RHODOLOGIA: A discourse on roses, and the odour of rose (1844)

Rose Perfumes
John Charles Sawer

THE great perfection to which rose growers (amateur and professional) in England, Belgium and France have now brought this magnificent flower is indeed marvellous; the number of distinct varieties named and cultivated is over 7,000, and with every horticultural exhibition this number is increased to such an extent as to justify the existence in England, France, and Germany of periodical publications exclusively consecrated to the "Queen of Flowers." In fact, there seems to be no end to what rose growers can do in the way of novelty. A leading London paper, commenting on a show of the National Rose Society, gives a glowing description of such novelties as the creamy, claret-tipped "Princess of Wales," the shell-white "Hon. Edith Gifford," the ivory Tea-rose "Souvenir de S. A. Prince," the pink-copper "Lady Penzance," and the exquisite double-yellow "Rosa Sulphurea," and remarks:—"What immortality can be more charming than that which is thus conferred by coupling some happy name with a new favourite in the world of roses."

Many a London house during the month of June, at the festal hour, is converted into a perfect Gulistan with this most delicate and most delicious of ornamentation. Some of them almost rival the lavish prodigality of ancient times. Cleopatra, according to Athenaeus, more than once covered the floor of her dining-hall with roses a cubit thick, and the Emperor Nero expended in one feast as much as £30,000 in roses alone; while everybody remembers the gorgeous canvas upon which Mr. Alma Tadema delineated Heliogabalus overwhelming his guests with an avalanche of roses showered from above. But the ancient rose lovers never possessed or saw blooms of their favourite flower half so perfect as those which are to-day quite common. They knew next to nothing of the amazing modern variety in the Queen of the Garden—white roses, blush roses, pink roses, pied roses, crimson roses, scarlet roses, vermilion, claret, carmine, purple-striped, and black roses, and roses of burning bronze and silver satin. Many a cottager has to-day upon his wall a "Cloth of Gold" rose such as the Caesars could not have bought for a thousand sesterces; and, beyond question, the finest roses in the world are to be seen in London during the months of June and July.

* London, 13th Edition.

Well is it said by the Dean of Rochester, in his perfectly charming "Book about Roses,"* that "he who would have beautiful flowers in his garden must have them in his heart." In this work, abounding in information concerning the culture of the choicest specimens of roses, are also to be found many anecdotes of those who have loved roses, and been successful with them (not by any means always the rich or the learned). The very reverend author tells a story of a Nottingham mechanic whose wife, with his entire consent, took a blanket from their insufficiently supplied bed to keep the frost out of the little green-house where they had their rose stocks. He tells us also how, at a June show of the Manchester Botanical Society, £1,100 was paid by the working-classes in single shillings to see the glory of the prize blossoms. Delightful indeed is the enthusiasm to which such a lover of the imperial flower attains! If once, indeed, a man is bitten with the passion for cultivating high-bred roses, anything like a cure of it is hopeless. The Dean himself confesses that, though he commenced with a dozen plants only (the first which enslaved him being the "D'Aguesseau Gallica"), he did not become contented when he owned 5,000. Master of many rosy legions, he frankly declares that if he had Nottinghamshire full of roses he would still desire to have Derbyshire for a budding-ground. If a grave and pious dignitary of the Church can fall into the splendid horticultural avarice indicated here, laymen may well be warned not to abandon themselves too lightly to the fascinations of the Rêve d'Or, La Belle Lyonnaise, or Souvenir de Malmaison.

Notwithstanding this great enthusiasm, it seems that both cultivators and admirers of this flower concentrate their whole attention on the development of colour, form, and size, and pay little heed to the great variety of perfumes generated in the beautiful petals. These different odours (although, of course, blended with that of rose) are distinctly recognisable, and are in some cases very marked.

The organs of the sense of smell can be trained to the appreciation of perfumes, especially by young persons, as easily as the palate can be trained for business purposes to the tasting of the flavours of wines, tea, or coffee. Of course a taster, sampler, or evaluer of such beverages must naturally be possessed of a finely developed nervous susceptibility to the slight variations occurring in every sample which comes under his notice. Such natural perfection of susceptibility is not common, although many possess the gift without being quite aware of it: they may not be in the tea trade or the wine trade, or in any trade at all; and so the gift is not trained, or even appreciated. To those who are not constantly occupied in the culture of the rose, it may seem that one rose is very much the same as another, and excepting a few variations in colour and habit of growth, there is very little difference distinguishable. Probably many persons would never believe that there are not only roses perfectly devoid of odour, but there are some which stink. There are experienced gardeners who can name many varieties of rose in the dark: this means that the perfume of roses is very varied, and that no two varieties possess the same odour. What is called the pure odour of rose is unique, undefinable, incomparable. It is, in fact, a type, and no imitation can approach it. It may be best represented by the R. centifolia and R. Damascena.

*P. J. Redoute, "Les Roses" Paris, 1817 and following years. Text by Thory; i. p.25, t. i. p.37, t.7; p.77, t. 26; p.79, t. 27; and p. l11, t. 40.
†Boissier, Flora orientalis, 1872, ii. p. 676.
‡Redoute, as above, i. p. 137. t. 53; i. p. 107, t. 38; and i. p. 121, t. 45. Also Lawrence, "Collection of Roses from Nature," London, 1799, t. 38.
§Wiggers and Husemann, Jahresbericht, 1867, p. 350. ||Dendrologie, 1869, i. p. 250.
¶Journal of Botany, Jan., 1875.

R. centifolia, Lin. Sp. 704* (R. provincialis Miller), is found in the wild state, with single flowers, in the eastern parts of the Caucasus.† Under cultivation its flowers are more or less double, and innumerable varieties of it are propagated and grown in every temperate region of the globe. In English gardens it is grown as the Cabbage Rose or Provence Rose. In the South of France it is cultivated commercially.

The second type of a pure rose odour is represented by the plant grown in Bulgaria; this is a variety of the Rosa Damascena, Miller.‡ Living specimens were sent from Constantinople by Professor Baur; these flowered at Tübingen, and were examined and identified by Hugo Mohl.§ R. Damascena is not known to exist in the uncultivated state. Koch|| affirms that at a remote period it was brought from Italy into northern countries. Baker¶ considers R. Damascena to be merely a variety of R. Gallica which was distributed from France to Mesopotamia.

For description and classification of roses the following works may also be referred to:— Guillemeau, Histoire Naturelle de la Rose, 1800; J. P. Buchoz, Monographie de la Rose, Paris, 1804; Raw, Enumeratio Rosarum, Wirceburg, 1816; Pronville, Nomenclature Raisonnée des espèces, variétés, and sous-variétés du genre Rosier, Paris, 1818; John Lindley, Rosarum Monographia, London, 1820; L. Trattinick, Rosacearum Monographia, Vienna, 1823; Hariot, Notes, pour servir à l'historie des classifications dans les espèces du genre Rosa; Thèse de l'Ecole de Pharm. de Paris, 1832; Boitard, Manuel complet de l'Amateur de Roses, Paris, 1836; Deséglise, Catalogue raisonné des espèces du genre Rosier, Paris, 1877; Chereau, Examen des roses officinales; Journal de Pharmacie, xii. p. 436; S. Reynolds Hole, A Book about Roses, 12th edition, London, 1892. Good lists of new varieties are also to be found in the catalogues of Louis Van Houtte, of Ghent.

* The Gardener's Monthly and Horticulturist, xxviii. 1886, p. 249, Philadelphia.
† Bull. de la Soc. Bot. de France, Fevrier, 1889.

The first attempt to classify the varieties of perfume in the rose was made as recently as 1886, by an anonymous writer in an American horticultural journal.* The subject was further studied by Dr. Blondel in 1889,† and reprinted in his admirable treatise, "Les produits odorants des Rosiers," Paris, 1889.

These observers record the fact that the odour of tea is not perceptible in the so-called "Tea-roses," indeed many Tea-roses are odourless, such as the Melanie Souppert, Marie Guillot, Marie Caroline de Sertoux, and Triomphe de Milan. Some Tea-roses possess very delicate fruity odours; for instance, the R. Socrates has the odour of peach; Elizabeth Barbenzien, odour of melon; Isabelle Narbonnand, odour of violet; Safrano, odour of pinks. Some of the Tea-roses possess odours which are far removed from that of pure rose, but yet are particularly sweet and fruity, recalling that of the raspberry, such as the Maréchal Niel amongst the yellow, Goubault amongst the red, and Madame Bravy amongst the creamy-white varieties. There are Tea-roses such as Gloire de Dijon, whose perfume is so sweet and subtle, that any definition or comparison is quite impossible.

The majority of Noisette roses are inodorous, but the variety known as Unique jaune has a faint odour of hyacinth, and some, such as Celestine Forestier, Claire Carnot, and Earl of Eldon have a slight fruity odour; this is particularly noticeable in the Desprez.

‡ Sowerby, Eng. Bot. Sup. t. 2653.
§ Bot. Reg. t. 424.

Amongst roses possessing fruity odours, R. bracteata and R. Macartnea are credited with possessing the odour of apricots, and Souveraine (a cross between a Tea-rose and R. centifolia) has exactly the odour of melons; in most cases, however, the fruit odours are blended with one somewhat analagous to that of bananas or quinces. The odour of mignonette is noticeable in R. canina (especially when the petals are developed in a somewhat darker shade than usual), the red R. saepium, and R. Alpina, Lin. (This last is now employed in the perfume industry specially by reason of its mignonette-like fragrance.)§

The odour of violet, besides being developed in R. Isabelle Nabonnand, above-mentioned, is also very noticeable in R. Banksia alba. (Banksia lutea is perfectly inodorous.) The odour of pinks is also exhaled by the R. caryophyllea, R. Ripartii, Deséglise, and R. moschata, Miller. This last, however, does not possess the odour of musk, as one might suppose by the name. The musk odour is only clearly developed in the R. Salet, a hybrid climbing Moss-rose, bright red with a pale centre.

* This is a very remarkable coincidence; the coriander plant and its unripe fruit both possess the intensely disgusting odour of bugs, but the fruit at complete maturity, acquires the pure odour of coriander. The nature of the chemical action producing this modification in the odour is not understood. An equally curious example of the transmutation or development of odours is instanced in the Acacia Farnesiana and the Tritelia uniflora, in both of which plants are associated the odours of violet and of garlic.

The flowers of most of the varieties of R. rubiginosa are inodorous, but those of R. platyacantha and R. Capucine (R. Eglanteria, Lin.), (especially the variety bicolor, Jacq.) belonging to this group, develop an odour of bugs and coriander. The same may be said of R. Beggeriana (R. coriosma, Decsn.).*

In further illustration of the capricious nature of this perfume, and the extraordinary complexity of its forms, it is stated that not only in the whole list of roses are there no two which develop precisely the same odour, but that in the same species, and even on the same plant, there are not found two flowers absolutely identical in odour,—even yet further, that it is a well known fact amongst rose growers that at different times in the day an individual flower will emit a different perfume.

As a rule the red roses are more odoriferous than the white.

Cut roses placed in a vase diffuse their fragrance more powerfully than when growing on the plant.

It has also been noticed that roses flowering under glass give off a greater amount of perfume than those cultivated in the open air; the reason of this is obscure, but it is perfectly certain that under no conditions is the odour fully developed except in very hot climates, where the power of the sun affords the maximum benefit of light and heat.

The flowers of Rosa gallica (which are used officinally) are but feebly odoriferous when freshly gathered; their perfume develops gradually in the process of desiccation, while that of the Damask-rose is almost destroyed by drying.

In Bulgaria the flowers grown for the distillation of the otto are gathered before they fully expand, and a little before sunrise. Were they gathered later in the day, when fully expanded by the heat, the perfume would be weaker, and not so sweet, and the resulting essence would be of less value.

It has been noticed that previous to a storm, or atmospheric disturbance, the odour of the rose seems strangely increased; this may be by reason of the oxidising influence of the ozone in the atmosphere, or it may be that our perceptive faculties are sharpened at such moments.

Rosa centifolia. Transverse section of a sepal, showing the pedicellate glands on surface of epidermis, and fibro-vascular bundle within. (After Blondel.)

Rosa Centifolia, var. muscosa (Moss-rose).
Fig. 1.—Single tooth of a foliole (x 80) showing the pedicellate and branched glands on the margins of the tooth, and non-existence of same at the summit. The dark coloured cells at the summit of the tooth and at the base of the glands are cells secreting tannin.
Fig. 2.—A single gland of the above, isolated and greatly magnified, showing the tannin-secreting cells at the base rendered opaque by their contents, and the upper cuticle raised by the accumulation of oleo-resin at the summit.
Fig. 3.—Transverse section of a foliole. (This drawing is inadvertantly made upside-down.) It shows the pedicellate, oleo-resin secreting glands on the lower surface, and the structural difference of the upper and lower parenchymatose tissue (x 300).
Fig. 4.—Base of a petiole. showing the stipules, and how the entire formation is clothed with pedicellate glands.
Figs. 5, 6.—Glandulous structures occurring in large quantities amongst the prickles and young thorns.


* A similar odour has been noticed in phenylnitro-ethylene chloride, which can be prepared by passing chlorine into a cooled solution of phenylnitro-ethylene in chloroform. On the evaporation of the latter it remains as a thick oil, which has a penetrating odour, resembling, when dilute, that of pippins. On standing for some time, large lustrous crystals are deposited, which are extremely soluble in ether and chloroform, and are again left on evaporation as an oil, which solidifies, when placed in contact with a. fragment of the original crystals, to a mass, which melts at 30°. This odour of pippins, akin to that of sweetbriar, is noticeable in the flowers of Agrimonia eupatorium, and in all parts of the Agrimonia odorata.

In estimating the quality of the odour of a rose, care should be taken in handling the stalk, the calyx, or any green part, as a very slight friction breaks the pedicellate glands which abound on these parts, and in which is secreted an oil or glutinous oleoresin of powerful odour, totally different in character to that developed in the microscopic cells of the petals. In some roses, such as the Moss-rose, these glandular formations attain quite a prodigious size, developing very rapidly on the young shoots, even overtopping the thorns. In some varieties the odour of the glandular secretion is very rank, and has been described as hot, peppery, clove-like and terebenaceous, but the word terebenaceous is misplaced and misleading; the odour may remind of the resins contained in certain woods, but it is generally more balsamic, like the odours of some varieties of pelargonium and geranium leaves. In some varieties of the group Rosa villosa the odour approaches that of olibanum and myrrh. In R. Brunonii, Lindl. (a moss variety of R. moschata, Miller), the hispido-glandulous system which thickly covers the pedicels and sepals, develops a fine odour of pinks. Sometimes this odour is fruity, as in many varieties of R. rubiginosa, Lin. (the "Sweetbriar") and is developed to such an extent as to be disengaged spontaneously, especially on a warm day, and by the gentle action of the wind after a light rain. All the green parts, especially the lower part of the leaves, contain innumerable oil glands, which, being broken by the slightest friction, exhale an agreeable odour which has been very rightly compared to that of an apple called "Pomme Reinnette," or in English "Pippin." The glands of R. micrantha, Smith; R. graveolens, Pers, and R. glutinosa, Sibth. secrete a very similar perfume. The composition of the body contained in these glands is apparently unstudied and little understood, but its odour is suggestive of valerianale of amyl, which is now prepared on a large scale for use in flavouring sweetmeats and liqueurs with the flavour of apples.*

The leaves of R. lutea, Dalech (R. Eglanteria, Lin.), (known also as R. Capucine) possess an odour which is even finer, recalling that of jasmin.

Rosa centifolia.
Fig. 1.—Transverse section of a petal (x 500,) after treatment with osmic acid; showing the reduction of the osmium in the vessels on both surfaces.
Fig. 2.—Fragment of section of a petal at the part where it is attached to the calyx, showing (x500) the non-papillary cells of the upper epidermis filled with essential oil.
Fig. 3.—Fragment of upper epidermis of a petal, dissected under the microscope, showing the formation of papillary cells.
Fig. 4.—A single cell (x 1000).
Fig. 5.—Longitudinal section of same, and Fig. 6 upper view of same.
Fig. 7.—Transverse section of filament of stamen, after treatment with osmic acid, showing the essential oil contained in the epidermal cells (x 200).

The glandular structures of the green parts of the rose are minutely described by Dr. Blondel in his work above cited, he also gives details of the cellular arrangement of the epidermis of the petals, and illustrates the appearance of the same under the microscope (the accompanying illustrations are copied therefrom). He found on examining the anatomical structure of a. transverse section, that both sides of the petal are equally odoriferous, that is to say, the otto is secreted in epidermal cells on both surfaces, the upper ones being papillary, and the lower ones cubic, in form. To exhibit the presence of the oil, he employed a reagent of great sensitiveness, viz., a solution of osmic acid (Os O4), 1 in 200. The section of a fresh petal is plunged in this bath for about twenty seconds, then carefully washed in distilled water and mounted in glycerine. Thus prepared, and examined under the microscope, the cells of both surfaces of the epidermis appear to be filled with an intense bluish-black pigment, due to the reduction of osmic acid by the essential oil and deposition of the osmium. The mass of osmium thus deposited is quite homogeneous in appearance, and this is thought to indicate that the oil exists in a very minute state of division, and not agglomerated in drops, as in the pedicellate glands on the leaves and other green parts of plants. In sections of plants containing essential oil in an agglomerated mass, or even in rather large drops, the action of osmic acid is very different. For example, if a section of a leaf or petal of the orange plant be treated with this reagent, the osmic acid coming in contact with the drops of oil, is only reduced by the smallest drops or particles of the oil, leaving the other drops intact and unacted upon. This is further exemplified by placing side by side, and in contact, on a slip of glass, a drop of otto of rose and a drop of osmic acid solution (even a concentrated solution), the reduction only takes place at the points of contact of the two liquids, and for the reason that they are not reciprocally miscible one with the other. To produce the complete reaction, it is necessary to actively stir the mixture of the two liquids with the point of a needle in order to produce an intimate mixture or emulsion of the oil in the aqueous solution, and so to multiply the points of contact with the reagent; the mixture then blackens almost immediately. From this experiment we can draw a conclusion of great importance as regards the elementary biology of the epidermal cell of the petal; viz., that as the osmium is reduced with the rapidity above stated, it is evident that the oil is located equally at all points of the substance of the phytoblast, and exists in a state of extreme division, so permitting contact to be established at innumerable points simultaneously. This proves that the cell is not only the reservoir or containing receptacle of the oil, but that it is in fact the seat of its manufacture, the actual locality where it is generated by a natural synthetic process, and not the simple container of products elaborated in sub-adjacent tissues. It is also inferred that as this oil does not accumulate or take the form of drops, it is given off as an odorous emanation almost as soon as it is generated, and that the formation and the dispersion by exhalation proceed simultaneously. Hence the flowers soon lose their perfume after being cut. This feasible theory applies to many other odorous flowers which exhale their perfume, are not furnished with receptacles to store their oil, and therefore cannot yield oil on distillation (or at all events very little). Considering this structural arrangement, it is not surprising that the quantity of essential oil obtainable from the rose is so very small, and if not immediately distilled the yield is even smaller. The perfume exhaling from many other flowers, such, for instance, as pinks, stocks, hyacinths, and in some instances from leaves, such as Mimulus moschatus is doubtless referable to the same cellular formation. Such flowers as those of lavender, and such leaves as those of mint, thyme, and geranium are furnished with cells and glands which store up the oil, consequently their perfume is only perceptible when these vessels are ruptured by friction. The conditions in which the various oils exist in the plant and their liability to be more or less changed or damaged by the action of heat, of course guide the manufacturer as to the process he shall adopt to extract the products.

* Recherches sur le mode de production du parfum dans les fleurs, Note de E. Mesnard, présentée à l'Académie par Duchartre. Comptes Rendus, 21 Nov., 1892, cxv., p. 282.

It may be remarked that as osmic acid is reduced by fixed oils and by tannin as well as by essential oils, the presence of stereoptene in the rose petals may have interfered with the accuracy of Blondel's observations, and that the conclusion he arrived at, viz., that the otto is produced in the cells of both surfaces, may be erroneous; in any case, it has more recently been affirmed by Mesnard that the essential oil of rose is found in the papilliform epidermal cells on the upper surface of the petals, and scarcely ever on the lower side.*

Mesnard conducted his investigation in the Botanical Laboratory of the Sorbonne, under the direction of Gaston Bonnier. His interesting and suggestive memoir may be abstracted as follows:—

"The imperfection of the micro-chemical methods usually employed has hitherto prevented an exact knowledge being obtained of the manner in which the perfume of flowers is generated (and localised in the flower). In this particular investigation I have followed the same method of research which was adopted in the localisation of fixed oils. The method is as follows:— The section being placed in a drop of pure glycerine is arranged upon a round cover-glass, which being then inverted, serves as a cover to a small chamber formed by cementing a glass ring to an object-slide. In the interior of this chamber is fixed another ring of smaller diameter, and somewhat less in height, thus forming an inner annular space in which the reagent may be placed. By adopting this arrangement the light passing through the central part of the cell is not weakened. The inner ring further serves to support a very small cover-glass upon which sections may be arranged which require to be exposed to the action of the reagent for some length of time, as occasionally happens in the case of fixed oils. The reagent invariably employed is pure hydrochloric acid, the hydrated acid vapours abundantly given off from which are absorbed with avidity by the glycerine. In this way, by a gentle and easily regulated action, I obtain complete hydration of sections in the presence of an acid. After a few moments of exposure to the presence of the reagent, the essential oils appear as minute spherical drops of a fine transparent golden yellow. If the action be prolonged, the drops disappear, being transformed into diffusible products. The disappearance or diffusion of globules of fixed oils never takes place, the process thus furnishes a means of distinguishing these two classes of products.

"As regards the localisation of essential oils in the parts of the flower, the following observations have been made:—

" Jasmin. In this flower the oil is situated in the row of epidermal cells on the upper side of the sepals and petals. Some exist also in the corresponding layer on the under surface, where the sepals are coloured by a violet pigment. If the evolution of the cell contents in flowers at different stages of development be followed, at first nothing but chlorophyll is found in the tissue; tannin is the next to appear, or rather intermediate glucosides, difficult to identify by means of the ordinary tests for these substances. These glucosides furnish the tannin and pigments of the lower surface of the sepals. The hydrochloric acid vapours furnish a means of distinguishing all the tannoid compounds intermediate between the chlorophyll and tannin, or pigments, on the one hand, and between the chlorophyll and essential oil on the other. The explanation of these facts seems to be as follows:— Whereas, upon the lower surface, which, in the bud, was exposed to the action of light and the oxygen of the air, the tannoid compounds were slowly oxidised, thereby generating tannin, the upper surface, on the contrary, being then hidden inside the bud, these agencies were inoperative (the parts not being exposed to the action of light and oxygen), and the same compounds were converted into essential oil, which oxidises when in contact with air, and so produces the sensation of perfumes."— [This confirms the theory of Liebig and others, that perfume is the result of eremacausis.]

He then states his opinion that the essential oil is generated in the rose precisely in the same way as in the jasmin; also from his investigations he draws the following conclusions:—

"1°—That the essential oil is generally found localised in the epidermal cells in the upper surface of the sepals or petals, though it may exist upon both surfaces, especially if the floral organs are completely hidden in the bud. The lower surface generally contains tannin or pigments derived from it.

"2°—The essential oil seems in all cases to be the result of a transformation of the chlorophyll. This transformation is readily understood if it be admitted, as it generally is, that the floral organs are but leaves modified for the performance of a new function. The chlorophyll being thus diverted from its original purpose or use, is transformed into permanent tannoid compounds or into essential oils.

"3°—The liberation or disengagement of perfume from the flower only becomes perceptible when the essential oil is sufficiently freed from the intermediate compounds which generated it. Its formation is to some extent in inverse proportion to that of the tannin and colouring matters in the flower."

The above opinions (of Mesnard) are of course speculative.*

*The subject has been investigated by other naturalists amongst the most important records of their researches, the following may be referred to:— Guettard: "Sur les corps glanduleux des plantes," Mém. de l'Acad. Roy. des Sc. 1745 à 1756. De Mirbel: "Mém. sur l'anat. des plantes, et Eléments de physiol. vég. & de bot.," 1815. De Candolle: "Organogr. véget." Meyen: "Ueber die Secretions Organe der Pflaenzen," Berlin, 1837. Ditto: "Neues System der Pflanzen-Physiologie," 1837. A. Weiss: "Die Pflanzenhaare, in Karsten's Botanische Untersuchungen," 1887. Martinet: "Organes de Sécrétion des végétaux. Thèse de la Faculté des Sciences Paris," 1871. Raw: "Enumeratio Rosarum," 1816.