SIMPLE FLORAL PERFUMES ready acceptance, although this represents the elements of the Freesia perfume but I feel that I can make my point more easily by drawing upon the analogy of colour. By this I do not imply any actual linking of a colour with the name and odour of a blossom. It is, of course, a matter of common knowledge that from the intermixing of pigments or solutions representing the three prim- ary colours--yellow, red and blue, the secondary shades of orange, green and violet are produced but little attention seems to have been paid to the admixture of the primaries and secondaries, the num- ber of chromatic units from these wkich are utilised in the evolution of the three tertiaries, or even the names and actual shades of these triads. However, it is found that: Yellow plus violet = citrine, a dull brownish green. Red plus sea-green ---- russet, a dull rust-brown. Blue plus orange -- olive, a dull bluish green-grey. These chromatic additions are not haphazard, but are based upon the arithmetical proportions of the prim- aries necessary to produce, as a final end-product, the theoretical "optical black." Hence the exact proportions of the basic hues can be calculated and graphs constructed, from which it is a simple matter to compute the admixture required for the produc- tion of such between-shades as mandarin-orange or turquoise-blue. As will be seen from Table III, a very important point emerges, namely, that there are no less than four alternative routes to each of the three tertiary shades. Incident- ally, these are dull and sombre hues and a sharp contrast to the bright secondaries and luminous primaries. In our elementary simulations of the six Basic Blossom Odours-- jasmin, neroli and rose, violet, lilac and lily, we select various propor- tions of the following group of isolates and synthetics: phenylethyl alcohol, linalol, terpineol, hydroxy- citronellal, geraniol, citronellol, TABLE III Citrine Russel Olive Yellow Orange Leaf-green Tangerin l•ed Orange Magenta Mandarin Blue Sea-green Turquoise Purple plus ,, ,, violet sea-green russet olive sea-green violet citrine olive orange violet russet citrine I87

JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS methyl anthranilate, benzyl acetate and ionone, of which at least two items must be used for each named floral. With portions of these various substrata common to several basic odours, the colour analogy explains (at least to my satisfaction) certain aspects of the frequently encountered phenomena of odours "cancelling-out," the pro- duction of toneless nondescripts, and the devious approaches that may be made towards achieving a specific type of perfume. In fact, I think that the alternative routes to the tertiary colours provide a striking parallel with the matching ef a perfume, whereby two or more workers using different raw materials will arrive at an end-point, not necessarily exact, but fundamentally in agreement, between themselves and the original. SIMPLE FLORAL ELEMENTS We have seen from Table I that the members of this group of ancil- lary or secondary blossom odours are aggregates, built upon a substructure of primary fragrances and, in certain instances, usage is made of the judici- ous inclusion of strong osmophores. It is, therefore, at least instructive to attempt an investigation into the behaviour upon alignment of these perfumery "radicals" in order to learn what is evolved osmically, the extent of the permutations and, finally, whether or not there emerges from these combinations any regular aromatic grading or progression. For the purpose of the construction of Table IV, it is assumed that we have available the six "radical per- fumes," namely: (3) Jasmin (5) Neroli (9) Rose (warm) (13) Violet (17) Lilac (21) Lily (cold) of such a quality that these odours would be recognised without hesita- tion by any audience these per- fumes being approximately of the same odour strength and combined at first in equal proportions. The numerical indices are retained as a check against repetition therefore in the first column of Table IV we have observed the mechanism of the permutations. It is, however, of vital importance to pause here and first examine the outcome of the initial interblending which gives the diads, observing in passing that we have no descriptive nomenclature in perfumery to name the outcome of such simple odour combinations, although we know, from those who possess any measure of artistic feeling, that chords in music and colour-blending with only two components are for the most part •esthetically unsatisfactory. Exactly how far this analogy can be extended to perfumery depends, I believe, on what can be deduced from Table IV. From this it will be seen that from six components, fifteen combinations are possible, but it is probaNe that the only useful con- clusion which can be drawn is that some couplings (for instance, jasmin- violet, neroli-violet and rose-violet) are harsh and contrasting, and the violet element seems to be in more 188