F.C. Stern


Journal of the Royal Horticultural Society.1943 vol. 68 (5) 124 – 131

By F. C. Stern, O.B.E., M.C., V.M.H.

Attention! This publication is subject to copyright by the Royal Horticultural Society. www.rhs.org.uk

A " Study of the Genus Paeonia" has been completed and will be published by the Royal Horticultural Society after the war. A synopsis of this study is being published in the R.H.S. Journal as it might be of interest to the readers of the Journal.

The late H. J. Elwes of Colesboume suggested to me many years ago that a new study of the genus should be undertaken. The names of Paeony species were in much confusion and since Huth published the last study of the genus in 1891, several new species had been discovered. The present study was begun in 1919. In order to identify each plant accurately, it was necessary to raise the different species from collected seed from the natural habitat of each species.

The revision of the genus has been based entirely on specimens collected in a wild state, which are in the herbaria of Kew, Edinburgh, British Museum, Paris and Manchester, and also from specimens kindly lent by the authorities of the herbaria' in Leningrad, Calcutta and Vienna. I am grateful to all the officials of these herbaria for their unfailing help and assistance.

The genus Paeonia is only found in the Northern Hemisphere. The species have been divided into three sections :—

1. Moutan—stems woody, perennial.

2. Onaepia—stems herbaceous ; petals not longer than the sepals ; disc produced as fleshy lobes at the base of the carpels.

3. Paeon—stems herbaceous ; petals much longer than the sepals ; disc not produced as conspicuous outgrowths.

Paeonies of the Section Moutan are all located in Eastern Asia; this Section is divided into two sub-sections, Vaginatae and Delavayanae.

The Section Onaepia contains two species, P. californica and P. Brownii, which are confined to Western North America and are the only Paeony species indigenous to the American continent. The word Onaepia, invented by Lindley, is an anagram of Paeonia.

The Section Paeon is composed of all the herbaceous species found across the continent? from Spain to Japan. This section is divided into two sub-sections, Foliolatae, those species with distinct entire leaflets, and Dissectifolia, those with leaflets much dissected.

Paeony species fall naturally into Groups under the sections and sub-sections described above. Each Group has been named after the best known species in that Group.

Suffruticosa Group contains P. suffruticosa and a variety found by W. Purdom in 1910 in the province of Shensi, named by Rehder, P. suffruticosa var. spontanea. P. suffruticosa, often known in the past as P. Moutan, covers all the beautiful tree paeonies introduced from China and Japan. Sir Joseph Banks introduced the first plants to this country in 1787. The home of the wild plant is in Szechwan, and seeds of what is probably the wild plant were sent to the United States by Dr. J. F. Rock about 1932. Plants raised from this seed are in cultivation in several gardens in the U.S.A., Canada and the United Kingdom. They are an acquisition to the garden; the flowers are large, single and white with maroon blotches at the base of the petals. It is perfectly hardy and easy to cultivate. P. suffruticosa var. spontanea differs from the species by being smaller, with the apex of the leaflets rounded instead of acute.

Delavayi Group is composed of shrubby Paeony species from the mountains of Western China and Tibet. P. Delavayi is the tallest of this Group and well known in European gardens. P. lutea differs from P. Delavayi by its yellow flowers and by the absence of the involucre before the calyx, so characteristic of P. Delavayi. It is not such a satisfactory plant for the garden as its flowers are often hidden by the foliage, but it has been crossed with P. suffruticosa with great success to produce some fine hybrids. The best of these are P. X ' Esperance,' bred by Messrs. Lemoine of Nancy, and P. x ' Argosy,' bred by Prof. A. P. Saunders, of New York.

The other three Paeonies of this Group are P. Potanini and a white-flowered form and a yellow-flowered variety of this species. P. Potanini was originally discovered by potanin and then collected by wilson, who considered it a variety of P. Delavayi. It differs from the latter by the absence of the conspicuous involucre, by the narrower segments of the leaves, smaller flowers, and in the manner of growth, and therefore should be regarded as of specific rank; this was the opinion of Stapf and Komarov ; Komarov had already described this species under the name of P. Potanini in 1921, so his name stands.

A shrubby, yellow-flowered species was collected by monbeig and forrest on the Mekong-Yangtse Divide and was considered by stapf a new species. He suggested the name ' trollioides ' owing to its Trollius-like flower, but never published the name or description. The differences between this form and the typical P. Potanini do not appear to be sufficient to justify its treatment as a separate species. I prefer to regard it as a variety of the species and have named it P. Potanini var. trollioides. It differs from P. Potanini in that the segments and the lobes are more oblong and shortly acuminate, the yellow flowers do not open widely and are shaped more like the flower of a Trollius, and it has a more erect growth and larger follicles.

All the species of this Group are easily grown garden plants and grow equally well in sun or shade. P. Potanini and its variety increase rapidly by stolons and are particularly effective in a woodland garden.

The species of the Section Onaepia are not of much garden value and are seldom seen in gardens in this country.

All herbaceous species with the exception of the two species in the Section Onaepia are contained in the Section Paeon. In the subsection Foliolatae there are nine groups. The Wittmanniana Group from the Caucasus region, the Russi Group from the Islands in the Western Mediterranean, the Mascula Group from the Continent of Europe, the Broteri Group from Southern Spain and Southern Portugal, the Obovata Group from East Asia and Japan, the Mairei Group from Western China, the Arietina Group from South-eastern Europe and Asia Minor, the Coriacea Group from Southern Spain, Morocco and Algiers, the Lactiflora Group from North-eastern Asia and Northern India.

The Wittmanniana Group contains the only yellow- or yellowish-flowered herbaceous species. Unfortunately a muddle in nomenclature has occurred among the species of this group. P. Wittmanniana was originally described by lindley in the Botanical Register in 1846, where he described the carpels as tomentose. steven, however, two years later described a form with glabrous carpels in Bull. Soc. Nat. Mosc. under the same name. This was again repeated in the Botanical Magazine in 1882. Since then the glabrous carpel form has been known as P. Wittmanniana, and the form with tomentose carpels has been named P. Wittmanniana subsp. tomentosa by busch in 1901 and P. tomentosa by Stapf in the Botanical Magazine in 1931.

To cut a long story short, P. Wittmanniana correctly describes the form with tomentose carpels; the form with glabrous carpels should be known by the earliest varietal name, that is the one of schipczinsky (1921)—P. Wittmanniana var. nudicarpa. These Paeonies only differ by the glabrous or tomentose carpels, so the one should be no more than a variety or sub-species of the other. There is another yellow-flowered species in the Caucasus, P. Mlokosewitschi, whose flowers are of a deeper yellow than the former. This Paeony is a diploid, while the other Paeonies in this group are tetraploids. All these Paeonies are excellent garden plants, perfectly hardy and easy to grow. They flower towards the end of April in the south of England.

The Russi Group consists of P. Russi and two varieties found in Corsica, Sardinia and Sicily, and P. Cambessedesii from the Balearic Islands. P. Russi is distinguished from the two varieties by its densely tomentose carpels. The variety P. Russi var. Reverchoni found in southern Corsica and Sardinia only differs from P. Russi by the glabrous carpels. This is another instance of two Paeonies indigenous to the same geographical area whose main disflaction is the presence or absence of tomentum on the carpels. This phenomenon occurs among a number of Paeony species, and does not in my opinion entitle each Paeony to specific rank but only justifies their classification as varieties. Another Paeony found in Corsica has glabrous carpels. It was named by sieber and described by tausch (1828) P. Corsica ; the differences of this Paeony from P. Russi and P. Russi var. Reverchoni are so small that it appears to be only a variety of P. Russi. I have made a new combination, naming it P. Russi var. leiocarpa (Cosson) following the varietal name of cosson. It differs from P. Russi by the glabrous carpels flushed with red and from P. Russi var. Reverchoni by the deeper green leaflets with reddish nerves, glabrous to slightly hairy below. This variety has been considered by some botanists identical with the other species in this Group, P. Cambessedesii, but this was probably because specimens were rare and because it resembles P. Cambessedesii in the red flush on the back of the leaves and on the carpels. It was collected by Mr. collingwood ingram and introduced by him to cultivation. It was thus possible to examine living plants. P. Russi var. leiocarpa was found to be a tetraploid and P. Cambessedesii a diploid, a finding which clinched the morphological evidence of their separation.

P. Cambessedesii is a most lovely Paeony and the earliest to flower in the garden. It is not very hardy but does well under the protection of a wall. The flowers are a beautiful shade of deep pink, and the whole plant is charming with deep green leaves, purple underneath, set off by the red petioles and stem.

The Mascula Group brings together three European Paeonies which are alike in having entire leaflets', tomentose carpels and unusually glabrous leaves. P. mascula is the well-known P. corallina, a name which has had to be dropped owing to the priority of miller's name of P. mascula (1768). P. corallina was described by retzius in 1783. P. daurica is the second of the Group. This Paeony has been known for many years as P. triternata, a name given by pallas in 1795, but the plant was never described by him. It was described by andrews in the Botanical Repository under the name of P. daurica in 1807. The third is P. banatica, a rare Paeony with a restricted location in Hungary. Though these three species have much in common, they can be easily separated. P. mascula, a tetraploid, has usually nine leaflets on a lower leaf with the terminal leaflet sometimes bifurcating, making the number of leaflets up to thirteen, broadly elliptic with the apex acute. P. daurica has usually nine leaflets which are oval to obicular, with margins undulating and the apex rounded to truncate ; the unusual character of the undulation of the margins of the leaflets make this Paeony easily recognisable. P. banatica has lower leaves like P. mascula and the upper leaves are oblong or narrow elliptic, more like -P. arietina. P. daurica comes from the Crimea and the Caucasus, with outlying locations in Bosnia and on the Island of Samos. The distribution of P. mascula is interesting. It is indigenous to the centre of France and South-east Germany. but there are a number of outlying stations where it is found wild—Steep Holme in the Bristol Channel, Cyprus, Sicily, Syria and Northern Armenia. In all these places there were monasteries in the past and it is possible that these plants are descended from Paeonies grown in the monastery gardens for the medicinal properties credited to Paeonies in the old herbals.

The newly discovered P. kesrouanensis from Syria should come into this Group as it is like P. mascula in the shape and number of the .leaves, but differs from it by the glabrous carpels and long stigma coiled only at the apex.

The best of these Paeonies for gardens is P. mascula. but it is not often seen in gardens. The rose red flowers are attractive and make an admirable contrast to the dark green foliage. It is easy to grow and, like all Paeonies, the flowers last longer when grown in half shade or in open woodland.

P. Broteri from Southern Spain and Portugal appears to be in a Group of its own. It has tomentose carpels and glabrous leaves, but it has nineteen to twenty-three leaflets on a lower leaf. This plant is seldom seen in gardens, probably because it increases slowly.  The shining leaves aad crimson flowers make an attractive combination.

In the Arietina Group there are two Paeonies well known in gardens, P. anetina and P. Bakeri, and two little known Paeonies, P. arietina var. orientalis from Syria and the islands of Cyprus, Crete and Rhodes, and P. rhodia also from the island of Rhodes.

P. arietina from Greece and Asia Minor differs from P. mascula in the greater number of the segments, the hairiness of the undersides of the leaves, petiole and stem and in the size of the segments, which are larger and sometimes wider. The Paeony from Syria mentioned above was named by Monsieur thiebaut P. corallina var. orientalis, but as it only differs from P. arietina in that the leaves, petiole and stem are glabrous, I have made a new combination and named it P. arietina var. orientalis. P. Bakeri is a somewhat mysterious plant. There is no specimen of a wild plant of this species in any herbarium where I have worked. It was described by lynch from a plant growing in the Cambridge Botanic Gardens. It no longer grows there, but was distributed by the late H. J. elwes and is found in several gardens. It is much like P. arietina with only slight differences in the size and shape of the segments of the leaves, but it is a taller and more robust plant. It comes true from seed and is a fine garden plant.

The Coriacea Group is found in Southern Spain and North Africa. P. coriacea is separated from the species in the Mascula and Russi groups by the leaflets which bifurcate into fourteen to sixteen, with lateral leaflets practically sessile. The carpels are glabrous. The variety found in the hills of Algiers only differs by the pubescence on the underside of the leaf, and is named P. coriacea var. atlantica.

The remaining Groups of the sub-section Foliolatae are found in Asia.

The Obovata Group is composed of the Paeonies with broadly oval leaflets and long attenuated carpels.  P. obavata comes from eastern Siberia, Manchuria and Northern China. It is said to have white and rose-coloured flowers, but only the white-flowered form is known in cultivation. In the provinces of Hupeh and Shensi in China there is another beautiful white Paeony which only differs from P. obovata by the amount of hairy tomentum on the back of the leaves, though when grown in the garden in the same aspect and soil this Paeony is a more robust plant with larger flowers and leaves. This plant has been named P. obovata var. Wittmottiae as it differs so slightly from the type. stapf, when he.first saw the plant growing in Miss willmott's garden at Warley Place, considered it a new species and described it in the Botanical Magazine (1916) as P. Wittmottiae. stapf raised it to specific rank owing to the larger flowers and leaflets. villons below, and by the length of the pistil. On examining wild specimens of both forms it would appear that the only difference was the tomentum on the back of the leaflets. The other species in this group is P. japonica from Japan. It differs from P. obovata and the variety, by the flowers which do not open widely, by the more concave petals and short stigma. When the specimens are looked at together, there are to be seen a number of small differences difficult to define. However, when the plants were examined, it was found that P. japonica was a diploid and the other two tetraploids.

P. obovata var. Willmottiae is one of the most beautiful Paeonies for the garden. It flowers in the first days of May with a charming white cup-shaped flower enclosing a mass of golden stamens. It is not very easy to grow, but does best in good loam facing west.  It seeds freely.

There are two species found in Szechwan in Western China, only known in herbaria by a few specimens which must go into a group of their own, to be known as the Mairei Group. P. Mairei and P. oxypetala only differ from each other by the latter having shortly acute or pointed apex to the petals. At first glance these Paeonies might be taken by their long leaves for P. lactiflora. but the leaves lack the minute hairs along the nerves and the papillose margin as in the leaves of P. lactiflora. These species resemble the Obovata Group with their long attenuated carpels. There are specimens of both these species in herbaria with tomentose and glabrous carpels ; carpels in both species, where they are tomentose, are densely covered with a silky tomentum of short golden brown hairs, unlike those of any other Paeony species.

The Lactiflora Group is composed of P. lactiflora and P. emodi and its variety. The Paeonies of this group all have more than one flower on a stem. P. lactiflora has been known for many years as P. albiflora, but pallas, who is responsible for both names, described this Paeony first as P. lactiflora in 1776. P. lactiflora, a native of North-eastern Asia, is well known in gardens ; the flowers have a sweet scent and there are many garden forms. A variety only known from cultivated plants, which only differs by its hairy carpels, has been named with a new combination, P. lactiflora var. trichocarpa (Bunge). P. emodi and its variety P. emodi var. glabrata, which only differs from the type by its glabrous carpel, are found in Chitral and Northern India. P. emodi and its variety differ from P. lactiflora in the leaflets which are larger and not scabrid along the margins, and in the carpel which is nearly always single, rarely two.

The remainder of the herbaceous Paeony species have been placed in the sub-section Dissectifolia in which the leaflets are much dissected into segments. The species in this sub-section are divided into three Groups.

The Peregrina Group contains only one species, P. peregrina, a native of the Balkans and the country around Smyrna. It can be easily distinguished from all other Paeonies by the leaf segments being lobed and coarsely toothed at the apex and also by the concave petals. clusius in the sixteenth century first drew attention to the great red Paeony of the Balkans and described it as P. byzantina. Much confusion has arisen over the name of P. peregrina, and several species have been given this name. How this confusion arose is fully described in the " Study " in the chapter devoted to the History of Paeony Literature. This beautiful Paeony goes by several names in gardens : P. decora, P. lobata, and ' Fire King ' and ' Sunbeam.' P. peregrina is perhaps the finest wild herbaceous Paeony for the garden. The great scarlet flower with its concave petals is a wonderful sight in the half shade under trees in the evening light.

The Officinalis Group is composed of a number of species, distinguished by having the leaflets deeply cut into twenty-one or more segments which are elliptic, oblong-elliptic or oblong. The species and varieties extend from Crete up the Dalmatian coast to Northern Italy and Switzerland and along the French coast to the centre of Spain.

The best known species in this group is P. officinalis. The name, which was first given to a Paeony by linnaeus, has been used in the past for several different species.  It was necessary therefore to find out which species linnaeus had in mind when he referred to P. officinalis. linnaeus originally " lumped " all the species together, but later in Species Plantarum (1753) he divided them into two forms, « femina and P mascula, giving references under each form to the works of earlier botanists. It was found by reference to the earner literature that the plant linnaeus had in mind' was the plant from Mt. Generoso, near Lugano. The history of this nomenclature is fully explained in the chapter on Paeony Literature. P. officinalis is found mainly in Northern Italy and Switzerland, and is distinguished by leaflets as described above and by the densely tomentose carpels. In the centre of Spain there is a smaller Paeony, P. humilis, which differs from P. officinalis by the leaflets being shorter and broader and not so deeply dissected and has glabrous carpels. Between Spain and Northern Italy another Paeony is found only differing from P. humilis by the tomentose carpels and villose stem. This species was named by anderson (1818) P. paradoxa, but as it so closely resembles P. humilis, it has been treated as a variety and a new combination formed with the name of P. humilis var. villosa (Huth).  This Paeony is known in the old literature as the ' Saint Loup ' Paeony. It is also found in the centre of Italy. These Paeonies from Italy and Southern France have been the most difficult to disentangle as many names had been bestowed on them by botanists in the past. There is also another Paeony named by anderson (1818) P. mollis which comes into this Group. It had been suggested that this Paeony came from the south of France. No wild specimens are preserved in any herbaria examined, but only specimens from gardens. It would appear, therefore, to be of garden origin. This Paeony differs from the other species in the group by its upright stalk, sessile leaves, and the flower being sub-sessile, on a short stalk. There is also a white form known in gardens as P. sessiliflora, a name given by sims in 1826. These plants are not of much account as decoration for the garden. The true P. officinalis is not often seen in cultivation, but P. humilis and the variety villosa are common. They are pleasant plants and seed themselves freely. By far the most attractive of the Paeonies in this group is P. Clusii from the Island of Crete. It grows about one foot high with finely-cut foliage and a white cup-shaped flower with golden stamens, a very delightful plant. It is not easy to grow and requires a sheltered place. It differs from P. officinalis by the smaller and narrower leaf-segments and the white flowers. The well-known P. tenuifolia has been included in this group but perhaps occupies rather an isolated position. It has been included because the leaflets are cut into many segments, but the segments are, narrower than those of any other Paeony species. It has a wide distribution from Transylvania to the Caucasus. The plant is very attractive with its fern-like foliage and deep red flowers and as it does not grow too rampantly or too high, is worthy bf a place in the rock garden. There is a well-known hybrid of this plant with P. lactiflora called P. x Smouthii which is a good garden plant. It takes after P. lactiflora in having more than one flower to a stem and in having a sweet scent. It may have been called after Monsieur smout, a chemist in Malines who was a keen hybridizer of plants and lived about 1840.

The Anomala Group, the last in this sub-section, is an interesting one as it contains species with only one flower to a stem and also species with several flowers to a stem. This character is the main difference between P. anomala and P. Veitchii, and appears to me to be of specific value. The leaflets of both species are much dissected into a number of leaf segments which are. again divided and have on all the main veins on the otherwise glabrous upper side of the leaflets, uninterrupted lines of delicate small hairs. P. anomala and P. anomala var. intermedia have a wide distribution extending from north-east European Russia, through central Asia to Turkestan. P. anomala is not often seen in gardens, but the variety is fairly common. It is a very beautiful plant with a large crimson flower.

P. Veitchii and its variety P. Veitchii var. Woodwardii are confined to a comparatively small area in the provinces of Szechwan, Kansu and Shensi in Western China. P. Veitchii var. Woodwardii, which is a new combination, differs from the species in its smaller stature, and in its leaflets which have longer bristly hairs on the midrib and nerves above, also prominently on nerves and midrib below ; the petioles and petioiules are also markedly bristly. Stapf described this plant as a new species under the name of P. Woodwardii, but he never published the description as he had some doubt whether it should be considered of specific rank. The differences between this Paeony and P. Veitchii do not appear to me to be sufficiently great to raise it to specific rank, but rather it would seem to be only a variety of P. Veitchii. Both these Paeonies flower later than P. anomala and its variety, usually towards the end of June. They are of the easiest cultivation. P. Veitchii has magenta red flowers and the variety rather pinker flowers. There is also a form of the variety with white flowers.

In the " Study " the cytological aspect of the genus and its relation to the geographical distribution has been fully discussed and raises a number of interesting speculations on their possible explanation.

It was found to be necessary to go through the whole of the literature of the genus in order to obtain the information as to how each name had arisen. The examination of all the literature was an engrossing study, and as it might be of interest and useful to future students of the genus, a short history of Paeony literature has been included in the " Study " as a separate chapter. All the Paeony species which have not already been figured have been illustrated with colour plates by Miss lilian snelling. The drawings of the leaves have been made by Miss stella ross-craig.