that the insecticidal activity of the pyrethrins and related compounds depends partly on the ci/cfopentenolone ring acting as a planar activating nucleus to hold the unsaturated side-chain in an appropriate stereochemical relationship to the alcoholic link and thence to the acid. In these benzyl esters the benzene ring fulfils this role. Neither ABC nor DMABC was well synergized by piperonyl butoxide in activity against house-flies.Details of the synthesis of the new compounds and of the biological data will be published. This work formed part of a project supported by the National Research Development Corporation. We thank Mr. F. P. W. Winteringhara for a helpful discussion at an early stage in this work and Dr. C. Potter for his help and advice.