Journal of Genetics, 1936: p 149
A Biochemical Survey of Factors for Flower Colour
Rose Scott-Montcrieff

Rosa polyantha: An interesting spray of flowers from a sporting Polyantha Rose ("Paul Krampel") was recently sent by Mr R. E. Cooper of the Royal Botanic Garden, Edinburgh. While the normal flowers were scarlet two distinct mutations had occurred; one to crimson and the other to a dog-rose pink.

A chemical examination of the pigments involved showed that the normal scarlet flowers were deeply pigmented with pelargonin and some flavone, while in the two sports cyanin took the place of pelargonin, the pale pink flowers having a smaller amount of anthocyanin than either the scarlet or crimson flowers, together with a proportional increase in flavone content. All three types of flowers contained large amounts of tannin.

The mutation from scarlet to crimson thus involves a change in pigment to a more oxidized anthocyanin with a similar 3-5-dimonosidic residue, while the change from scarlet to pink appears also to involve co-pigmentation and a change in the anthocyanin-flavone balance, and is apparently due to a double mutation.


Cytologia, 2. 1931: p 268-269
Chromosome organization in Rosa
Eileen Whitehead Erlanson

(e) Structural Hybridity

In the rose "Orleans", among forty whole nuclei, at diakinesis, twenty-nine showed seven bivalents. In seven others one pair of chromosomes had failed to form chiasmata, giving six bivalents and two univalents. In three nuclei five bivalents and a quadrivalent were found. One nucleus among the forty showed four bivalents, a quadrivalent and two univalents. The proportion of potential bivalents with no chiasmata is 2.5 percent at diakinesis, and 2.9 percent at metaphase. The proportion of chromosomes involved in quadrivalents is 2.9 percent at both stages. This is a statistical proof that the end to end association at metaphase is functionally equivalent to the interstitial chiasma of prophase. The end to end association is a terminal chiasma.

[background discussion]

The occurrence of rings of four in the rose "Orleans", therefore, shows that it is a structural hybrid as well as a hybrid in the general sense. The low frequency of quadrivalent formation in this plant is due to the smallness of the exchanged segments, since chiasma-formation is fortuitous and random (Gairdner and Darlington 1930). Chiasma-frequency at diplotene is only 2.66 per entire bivlent, and there is no reason to suppose that the exchanged fragments are even a quarter of the length of a whole chromosome. In Campanula Gairdner and Darlington (1930) found complete terminalization which gave rings or chains of four at metaphase. In the rose "Orleans" terminilization is partially arrested and some interstitial chiasmata persist in the quadrivalent groups at metaphase.


The World of Irises, p 369
Iris Pigments
Norlan Henderson

Werckmeister (1969) discussed the breeding of roses in which a kaempferol-bearing rose was crossed to a cyanidin-bearing plant and which gave rise, eventually, to a scarlet pelargonidin-bearing flower. He suggested that the same sort of breeding program might be useful in the eventual development of a red iris.

Megson (1974) pointed out that the similarity of kaempferol to pelargonidin and the discovery of kaempferol in the tall bearded irises is good news to hybridizers working toward a red iris. It should also be noted that quercetin is closely related to cyanidin structurally. A similar condition exists in the flavones in which the apigenin-pelargonidin and the luteolin-cyanidin relationships are obvious. However, little has been done to determine the biosynthetic pathways of these pigments. Although they are structurally similar, it is not known whether they are or can be precursors of the anthocyanin pigments.

  1. Werckmeister, P: Red Irises and cyanidin. Bull. Amer. Iris Soc. no. 194: 7-13. 1969
  2. Mergson, F. H. A review of the article by Ashtakala and Foward. The Medianite 15:7-8. 1974

Comments: I am sceptical of "double mutations" occurring in rapid succession. Paul Crampel was itself a sport from Orleans Rose, which means that one of these mutations was a reversion. Many other sports came from Orleans Rose, and from its sports, which vary in the proportions of the pigments and co-pigments. If we accept the opinion of Scott-Montcrieff, we must suppose that a few genes are wildly mutating back and forth and sideways.

A more plausible explanation is that position effects, possibly associated with the chromosomal interchange reported by Erlanson in Orleans Rose, account for the color instability in the Polyanthas and their descendents. The genes themselves do not mutate; rather, changes in the attachments of the homologous segments of different chromosomes alter the expressions of the genes. This explanatation is supported by the observation that decreased anthocyanin was matched by an increase in flavones. Anthocyanins and flavones are synthesized from the same precursor molecule.