Repeat-flowering old roses part I by Peter Harkness

Much ink and many words have gone into defining the term “repeat-flowering”. For my purpose I take it to cover any rose that can be depended on in a normal season to bear blooms in summer and in autumn.

Today we take such behaviour in a rose for granted, but in terms of rose history it is not so long ago since things were very different. Jamain & Fomey, writing in Les Roses published 1873, noted that: “In 1815, remontant roses were still rare; one can only find two in the Portland series, the biferas ‘Talmyre’ and ‘Venusta’.”

And yet, comfortably within one life span, the world of roses underwent such amazing developments that Thomas Rivers, born in 1798, could declare in his book The Rose Amateur’s Guide: “Summer roses will soon be out of favour except a few of the very best.”

Now step back in time to 1763. What repeat flowering varieties were on offer then? There survives in my local Record Office an order for plants from Thomas Emmerton of London for John Radcliffe, the squire of Hitchin Priory in Hertfordshire. Here is the part that is of particular interest to us:

John Radcliffe Esqr. Apr 8th 1763 to Thos. Emmerton
5 Monthly Roses at 4d£ 0-1-8
3 Damask Do.£ 0-1-0
2 Red Belgick Roses £ 0-1-0
2 Blush Hundred Leav’d Roses £ 0-1-0
2 Double Velvet Roses £ 0-1-0
1 Rose of Monday£ 0-1-8

Received September 16th 1763 the full contents of this Bill and all demands by Mr. Thos. Emmerton

How many of these names can we identify?

If we start at the bottom with ‘Rose of Monday’, we are reminded how easily rose names become corrupted, for it is in fact our old friend ‘Rosa Mundi’, also known as R. gallica ‘Versicolor’.

‘Double Velvet’ must be ‘Tuscany’, and ‘Blush Hundred Leav’d’ a Centifolia rose, the word ‘Leav’d’ actually referring to the petals of the flower. ‘Red Belgick’ appears in an English list of 1750 from Henry Clark of Chipping Campden in Gloucestershire, and may, as the name suggests, be a form of Gallica from Flanders, where important developments of this group were being made.

Damask roses

Above that appears the entry ’3 Damask roses’. Now there were at that time ancient forms of the Damask Rose, which flower in summer only, and later ones that re-flower, and this entry I believe refers to the summer type.

Above that we see ’5 Monthly Roses’, and these cannot be the China rose called ‘Old Blush Monthly’ because that is not thought to have reached England until some years later. No, these in all probability are Damask roses of the re-flowering kind, known by several names including ‘Autumn Damask’, ‘Quatre Saisons’, ‘Four Seasons’ or R. damascena bifera – ‘bifera’ meaning ‘twice bearing’.

It is said by Richard Hakluyt in 1582, that the ‘Autumn Damask’ was brought to England from Italy about the 1520s, and the name Damask tempts one to suggest a Middle Eastern origin, but really no one knows. Although the autumn flower is sporadic and sparse by modem standards, it was an obvious favourite as almost the only variety that, in the English climate, could be relied on to provide late blooms.

An improved form, known as ‘Tous les Mois’ or ‘All the Months’ seems to have arisen in seventeenth century France in the gardens of Louis XIV, and some colour variants of ‘Autumn Damask’ also evolved. In a select list of ‘the best roses’ in 1838 we find four ‘Autumn Damask’ roses, the blush, the pink, the scarlet and the white.

Also available by 1838 was a sport of the white ‘Autumn Damask’, with a curious soft growth over the stems and calices. The French introducer called it ‘Quatre Saisons Blanc Mousseux’ and the English ‘Perpetual White Moss’. This was a somewhat misleading name, for it is not reliably perpetual.

Rosa moschata

Whence did the Autumn Damask gets its repeat-flowering character? For years it has been guessed that this rose came about as a natural cross between a Gallica and a plant we call R. moschata. Now R. moschata holds an important place in rose history and philologically bears as ancient a rose name as can be devised, though not perhaps the most romantic one, for it derives its name from the Sanskrit word mushka, with reference to an intimate part of the anatomy of the musk deer, from which musk fragrance can be obtained, and which the scent of the rose is said to resemble.

What we now call R. moschata was pictured in a book of 1586, labelled with its French name ‘Rose Muscade’ and the English name ‘Muske Rose’. But although rosarians have bestowed on it the name of a species, there is no convincing evidence that it has ever been found in the wild; and since also it can vary its petallage, the indications are that it is likely to be a hybrid.

It was first botanically described by Johann Hermann in 1762, and is properly called R. moschata J. Herrm. This was a long time after its first appearance in Europe, whenever that was. We have the evidence of the book of 1586 just mentioned, but it has been claimed that the Musk Rose reached England from Italy before 1540, supposedly through Thomas Cromwell, who as a young man had acted as a mercantile agent in Venice. This is the same Thomas Cromwell who became Henry VIII’s trusted servant, and the infamous prime mover in the destruction of England’s monasteries.

An unusual feature of R. moschata J. Herrm. for a rose of this period is that, like the ‘Four Seasons’ it is supposed to have fathered, it is considered repeat-flowering. Having heard varying accounts of its ability in this respect, I was interested to receive testimony from rose expert Professor Michael Manners of Florida:

“I have in my collection” he says, “all of the known accessions of R. moschata, including the one Graham Stuart Thomas found, as well as the several Virginia, North Carolina and South Carolina finds. As far as I can tell, they are the same rose, apparently existing as three cultivars, one single, one semidouble and one quite double. While they may be very late once-blooming in a cool climate like Britain’s, in a warmer climate they start flowering late, then they continue, as prolifically as any Floribunda, until stopped either by frost or hard pruning. My plants normally flower, non-stop, from about the middle of May (very late for central Florida roses) through February, when we cut them back.”

In cooler England, we can expect R. moschata to come into bloom so late in the summer that the hoped-for later blossoms may be forestalled by autumn chills. The name it was sold under in eighteenth century England was ‘Cluster, Late White’, and some customers familiar with the ways of catalogue writers might have taken that as a coded warning, and decided not to splurge one shilling and sixpence on it. Nevertheless, the genes for repeat flowering were there.

Alternative parents

The other parent of ‘Autumn Damask’ is supposed to have been a Gallica, some form or descendent of R. gallica ‘Officinalis’. Here there is a mystery, because Gallicas do not repeat their flower, and the received wisdom is that a first generation hybrid raised from one repeating and one non-repeating parent will not repeat.

There are two possible solutions to the puzzle. There could have been a second generation cross, again with Moschata; or the hybrid could have pollinated itself and produced a seedling which by one of nature’s freakish bounties reinforced the genes for repeat flowering to bring about the end result.

Ivan Louette of Belgium has done much research on the parentage of ‘Autumn Damask’, and has put the case for a parent other than R. moschata J. Herrm. He gives us a choice of either R. abyssinica, whose structure, proportions of various parts of the flowers, shape of buds, prickles etc. is much closer to ‘Autumn Damask’ than R. moschata; or a related Moschata called ‘Nastarana’, called the ‘Persian Musk Rose’. And in a recent article, he has suggested R. phoenicia. All three have a lengthy period of bloom and are therefore likely candidates.

Against this, Charles Quest-Ritson questions whether any form of R. moschata could have been the parent of ‘Autumn Damask’, on the grounds that the two have so few characteristics in common, and also because of the number of their chromosomes – Moschata being diploid and Gallica (like ‘Autumn Damask’) tetraploid.

DNA research

However, the May-June 2001 issue of Rosa Gallica includes a fascinating article offering a solution. DNA research carried out in 1998-99 by Hikaru Jwata and Tsuneo Kato of Hiroshima on the summer flowering Damasks `Kazanlik’ and `York and Lancaster’ and the `Autumn Damask’ and `Quatre Saisons Blanc Mousseux’ revealed that all of them derive from the same three species, and that the original natural hybrid that gave rise to them would have been R. moschata (seed parent) and R. gallica (pollen parent). This hybrid was then pollinated at some stage by R. fedtschenkoana. The parentage of these ancient Damask roses should therefore be correctly expressed as (R. moschata x R. gallica) x R. fedtschenkoana.

This meets Quest-Ritson’s objection as far as plant characteristics are concerned because R. fedtschenkoana has features in common with the Damasks, notably in respect of foliage and hips; the researchers claim that this species, which bears a dense covering of fuzzy bristles on its new young shoots, also accounts for the mossiness of `Quatre Saisons Blanc Mousseux’. Furthermore, the somewhat lanky, lax habit of R. fedtschenkoana seems to me typical of many Damasks.

On the question of length of flowering, those familiar with R. fedtschenkoana in a garden setting will have noted that its blooms appear sporadically from summer through to late autumn. It is perhaps not unreasonable to conclude that the genes of R. moschata and R. fedtschenkoana have combined successfully in the second stage cross to produce the autumn flowering of `Quatre Saisons’ and its related forms, but this particular gift has not passed to the summer Damasks.

The Mystery of the Portland Rose

So much for ‘Autumn Damask’ and Moschata. Now if Squire John Radcliffe had been ordering a few years later than 1763, he could have requested another repeat-flowering rose, the Portland. This rose has puzzled rosarians for many years. Seeking to nail down facts about it is like handling mercury. Here is a selection of the opinions of various experts :

  • It originated in Beaconsfield; in Dorset; in Naples.
  • In the 1770s; in the 1790s.
  • The name is Portland, Portland Crimson Monthly Rose, Portlandica, R. portlandica, R. Paestana, Rosier de Portland, Duchess of Portland.
  • It’s a cross between ‘Autumn Damask’ and R. gallica ‘Officinalis’; no it’s not, it’s a cross between ‘Autumn Damask’ and ‘Slater’s Crimson China’; it’s impossible that it could be a cross from ‘Slater’s Crimson China’; all these suggestions are wrong – it derives directly from Gallica roses and is not a hybrid with anything else at all.

If you look in recent publications like Botanica’s Roses or The Quest for the Rose you find a sort of conflation of some of these contradictory statements. Even the most learned of our rosarians has betrayed some confusion. In The Graham Stuart Thomas Rose Book , published in 1994, the main text declares that “From the colour and dwarf habit of the Portland it may be presumed that the China parent concerned was ‘Slater’s Crimson China’.” But that statement is then followed by an asterisk, which directs our attention to this footnote: “It is now considered that the parents of the Portland Rose were R. gallica and R. damascena semperflorens and that they owe nothing to R. chinensis.”

In his earlier The Old Shrub Roses Graham Thomas published the findings of the pioneer rose geneticist Dr C.C. Hurst, who wrote of the Portland Rose that it “was a bright red verging on scarlet, and if treated well and pruned in a certain way it flowered twice a year. Judging from Redouté’s accurate figure… it is evidently a China-Damask-French hybrid… From the colour and dwarf habit… it may be presumed that the China parent concerned was ‘Slater’s Crimson China’.”

In the same book Graham Thomas thought Hurst must be wrong about the parentage on the grounds that ‘Slater’s Crimson China’ was not known in Europe before 1791, whereas English records showed that the Portland Rose was already in commerce in the 1770s.

So much for the confusion as to the Portland Rose’s parentage. Now for the evidence of my own eyes. For many years I used to see and photograph a variety labelled ‘Duchess of Portland’ in the Royal National Rose Society’s Gardens of the Rose. Some years later I saw a rose labelled ‘Paestana’ at Bagatelle in Paris. I was puzzled because although ‘Duchess of Portland’ and ‘Paestana’ are supposed to be synonymous, the roses appeared distinctly different, the flowers at Bagatelle being a much brighter red, and the growth taller and more lax.

What conclusion should I draw? That the differences were accounted for by climate and location? That one of the roses must be incorrect? Or that I’d better cut down on the red wine? I am wondering if there is another explanation. What if there were two different roses, each bearing the Portland name?

Two different roses?

According to Henry Andrews in his book Roses of 1805, ‘Portlandia’ as he terms it may have originated in England, in the gardens of the ducal Portland family at Bulstrode, near Beaconsfield. That it had some degree of remontancy seems evident from a comment in 1775 by the agricultural writer Richard Weston, who calls it the ‘Portland Crimson Monthly Rose’, and says it is an item “to be readily found” (mentioned in Brent Dickerson The Old Rose Advisor).

It was in cultivation in 1775 under the name ‘Portlandica’, according to Trees & Shrubs Hardy in the British Isles, and in 1782 appears as ‘Portland’ in the catalogue of Mr Brunton of Birmingham. He priced it at I/- which suggests it was not a very recent novelty. (See John Harvey’s ‘Prices of Roses in the Eighteenth Century’ in the 1979 RNRS Annual ). Another source says that in 1785 it arrived in France.

Three twentieth century English rosarians, all familiar with the ‘Duchess of Portland’ at St Albans, have described it as follows: “…more like a Gallica than a Damask, the flowers taking after R. gallica ‘Officinalis’.” [J.Harkness, Roses, p.184]; “a neat upright suckering bush of 3ft average height, with Gallica flower form, cupped to flat and semi-double with yellow stamens, of a brighter and deeper crimson-pink than ‘Officinalis’ – though by no means dissimilar except that the blooms repeat in the autumn.” [M.Gibson, Shrub Roses, Climbers & Ramblers p.82 & plate 8]; “Its colour distinguishes it from all the old Autumn Damask variants, which were never more than pink.” [G.S.Thomas, The Old Shrub Roses. p.209].

Compare them with what William Paul writes in The Rose Garden 1848: “Flowers deep rose, tinted with purple, large and semi double; form, cupped… a most abundant summer bloomer.” That sounds consistent with the modern English writers and with the rose at St Albans, though it is interesting that Paul did not consider its autumn flowering capability worth a comment.

Now for my theoretical later Portland rose. Some sources claim that this originated in Italy, perhaps on account of the name it was given, R. paestana, though this of itself need not denote an Italian origin as it may simply be a classical allusion to the historic “twice-bearing” Rose of Paestum.

Originating perhaps about 1795, it found its way first to England and then, in 1803, to France, where, some time between then and 1809, Andre Dupont, gardener to the Empress Josephine, named it ‘Rosier de Portland’. Its picture by Redouté‚ appeared in 1817 titled ‘Le Rosier de Portland’.

In the Almanach des Roses of 1811, by Claude-Thomas Guerrepain, there is a description of the rose which is valuable because it is both early and detailed: “One of the most striking by the brightness of its colour, and its precious quality of blooming from spring to fall. The flower buds are pointed and crowned with lacinated leaflets formed by the elongated sepals… its colour is a handsome red which makes it distinct and noticeable among all other sorts, especially in the autumn, when no others of that colour are to be found.”

Louis Noisette described it thus in 1825: “Flowers semi-double, sparkling maroon”.

An interesting description from Thomas Rivers seems to bear out the idea of two Portland roses. Writing in 1837 reasonably close to the events, his words can be taken to suggest he does not consider ‘Portland’ and ‘Paestana’ as one and the same, for he refers to “Rosa Portlandica, a semi-double, bright-coloured rose, much like the rose known in this country as the Scarlet Four Seasons, or Rosa Paestana.”

The “later” Portland Rose, with its brighter colour and laxer growth, as seen in Paris (and indeed now to be found at St Albans also, under the name R. paestana) could be the result of a different cross from the “earlier”, involving ‘Autumn Damask’ and some degree of genetic input from ‘Slater’s Crimson China’ which has had the effect of giving richer colour and reinforcing remontancy, without changing the old rose character of the foliage. Since ‘Slater’s Crimson China’ seems to have been growing in England in 1791, the objection which Graham Thomas raised to its possible parental role as suggested by Hurst would not apply to the “later” Portland.

In the group of roses that came to be called Portlands, not all were dependably repeat flowering. This is borne out by the quotation at the beginning of this article, which says only a couple were considered so as late as 1815. Rosarians also noted that the Portlands that gave more bloom in autumn tended to be less vigorous in growth. If there were an “earlier” and a “later” Portland Rose, these developments could reflect their respective genetic influence, the ‘Autumn Damask’ strain tending to grow more robustly but give fewer flowers than the ‘Slater’s Crimson China’ strain.

Duchesses of Portland

Before leaving the Portland rose, it is appropriate to comment on the supposed connection with the Duchess. Both the second and third Duchesses of Portland have been proposed as having inspired the name. Perversely, my suggestion is that it was neither of them, or at least neither of them directly.

It is true that the second Duchess was a noted gardener, but there appears to be no evidence for the Portland Rose being called ‘Duchess of Portland’ in her lifetime or for long afterwards. She died in 1785, and it was simply not the practice in the eighteenth century to name roses after individuals.

Therefore, despite Henry Andrews’ report in 1805 that the rose from Bulstrode “is said to have received” its title “in compliment to the late Duchess of Portland”, it seems more in keeping with the times that the name ‘Portland’ or ‘Portlandica’ was bestowed to indicate the estate associated with its cultivation, much in the same way as the important variety ‘Trianon’ came from the French royal garden of that name.

The variety from Italy, known as R. paestana, was, we recall, renamed by Andre Dupont ‘Rosier de Portland’ in or after 1803. Some authorities claim that by this act Dupont was showing his gratitude to the third Duchess for her help in persuading the Board of Admiralty to grant a passport to nurseryman John Kennedy to carry this and other roses through the wartime blockade. And another story credits the Duchess with having found it in Italy about 1800 and arranged to take it back to England with her.

Both suggestions are impossible, for the simple reason that during those years there was no Duchess of Portland – the third Duchess died in June 1794, and the Duke, who did not remarry, lived on until 1809.

Besides, the actual name chosen by Dupont and used by Redouté was not ‘Duchess of Portland’ but ‘Rosier de Portland’, meaning rose of Portland. As we have seen, it cannot have been the Duchess he was honouring, but what about the Duke? The Duke was certainly in a position to assist him with a passport for his roses in spite of the blockade, for he held high office almost uninterruptedly throughout the period, as Home Secretary 1794-1801, then briefly in opposition, then from 1803 Lord President of the Council and then Minister without Portfolio, culminating in his second stint as Prime Minister from 1807 until shortly before his death.

That still leaves the question unanswered of when and why the name ‘Duchess of Portland’ gained circulation. It does not appear among the Portlands and Perpetuals in Mr Wells’ list of 1838 in which he recommends the best sorts to grow, nor in William Paul’s The Rose Garden of 1848 (where the Portlands are considered as Damask Perpetuals), nor in the list of Hardy’s roses at les Jardins de Luxembourg in 1859, which gives R. portlandica. It’s a question that deserves farther exploration.

This article appeared in the Autumn 2001 issue.

Part II of this article (Spring 2002) examines the impact of the China rose.

← Singing the blues / Repeat-flowering roses part II →