Rimmel, 1865: mười tám lớp mùi, mỗi lớp quanh một mùi mẫu

Rimmel borrowed from painting. There are primary colours from which every other shade is mixed; he argued that there are also primary odours with a perfect type, and that other odours are tied to them more or less closely. Each class in the table therefore has three columns: the class name, the type odour, and the odours he placed in the same class.

The table covers pleasant odours only. Rimmel noted two limits himself: some single odours fit no class, and he gave wintergreen as the example; and the table does not cover the scents produced by blending several classes. He called 18 the smallest number of types he could reduce it to.

The class names follow no single criterion. Some are named after a flower (rose, violet), some after a substance (camphor, musk), some after a quality (balsamic, spice). The type is always a specific material found on a perfumer's shelf of the time. The almond class even includes mirbane, that is nitrobenzene, a synthetic that Rimmel, in chapter XII, calls the most widely used artificial material in perfumery.

The table keeps Rimmel's original English names. The last column links to library pages only where the material clearly corresponds.

Just before the table, Rimmel reviews three older systems. Linnaeus made seven classes, three of them pleasant. Fourcroy made five series, Haller three. Rimmel faults Linnaeus for putting carnation with laurel leaves and saffron with jasmine, pairs he found wholly unlike. His conclusion: the earlier systems were more theoretical than practical, and none grouped odours by their resemblance to one another.

That is a real difference. Linnaeus grouped odours to infer the use of a remedy. Rimmel grouped them to work with them: a compounder needs to know which material stands nearest when one is missing. The column of related odours reads like a list of substitutes or companions. Geranium and rosewood sit next to rose; civet and musk seed sit next to musk.

Chapter XII, on materials, uses a quite different division. Here Rimmel asks not what an odour resembles but where it comes from: the animal series (musk, civet, ambergris), floral, herbal, andropogon (aromatic grasses such as lemongrass and citronella), citrine, spice, ligneous, radical (roots), seminal (seeds), balmy and gummy, fruity, and artificial.

The floral series then held only eight flowers usable in perfumery: jasmine, rose, orange, tuberose, cassie, violet, jonquil and narcissus. The radical series held only orris and vetiver. The fruity series grouped bitter almond, tonka bean and vanilla together, since all are fruits or seeds.

Side by side, the two groupings answer different questions. Vanilla is the type of the balsamic class yet sits in the fruity series. Camphor is the type of an odour class yet sits in the balmy and gummy series. Patchouli belongs to the camphor class by odour but to the herbal series by source. Grouping by source later became the usual order of raw material catalogues; grouping by odour led towards fragrance families.

When Hendrik Zwaardemaker drew up nine odour classes in 1895, he cited Rimmel's series directly. His aromatic class combines the spice, clove, camphor, sandal, citrine, herbal, mint, aniseed and almond series; his fragrant class takes the jasmine, orange flower, tuberose, violet and balsamic series. Zwaardemaker refers to these series by French names (série épicée, camphrée, santalée).

So a perfume maker's table entered a physiology textbook. Zwaardemaker kept Rimmel's grouping but replaced the type odours with compounds: eugenol for the spice group, benzaldehyde for almond, citral and geraniol for citrus and rose. The Zwaardemaker page in this series gives the details.

Publication year: Doty's 2025 review gives 1867, the year of a later printing. The 1865 scans held by Cornell and Wellcome carry the same table.