Studies on the Expression of Color Tone in Rose Petals
Hitoshi Yasuda

Jour. Fac. Lib. Art. Sci. Shinshu Univ. 14: 31-37 (1964)
Studies on the Expression of Color Tone in Rose Petals
I. Characteristic Structure of the Epidermis of Petals in Velvety Dark Red Roses.
Hitoshi Yasuda

We have a group of horticultural roses known as "Black Rose". It is how the varieties had this name that the petals of these varieties are dark red rather than red.

Many varieties of this group have petals with velvet-like surface changing in color tone depending on the strength and direction of light. Sometimes they look red while another times they are typical blackish red. Rarely they even seem deep black. Such color tone will be better called "velvety dark red".

In the preliminary experiment the author (unpublished) could proved that the quantity and quality of pigment estimated in the petals can not show any actual difference among these rose varieties.

In this paper, the surface structures of petals of the group beeing velvety dark red were compared morphologically with those of other red group. And the author observed that a certain structural difference between the two groups was a main cause for the different expression of color tone.

Fig. 1~4 Sections of the petal of BONNE NUI.
  1 : a longitudinal section of the area which is velvety dark red.
  2 : a cross section of the same area.
  3 : a longitudinal section of the area which is not velvety dark red.
  4 : a cross section of the same area.
Fig. 5~6 Sections of the petal of RADAL.
  5 : a longitudinal section.
  6 : a cross section.


BONNE NUI, CHARLES MALLERIN and JOSEPHINE BRUCE are the most typical varieties among so called "black rose", the petals of which are velvety dark red with black tone.

The author made a comparative morphological study on the epidermal structure of petal between a group of 3 varieties of the velvety dark red rose and another group of 7 red varieties. The study made it clear that the epidermal cells of the velvety dark red area in the petals were longer vertically to the petal surface than those of other varieties, and that each epidermal cell was separate without contacting each other, having more spaces among the epidermal cells than in those of other varieties of the roses.

These characteristics of the epidermal structure of the petals were thought to be responsible for casting the shadow of epidermal cells on the petal surface. It was concluded that the conditions above mentioned were a main cause of expression of black tone in the petals of the velvety dark red roses.

A term "morphological color tone" was proposed by the author to the histological structure affecting development of color tone.

Jour. Fac. Lib. Art. Sci. Shinshu Univ. 15: 15-21 (1965)
Studies on the Expression of Color Tone in Rose Petal
II. Changes of the Epidermal Structure in the Velvety Dark Red Petals as Flower Develops

It is well-known among many rose growers that the black tone of petals of velvety dark red roses becomes gradually lower in contrast with the red color coming out as the flower develops. On the view point of the author's previous paper that the expression of black tone of the velvety dark red petals was attributed to the shadow of epidermal cells, this phenomenon may be imputed to the morphological change of the petal surface as flower develops.

Various physiological and morphological changes may occur in opening flowers. In the case of rose flowers, there are found only a few physiological investigations, but so far as the author is aware we have none as to their morphological changes. For instance, SIEGELMAN et al.(1) showed that the respiration

of rose petals begun to increase just before flower development, became to maximum at fully-opened stage and thereafter depressed; WEINSTEIN et al.(2) studied the fixation of carbon dioxide in the flower of various opening stages; WEINSTEIN(4) reported that the contents of various organic compounds in petals varied diversely with flower development; and AHUJA(5) et al. and YASUDA(8) indicated that the quantity of the pigment in petals diminished slightly as flowering progressed.

In this paper the author studied the morphological change of epidermal structure in the petals of velvety dark red roses during the flower development, discussing the relation of this change to the gradual diminution of black tone.

Figs. 1-6: —The schematic sketches of epidermal structures of the petals in various flowering stages.


Three cultivars of velvety dark red roses, BONNE NUIT, CHARLES MALLERIN and JOSEPHINE BRUCE, were investigated concerning to the change of epidermal structure of petals during the flower development.

It was observed that the more developed flowers had the wider spaces among epidermal cells and the slant side wall to petal surface of epidermal cells, resulting in the condition that the epidermal structure grew less suitable to produce the unbroken shadow of epidermal cells.

The findings mentioned in this paper led the author to the following conclusion: the fact that the black tone becomes gradually lower and red color appears slowly with the progress of flowering in the petals of velvety dark red roses, is attributable mainly to the gradual elimination of the shadow of epidermal cells onto the petal surface.

Jour. Fac. Lib. Art. Sci. Shinshu Univ. 15: 23-30 (1965)
Studies on the Expression of Color Tone in Rose Petals
III The Role of Anthocyanin in the Expression of Black Tone in the Petals of Velvety Dark Red Roses.
Hitoshi Yasuda

The author suggested that the appearance (1) and disappearance (2) of black tone in the petals of velvety dark red roses were attributable presumablely to the shadow cast by the long nipple-like epidermal cells situated at some intervals on the red petal surface, without dependence on the quantity and quality of anthocyanin in these petal tissues. But this suggestion was made by some preliminary experiments, and therefore it was expected to estimate practically the pigments in these tissues in more detail.(3)


The anthocyanin in the petals of a velvety dark red cultivar BONNE NUIT, and a red cultivar HAPPINESS, was identified as cyanin according to the paper-chromatography recommended by SHIBATA et al.

Comparative determinations of cyanin contents in the petals were made between velvety dark red and red cultivars, between the black and red portions within the same petal and among the different stages of flower development respectively. It was found that cyanin contents of rose petals were not necessarily higher in the dark red cultivar nor in the dark red portion of a single petal.

Furthermore it was also evidenced that the cyanin content in the petals had no causal relation to more blackish appearance of the opened flowers.

Spectrophotometrical determinations indicated that there was no significant difference in reflectances between the velvety dark red and red petals.

According to these findings, it may be concluded that the quantity and quality of anthocyanin in the petals of velvety dark red cultivars are not concerned directly in the expression of black tone.

Bot. Mag. Tokyo 80(944): 86-91. (Feb. 1967)
Studies on the expression of color tone in rose petals
IV. Comparison of spectro-reflectance of red rose petal and spectro-transmittance ofanthocyanin solutions.
Hitoshi Yasuda
(mostly in Japanese)


The spectro-reflectance curves of red rose petals (Fig. 1) of cultivars, Happiness, Karl Herbst and Radar, were compared with the spectro-transmittance curves of the pigment solutions containing 25~450 γ/ml of cyanin at pH 2.0~3.5 (Fig. 2).

When cyanin in higher concentrations was applied, its transmittance curve came near to the reflectance one.

The transmittance curve of cyanin solution saturated with rutin was more similar to the reflectance curve.

To make the better comparison between two kinds of curve, several per cent. of reflectance must be deducted from the total reflectance measured, taking the surface reflection of red petals into account (Yasuda, 1964)10)

Bot. Mag. Tokyo 80(951): (1967)
Studies on the Expression of Color Tone in Rose Petals
V. Colorimetric study on the quantitative effect of anthocyanin in red petals.
Hitoshi Yasuda
(mostly in Japanese)


The relationship between anthocyanin content and flower colors was studied in roses using two groups of cultivars, the red one including Happiness, Crimson Glory and Karl Herbst, and the black one including Bonne Nuit, Charles Malleriri and Josephine Bruce.

I.S.C.C.—N.B.S. color name was used for the designation of the flower colors.

In red cultivars, the petal color varies with increasing anthocyanin content ranging from 150 µ/cm2 to 800 µ/cm2 in the following order: deep or dark purplish Red—very dark purplish Red—very dusky Red.

In black cultivars, increase of pigment content within the range of 300 µ/cm2 to 1000 µ/cm2 brings about the color change in the following order: very dusky purplish Red—very dusky Red Purple—purplish Black—reddish Black.

The order of color variation in red petals within the black cultivars was the same as that within the red cultivars.

No quantitative effect seemed to exist between the red and black cultivars.

Bot. Mag. Tokyo 80(953): 459-465 (Nov 25, 1967)
Studies on the Expression of Color Tone in Rose Petals
VI. Role of Surface Reflection in Red Petals.
Hitoshi Yasuda
(mostly in Japanese)


Using the method in color science, attempt was made to demonstrate the role of surface reflection in petal color of roses including the red and black cultivars: Radar, Fire King, Independence, Happiness, Karl Herbst, Christian Dior, Charles Mallerin, Josephine Bruce, Crimson Glory, and Bonne Nuit.

Two I.S.C.C.—N.B.S. color names were determined by calculations from each spectral reflectance curve of the petals, one on the basis of the usual horizontal axis (Fig. 1, O—X) and the other on the translated horizontal axis (Fig. 1, O'—X'). The latter was regarded as the level corresponding to the reflectance by surface reflection1,2). From the difference in these two designations (Table 1), it follows that the effect of surface reflection on red petal color may be devided, according to the lightness Y, into the three types, as follows: (1) when Y lies within the range of 8 to 13%, surface reflection exhibits no perceptible effect; (2) when Y is within the range of 6 to 8%, the surface reflection affects only the level of "tone", showing a tendency to reduce the saturation in reddish color; (3) when Y is below 6%, the reflection shows an effect on the main color" level in the direction of YR→R→RP in its hue.

Thus, the surface reflection is to be regarded as one of the essential factors in an expression of petal colors.

Bot. Soc. Japan 81(961-962): 377-384 (1968)
Studies on the Expression of Color Tone in Rose Petals
VII. On the Spectral Transmittance Curves of Red Petal Extracts
Hitoshi Yasuda
(mostly in Japanese)


The spectral characterization was made on aqueous acid extracts of the petals of red rose cultivar Happiness and those of black cultivar Bonne Nuit. When these extracts were heated at 68-75°, the spectral shift from 508mµ to 502mµ came into appearance. The wave length showing minimum transmittance of the petal extract was ca. 6mµ longer than that of the acidic solution of crystalline cyanin, i.e., an anthocyanin component of rose petals. Followingly, it is conceivable that co-pigmentation takes place in these extracts.

From the comparison of spectral transmittance curves in detail, it was shown that the petal extracts exhibit some hyperchromic and bathochromic effects on anthocyanin color at pH 3, whereas only a bathochromic effect at pH 1. Colorimetric calculation based on these curves has revealed the existence of reddening effect in these extracts.

No significant spectral difference was observed between the petal extracts of red and black cultivars. This fact seems to indicate that the co-pigmentation is not involved in an expression of the black color in rose petals.

In the extracts showing an obvious spectral shift, certain flavonoids, tannins and metallic elements (K, Na, Mg, Fe, Al) have been detected. It is likely that the spectral shift observed above is due not to flavonoids and catechol tannin, but chiefly to pyrogallol tannin.

Bot. Mag. Tokyo 82(973): 308-315 (July 25, 1969)
Studies on the Expression of Color Tone in Rose Petals
VIII. Transmittance Curves of Red Petal Extracts by the Use of Acids Inherent to the Petals
Hitoshi Yasuda
(mostly in Japanese)


The transmittance curves were studied on petal extracts from red roses which were prepared using the acids inherent to the petals. Cultivars used were 'Happiness' and 'Bonne Nuit', the representative ones in a whole series of this study. The measurement of component acids in the petals indicated that the total amount of non-volatile acids was about fifty times as much as that of the volatile acids, in which carbon dioxide was a major component. Non-volatile acids in the petals were identified by paperchromatography using the following solvents: butyl formate 10ml, formic acid 4ml, water 1ml; phenol 3ml, water 1ml, 90% formic acid 1%; chloroform 2ml, ethanol 1ml, 90% formic acid 2%. The chromatograms showed that in the petals examined, phosphoric and malic acids predominated, while citric and succinic acids were present at lowest level. Fumaric acid was scarecely detected.

On the basis of this fact, petal extracts were prepared using 0.1N solutions of phosphoric, malic, citric and succinic acids, respectively. Similar solutions containing both phosphoric and malic acids in the ratios of (2:1), (1:2) and (1:1) were also applied for extraction.

Spectral transmittance was measured on these extracts. No significant difference was observed in the curves obtained here. At a higher concentration of cyanin, the transmittance curves were similar to the reflectance curves of red petals.

Bot. Mag. Tokyo 84: 256-260 (June 25, 1971)
Studies on the Expression of Color Tone in Rose Petals
IX. Quantitative Effect of Anthocyanin in the Petals of Pinkish Cultivars.
Hitoshi Yasuda
(mostly in Japanese)


The quantitative effect of anthocyanins on the color of petals of weak pinkish and deep pinkish rose cultivars was investigated, and results were compared with the data previously obtained on red petals.

The major pigment in those petals was identified as cyanin by paper chromatography. Cyanin contents were: 1.5~10 µg/cm2 in weak pinkish petals, and 20~80 µg/cm2 in deep pinkish ones (cf. 100~900 µg/cm2 in red petals, as reported previously3)).

Chromaticity coordinates (x, y) were calculated from the transmittance curves for pigment solutions containing different amounts of cyanin, and also from the reflectance curves for the petals of cultivars used.

From the correlation between the colorimetric data and the cyanin contents in pigment extracts or petals, the following possibilities may be considered:

1) Visible color difference between pinkish and red petals seems to be determined by difference in cyanin content in the petals.

2) Color difference between deep pinkish petals and weak pinkish or red ones can not be ascribed solely to the quantitative effect of cyanin alone; in addition a bluing effect must take a part in the development of deep pinkish color in rose petals.

Jour. Fac. Sci., Shinshu Univ. 8: 127-134 (Dec 1973)
Studies on the Expression of Color Tone in Rose Petals
X. Transmittance Curve of the Red Petal Extract, Measured with Translucent Cuvette.
Hitoshi Yasuda


On the reflectance measurements of red petals, part of incident light should be absorbed into some inner constituents of petals, and the remaining part passed away through the petals. The only part of incident light is detected by the spectrophotometer as the reflected light.

In the present study, the transmittance measurements on pigment solutions were taken with the translucent cuvette to give an optical state similar to the translucent property of the petal.

The translucent cuvettes were prepared by inserting the semitransparent pieces of plastic sheets along the inner walls of ordinary cuvettes.

When the cuvettes having the transmittance of 40 to 60% were used, the transmittance curves of red petal extracts became more approximate to the reflectance ones of red petals.

From the results of colorimetric calculation it is suggested that the translucent property brings some modification of the color tone in the petals having higher content of pigment.


According to SHIBATA'S suggestion6),7), it may be expected that when the petals or the translucent cuvettes are irradiated, various light paths are distinguishable. The representative light paths in these cases are illustrated schematically in Fig. 4 A and B. The nomenclatures and the symbols of light paths were refered to SHIBATA'S. It was pointed out in the previous paper8) that the light reflected by the red petal is composed of the following two parts: the directly reflected light and the diffused reflected light. On the other hand, transmittance measured with the translucent cuvette must be based on the next two parts: the parallel transmitted light and a part of diffused transmitted light.

Fig. 4 Schematic figures showing representative light paths in the petal (A) and in the translucent cuvette (B), Io : incident light Isr : directly reflected light Idr : diffusely reflected light la : absorbed light I diffuse transmitted light Ip parallel transmitted light.
In (A) spotted portions denote red color.

There are differences in the names of light captured by the light detector between reflectance and transmittance measurements, but two spectroscopic measurements are common in the next point: a certain part of incident light can be attenuated by the translucent materials, i.e. petal or polyvinyl sheet, and the remainder can be detected by the photometer, after being characterized with spectra of cell sap or pigment solution. The problem as to the significance of directly reflected light in the red petal was discussed in the previous paper8). Accordingly, it can be said that by the application of translucent cuvette the translucent property of the petals may be demonstrable to some degree upon the transmittance curves.

Fig. 5 Schematic figure pointing various effects on the reflectance curve of a red rose petal.
  1 : Effect of anthocyanin content.4)
  2 : Effect of surface structure.9),10)
  3 : Effect of copigmentation-like factor.11)
  4 : Effect of surface reflect.8)
  5 : Effect of translucent property.

The present author failed in this experiment to devise a cuvette having the transmittance curve closely similar in shape to the reflectance curve of a white petal. Furthermore, when a red petal deprived of the red pigment is considered a question remains as to whether this imaginary ground of pigmentless red petal can be replaced by the white petal. Hence, curve 3 in Fig. 3 A~C respectively is only an approximate indication of the translucent property in the petals. However, it seems evident that the translucent property has an effect to change the S‑shape of transmittance curve toward lower transmittance. Thus, one can give the transmittance curve of pigment solution more similar in shape to the reflectance curve of red petals.

Fig. 5 illustrates summarily the effects of principal factors on the development of petal color in red roses. Effects 1 to 4 were pointed out in the previous papers4),8),9),10),11), and effect 5 was revealed in the present work.

From the results shown in Table 1, it can be pointed out that the translucent property brings about some modification, e.g. a change of color in the "tone" level, in the cell sap having the higher concentration of pigment. It is natural to propose that the translucent property of petals depends prominently upon the inner structure of petal. Therefore, it can be said that the result presented here is a possible example of a color modification of pigment caused by the inner structure of the petal.


1) YASUDA, H. (1967) Bot. Mag. Tokyo 80: 86
2) _____ (1969) ibid. 82 308
3) _____ (1965) Jour. Fac. Lib. Art. Sci. Shinshu Univ. No. 15; 23
4) _____ (1967) Bot. Mag. Tokyo 80 : 357
5) HIOKI, R. (1964) Handbook of Color Science, p.71 (Shikisaikagakukyokai, 2nd ed. Nankodo, Tokyo)
6) SHIBATA, K. (1959) Method of Biochemical Analysis VII: 77 (GLICK, D. ed. Interscience Inc., New York)
7) _____ (1962) ibid. IX : 217
8) YASUDA, H. (1967) Bot. Mag. Tokyo 80 : 456
9) _____ (1964) Jour. Fac. Lib. Art. Sci. Shinshu Univ. No. 14: 31
10) _____ (1965) ibid. No. 15 : 15
11) _____ (1968) Bot. Mag. Tokyo 81 : 377