384 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Figure 9 (b) Integrams of oil of lemongrass in the neighbourhood of the citral peaks, with isopropyl benzoate added as internal standard.--High magnification. (Column temperature 100øC.) oil in the neighbourhood of the two citral peaks. The internal standard, isopropyl benzoate, will occupy a position on the level section of base-line preceding them. Fig. 9 shows an integram of the three compounds at low magnification, also a similar integram at higher magnification. Evaluation of the citrals depends upon linear measurements of the total traverse of the pen in the three groups. Quantitative gas-liquid chromatography analysis is sometimes less simple in practice than in theory, because with certain essential oils it is not easy

GAS-LIQUID CHROMATOGRAPHY AND THE PERFUMER 385 to comply with the conditions of separation previously described. However, when gas-liquid chromatography can be used, it is capable of yielding more realistic results than those obtained by chemical analysis. The principle of completely isolating the component to be measured is sound from the analytical standpoint, and we may anticipate further advances in gas-liquid chromatography analysis, particularly if a more satisfactory measuring device can be discovered than the detectors now available. GAS-LIQUID CHROMATOGRAPHY AND IDENTIFICATION OF •ESSENTIAL OILS The odour of an essential oil depends upon the plant from which it is derived, but it may be modified by the environment in which the plant grew and by how, and when, the oil was extracted. On the whole, oils from species within the same genus are more likely to resemble each other than oils of plants from different genera. Oils from related hybrids, varieties or strains will tend to differ to a lesser degree. Variations due to environ- ment and to deta/ls of manufacturing processes are usually the least marked, but perfumers can distinguish and recognise consistent odour categories, even of this order, without great difficulty. This has led to a system of classification, based upon botanical and geographical considerations, which is used for trade in essential oils. Consistent differences of odour, however small, must be due to corre- spondingly consistent divergences in composition. These underlie the commonly accepted physical and chemical specifications for essential oils. Unfortunately, such specifications are almost valueless by themselves for checking the identity of essential oils because they yield so little information about details of composition. Gas-liquid chromatography, being a separa- tion technique, can provide more reliable evidence of identity, even though the compounds which give rise to peaks on chromatograms are not all known. It remains to demonstrate what degrees of difference can be detected with certainty by gas-liquid chromatography. My experience is that chromatograms made with good standard instru- ments under properly chosen conditions may be relied upon to distinguish between any two recognised types of essential oil, even those distilled from the same sort of plant in distinct geographical areas. Variations among samples from a single location are also detectable, though they have proved to be less characteristic and of a lower order. Previous illustrations (Figs. 1, 3, 6 and 7) have shown the very distinct chromatographic patterns produced by unrelated essential oils like citronella, rosewood, spearmint and lavendin. Some similarities and differences between oils from more closely related plants are shown below. I have chosen certain oils from the genus Lavandula, because having collected