It is shown that in the assay of pyrethrum flowers the customary use of warm solvent (soxhlet) gives erroneously high results for pyrcthrin 11, especially when the higher-boiling grades of ligroin arc used. Cold exvaction gives lower results for pyrethrin 11, and largely eljminrtcs the variations in the results obtained with diffcrent grades of ligroin. It is also shown that cold ligroin extract:c all the inse:ticidally-active pyrethrics, and that the additional pyrcthrins exrncted by hot ligroin are non-toxic to house-flies. Chloroform. is"shown to extract a particularly large quantity of such non-toxic pyrcthrins." The use of cold ligroin both for the assay, and for the commercid extraction, of pyrethrum flowers is therefore recommended; and it is also suggested that the flowers should be standardized only on their content of pyrethrin I.Pyrethrum flowers are usually analysed by the Seil'. method, or, less frequently, by the mercury r e d~c t i o n~.~ method. In a recent paper (Mitchell, Tresadern and Wood)6 the accuracy of these methods was studied, using the pure chrysanthemum acids for this purpose. It was confirmed that the mercury reduction method gave satisfactory results, provided that it was carried out under carefully standardized conditions; a small modification, designed to improve the results for pyrethrin 11, was proposed. It was also confirmed that the Seil method gave low results for pyrethrin I, and it was shown that these low results were not caused by the presence of mineral acid, as had previously been supposed, but were due to D temperature effect in the steam distillation. Experiments suggested that the .apparent loss of chrysanthemum monocarboxylic acid was due, not to decarboxylation, but to direct hydration to a saturated hydroxy-acid. This latter was much less readily voIatile in steam than chrysanthemum monocarboxylic add, and was also sparingIy soIubIe in Iigroin. A modified Seil method which eliminated these low results was described. A series of comparative analyses of commercial pyrethrum extracts confirmed the accuracy of the above results; but the methods were not applied to the analysis of pyrethrum flowers.In the Seil, mercury reduction, and most of the earlier methods which have been proposed for the assay of pyrethrum flowers, the crushed drug is extracted with ligroin in a soxhlet, or other continuous extractor. In such extractors the drug is extracted with the o u~y l n solvent at a temperature just below its boiling point. In commercial extraction of the flowers, however, cold solvents are generally employed, and the yields of total pyrethrins are somewhat lower than those calculated from the laboratory assays using the soxhlet method. Thus, large-scale experiments on Kenya pyrethrum flowers, the authors have found that cold ligroin rarely extracts more than 95% of the total pyrethrins on the basis of the laboratory test (Seil, or . mercury reduction). Using old flowers, or those appearing to have been carelessly dried or stored, the yields have been only 8 0 %~ or even lo...