Hamamelidaceae, the witch-hazel family, includes approximately 30 genera representing around 100 species of deciduous trees and shrubs. Members of the family are found in both temperate and tropical regions of North and Central America, Eastern Asia, Africa, the Pacific Islands, and Australia. The Arnold Arboretum has a rich history with the family, from plant exploration to the naming and introduction of its members to cultivation. The Arboretum’s Hamamelidaceae collection, which currently comprises ten temperate region genera, can be found in groupings throughout the Arboretum landscape. Specific locations include the area around the Hunnewell Visitor Center, the Leventritt Shrub and Vine Garden, scattered among the trees in the North Woods, on the edges of the hickory (Carya) collection, near the summit of Bussey Hill, and among the jewels of the Explorers Garden.
As autumn arrives at the Arboretum, the flowering season for the witch-hazel family begins, and will carry through until spring. Starting in October, common witch-hazel (Hamamelis virginiana)—a New England native—begins to bloom, the straplike yellow petals of its fragrant flowers extending on warm days and curling up when temperatures drop near freezing. This show can persist into December even as the snow begins to fall. Other members of the witchhazel genus represent the earliest of bloomers, starting in January and lasting well into March—a remarkable sight in the depths of winter.
As the ground begins to warm in April, several species of Corylopsis—commonly called the winter-hazels—produce many pendulous clusters of bell-shaped yellow flowers. The fothergillas (Fothergilla) round out the family’s flowering season in the Arboretum with their bottle-brush-like white blooms in May. Beyond the showy flowering of these genera, many are also aesthetically valuable for their unique foliage, vibrant fall colors, and, in the case of Parrotia, attractive exfoliating bark. Given these attributes, perhaps no other plant grouping holds greater ornamental potential and yet is so underutilized in today’s landscape than the witch-hazel family. This two-part article explores various historical, taxonomic, and horticultural facets of Hamamelidaceae taxa in the Arboretum’s collection. We begin with Hamamelis, the genus for which the family is named.
Whilst winter’s hand is yet heavy on the land the Witch-hazels boldly put forth their star-shaped yellow blossoms but the native Hamamelis vernalis is over-shadowed by its more brilliant Chinese and Japanese relatives.Ernest H. Wilson, Plant Hunting, 1927
Witch-hazel (Hamamelis) is the most well-known Hamamelidaceae genus among gardeners and includes the only native New England representative of the family—Hamamelis virginiana, the common witch-hazel. There are two other North American species, H. vernalis (vernal or Ozark witch-hazel) and H. ovalis (big-leaf witch-hazel), and two Asian relatives, H. mollis (Chinese witch-hazel) and H. japonica (Japanese witch-hazel). All of these species are shrubs or small trees inhabiting temperate regions. They share characteristically narrow, straplike flower petals and capsulate fruit that is explosively dehiscent, capable of ejecting seeds as far as 10 meters (33 feet). Much work has been done to create hybrids (H. × intermedia) between the Chinese and Japanese species, resulting in the development of horticulturally desirable selections. Today these hybrids, as well as cultivars of H. mollis, are the witch-hazel family members most popular with American gardeners. Both the North American and Asian witch-hazels have a rich history, with fascinating stories of discovery and much horticultural potential.
North American Discoveries
Common witch-hazel has a wide-ranging native distribution along the east coast from Nova Scotia to Florida and west to the Mississippi River, petering out in the Ozarks. Its western limit runs from eastern Texas to Minnesota. It is commonly found in forest understories as a large multi-stemmed shrub. For non-gardeners, witch-hazel may be a familiar name not as a forest-dweller but for its use as a component in first-aid and skincare products. Native Americans used witch-hazel for its healing properties, and open-minded New Englanders soon recognized its potential. In 1866, the first commercial witch-hazel extract distillery was founded in Essex, Connecticut, by Thomas Newton Dickinson. Today, the distilling facility is located in East Hampton, Connecticut, and is the world’s largest source of witch-hazel extract, still produced from witch-hazel wild-harvested from New England’s woodlands.
The flowering time of common witch-hazel is definitely unique. Just as it seems that the last of the years’ blooms have faded, H. virginiana comes into flower. Depending on the specimen in question, blooms start as early as October and can last into December. The species’ fragrant flowers are composed of four yellow, straplike petals that furl and unfurl with the temperature swings of late autumn. In many cases, full bloom occurs when the plant’s yellow fall foliage is still present, making it difficult to appreciate the flowers’ full grandeur. This is viewed by some as an aesthetic fault of the plant, but to those who know what to look for, it is quite a remarkable display. With nothing else in bloom, the common witch-hazel has little competition for pollinators seeking a late season food source.
The Arboretum’s largest concentration of H. viriginiana can be found in the North Woods, just past the Aesculus collection along Meadow Road, uphill from the short stretch of post-and-rail fence. A visit to explore this nook should be part of any autumn walk in the Arboretum. A bit farther down Meadow Road, at the northern end of Rehder Pond, is the Arboretum’s oldest accession (14693-D) of common witch-hazel, wild-collected as a plant from western Massachusetts and brought back to the Arboretum in 1883 by Jackson Thornton Dawson, the Arboretum’s first plant propagator.
Steps away from this specimen grows another one of the Arboretum’s centenarian witchhazels, Hamamelis vernalis accession 6099-D. Unlike common witch-hazel, the vernal witchhazel, as the name suggests, flowers very early in the year (January through March). Although its geographic range overlaps with that of common witch-hazel, it only grows natively in the Ozark highlands of Missouri, Arkansas, and Oklahoma, and in small populations in Texas and Louisiana. A truly grand representation of the species, accession 6099-D was wild-collected as a seedling in Missouri and sent to the Arboretum in 1908 by Benjamin Franklin Bush under the consignment of Charles Sprague Sargent, the Arboretum’s first director.
At the time of this plant’s collection, H. vernalis had yet to be officially named and described by science, although, as herbarium records show, it was found growing in Missouri by Saint Louis botanist Dr. George Engelmann as early as 1845. Nonetheless, common witchhazel was the only identified North American species at this point. In fact, Bush authored the 1895 publication A list of the trees, shrubs and vines of Missouri in which H. virginiana is mentioned as the sole representative of the genus. Sargent’s 1890 publication The Silva of North America, a description of the trees which grow naturally in North America exclusive of Mexico, made the same conclusion.
The story of vernal witch-hazel’s discovery begins with Sargent’s and Bush’s plant explorations in Missouri and Arkansas in September–October of 1907, the main goal of which was to search for new Crataegus (hawthorn) species. On October 8, 1907, the explorers collected a herbarium specimen in Swan, Missouri, of a Hamamelis in fruit, but lacking flowers; this certainly sparked their curiosity since it appeared dissimilar to the known fall-blooming species, H. virginiana. Returning to Boston, Sargent anxiously requested of Bush that he return to Missouri to collect seeds and flowering herbarium specimens that winter. In a letter to Bush dated January 22, 1908, Sargent wrote, “Are you doing anything about the flowers of that Southern Missouri Hamamelis? I am very anxious to get these this spring if possible and I am counting on you to do it, either through our friend at Swan or through your brother.” Sargent received his first flowering vouchers of the suspicious witch-hazel on March 14, 1908, and wrote:
Dear Mr. Bush:
I am very much obligated for the Hamamelis specimens which arrived today. They were gathered a little too soon and if you had only put them in water a few days before pressing then the flowers would have fully expanded. I think there is no doubt, however, that this is an undescribed species. We want to describe and publish a figure of it in an early number of Trees and Shrubs, so I hope you won’t “give it away” to anyone else … We must manage to get some young rooted plants of the Hamamelis as none of the seeds we got last autumn were good. Apparently after they were gathered they were destroyed by the weevil …
Yours very truly, C. S. Sargent
In October 1908, the Arboretum did receive and accession (6099) the rooted plants from Bush, as Sargent requested. Additional herbarium specimens of the wild plants in flower were received in February 1909 and are also held in the Harvard University Herbaria; they were undoubtedly sent to Sargent by Bush as he developed his new description of the species. In Sargent’s 1911 publication, Trees and Shrubs, Illustrations of New or Little Known Ligneous Plants, he first described H. vernalis as a new species:
The different species of Hamamelis offer no good morphological characters, the structure of the flowers, fruit and seeds being the same in them all. The plant, however, from southern Missouri, Arkansas and Louisiana is so distinct in its time of flowering, in the bright red color of the inner surface of the calyx-lobes, in the pale color of the under surface of the leaves, and in the amount and persistency of the pubescence on the leaves and branches that it appears desirable to distinguish it specifically from Hamamelis virginiana. The habit, too, of spreading by stolons into great thickets, and the fact that it grows so far as I have seen it only in the gravelly beds and margins of streams, also seem to separate it from the eastern species, which inhabits rich woodlands and upland pastures. In the color of the inner surface of the calyx-lobes and in its time of flowering Hamamelis vernalis resembles the Japanese species.
The individual specimen 6099-D from this original collection, which was planted and remains in its location near Rehder Pond, was first noted in bloom on January 15, 1913, by Ernest Henry Wilson, Arboretum plant explorer and “Keeper” of the Arboretum following Sargent’s death in 1927. The news of the bloom was reported in the Arboretum’s Bulletin of Popular Information that spring: “Hamamelis vernalis is an interesting plant with considerable decorative possibilities. It is a native of southern Missouri and, although the existence of a Witch Hazel in that part of the country has long been known, it has only recently been distinguished from the autumn flowering species of the northern states. This Missouri species flowered this winter in the Arboretum for the first time in cultivation and is still little known in gardens.”
In the 1920s, Alfred Rehder, renowned Arboretum taxonomist, described two variations (forms) of the species: H. vernalis f. tomentella, with pale, pubescent leaves, and H. vernalis f. carnea, with reddish petals. Specimens of both these forms (accessions 18885-A and 18886- A, respectively) can be found just north of the Hunnewell Visitor Center.
Astonishingly, almost one hundred years after Sargent named H. vernalis, a third species of North American witch-hazel was named. Hamamelis ovalis, known commonly as bigleaf witch-hazel, has a known range that is limited to a handful of counties in southern Mississippi and Alabama. It was discovered in July 2004 during a botanical inventory of a National Guard training site in Perry County, Mississippi. Officially described in 2006, this species has several distinctive traits including varying red petal coloration, musty floral scent, relatively large pubescent leaves, and the habitat in which it is found. Although known only to the far south, it carries the hardiest aspects of its relatives, growing remarkably well here at the Arboretum. Three specimens can be found planted in the collections, including accessions 113-2009-A and 114-2009-A in the Leventritt Shrub and Vine Garden. The latter of the two in particular expresses the large leaves for which the species is commonly named.
The potential horticultural merits of our native witch-hazels have long been recognized, yet remain underutilized. It is fairly rare to see these species grown in cultivation outside of botanic gardens and arboreta. Several cultivars of these species have been introduced, probably the most commonly seen being H. vernalis ‘Autumn Embers’, named for its noteworthy burgundy red fall foliage. Among the North American species and their variants growing at the Arboretum, there are three cultivars of vernal witch-hazel—‘Lombart’s Weeping’, ‘Orange Glow’, and ‘Sandra’—and one of common witch-hazel, ‘Champlin’ (synonym ‘Champlin’s Red’).
Witch-hazels of Asia
In 1907–1908, as Sargent worked with Bush to describe Hamamelis vernalis, E. H. Wilson was collecting living plants and seeds of Chinese witch-hazel (H. mollis) in central China under the sponsorship of the Arboretum. Wilson wrote of his explorations:
April 21, 1907
Dear Professor Sargent,
… I visited a part of the mountainous region to the West-south-west of Ichang [Yichang], a part where I had not previously been. Reaching an altitude of about 7,000 feet, the woods in the mts. were still as dormant as in mid-winter, and the snow was still lying in the crevices. In the ravines & open valleys vegetation was advancing, and I made a collection of about 180 species of trees and shrubs … Of shrubs, in the Mt. Hamamelis mollis was the most striking with its wealth of yellow flowers, on the low hills. Loropetalum chinensis [another member of the family] was a sight for the Gods …
With kindest regards,
I am, Dear Professor,
Faithfully and obediently yours,
E. H. Wilson
Although still quite underutilized in our cultivated landscapes, the Asian witch-hazels and their hybrids have certainly received greater horticultural attention. Chinese witch-hazel is the least hardy of the species but is also the showiest, with bright yellow petals and red calyx-lobes, and fragrance well beyond the others. The previous season’s leaves on H. mollis can be persistent, appearing dried out, brown, and obscuring the blooms even in late winter. It is found growing natively in the forests and thickets of central and eastern China. Japanese witch-hazel (H. japonica) differs from Chinese witch-hazel in several ways; it has a more flat-topped form, blooms slightly later, has cold-hardier flower buds, and its flowers have slightly longer and wavier petals but are often produced less abundantly and are muted in color. As the name suggests, H. japonica is endemic to Japan. The fall foliage color of both species can be quite spectacular.
Sargent himself collected seed of Japanese witch-hazel in 1892 as he explored the island’s flora. He wrote observations of the encounter in his 1894 publication, Forest Flora of Japan:
The Japanese Hamamelis … is already an inhabitant of our gardens, where, unlike the American species which flowers in the autumn, it produces its orange or wine-colored flowers in March [H. vernalis had not yet been described]. Hamamelis japonica is one of the common forest-shrubs or small trees in its native country, where specimens occasionally occur thirty or forty feet in height, with stout straight trunks and broad shapely heads. In autumn the leaves turn bright clear yellow; but on one form which we found on Mount Hakkoda, near Aomori, with small thick often rounded leaves (Hamamelis arborescens of Hort., Veitch), they were conspicuous from their deep rich vinous red color. This may, perhaps, prove to be a second Japanese species.
As Sargent’s observations in the wild suggest, the fall color can be quite variable, with combinations ranging from yellow to purple. None of the witch-hazel representatives from Sargent’s 1892 voyage remain in the collections today, but a rather stately, vase-shaped specimen of H. japonica (accession 475-90-A) can be found growing in the Leventritt Shrub and Vine Garden.
As well, while none of the witch-hazels Wilson collected directly for the Arboretum remain in our collections today, the lineage of his 1907–1908 voyage lives on in a very significant way. On February 21, 1908, Wilson wrote Sargent detailing the contents of cases of plant material that he was sending to the Arboretum, and explained: “The uncertainty regarding the arrival of these plants in a living state makes one anxious. Should fortune favor us and they arrive in a satisfactory condition you will possess many plants of more than ordinary interest and which are worth much from the scientific and arboricultural standpoint. If it fails we must try again on other lines.” Two months later the cases did arrive, but the contents were in poor condition as he feared. Of the news, Wilson wrote “I need not enter into my feeling of bitter disappointment and vexation … In slang language I was knocked all of a heap.” Although the mortality in Wilson’s early shipments to the Arboretum may have been high, some material did survive the journey. Among the survivors were seeds from H. mollis collected in Changyang Hsien, Hubei, under Wilson Collection Number 624, later becoming Arboretum accession 14691.
After accession 14691 had grown in the collections for a number of years, William Judd, propagator at the time, collected open-pollinated seed from this remarkable specimen of Chinese witch-hazel in 1928. The plant was growing in close proximity to other witch-hazels in the Arboretum’s collections. Germinated the following spring, seven of the seedlings would eventually be planted in the collections carrying with them accession number 1173-28. Two of these original seedlings (1173-28-A and G) can still be found growing on the grounds today. In the years that followed, these plants were under careful observation, as several herbarium vouchers from the mid-1930s in the Harvard University Herbaria attest. Through such observations, it was determined that the open-pollinated nature by which the seeds were produced led to none of the seedlings being true H. mollis, but that they were in fact hybrids between H. mollis and H. japonica, displaying traits of both parent species. In 1945, Rehder named the hybrid Hamamelis × intermedia, given the “intermediate” traits of the parents exhibited in the new hybrid.
The Best of All in Flower
Witch-hazels seem to be the true harbingers of spring … However the Japanese witch-hazel has not proven an outstanding plant in bloom because the flowers are not profusely borne and mixed in color with some red, which detracts from the brilliance of the color display in early spring. On the other hand, the Chinese witchhazel, long noted as a good and fragrant blooming plant, has proved disappointing many years in the Arnold Arboretum because the flower buds have been killed by cold winter.Donald Wyman, 1963
One of the Judd hybrid seedlings (accession 1173-28-B) had been planted beside the Hunnewell Building, and as it grew it was noted as exceptional among its siblings. In a plant records entry from March 24, 1959, Arboretum horticulturist Donald Wyman referred to it as the “best of all in flower.” It captured all of the best floral traits of its maternal parent (H. mollis) with profusely borne, fragrant, bold yellow blossoms, as well as desirable characteristics of H. japonica, including better winter hardiness, larger petals, and less leaf retention through the winter. It also displayed signs of hybrid vigor, with a more upright form compared to its spreading parents. In the October 25, 1963, issue of Arnoldia, Wyman announced that a new clonal cultivar had been registered: Hamamelis × intermedia ‘Arnold Promise’. He compared the plant to an “old friend,” which could be observed out the windows of the library and herbarium and was known for its performance and counted on because it had been there a long time, yet was not “unusual” to the people who got used to enjoying it on a continuous basis. Only after several well-traveled visitors called special attention to the specimen was the plant considered for introduction. Although the original plant no longer graces the Hunnewell Building, a cutting taken from it in 1969 (accession 396-69-A) was grown out and planted near the original location in 1979. It survived a temporary relocation to the nursery in 1992 during the renovation of the Hunnewell Building and in 1995 it was returned to the same spot where it still thrives today. Though there are now a great number of H. × intermedia cultivars that have been introduced to the horticultural industry, ‘Arnold Promise’ remains among the leaders in the garden world.
Along with the species accessions, the Arboretum’s Asian witch-hazel holdings include a number of introduced cultivars. There are currently three Chinese witch-hazel cultivars in the collections: ‘Brevipetala’, ‘Pallida’, and ‘Princeton Gold’. And in addition to the 14 ‘Arnold Promise’ specimens that adorn the grounds, six other cultivars of H. × intermedia can be found throughout the landscape, including yellow-petaled ‘Moonlight’ and five others selected for their unique petal coloration in varying hues of red and orange: ‘Diane’, ‘Feuerzauber’, ‘Hiltingbury’, ‘Jelena’, and ‘Ruby Glow’.
Continuing a Legacy of Discovery
Collection, evaluation, and scientific study of witch-hazels continues at the Arboretum. An accession of particular interest and value is a Chinese witch-hazel that was wild-collected in Wudang Shan, Hubei, China, as part of the 1994 North America–China Plant Exploration Consortium (NACPEC) expedition. One of the expedition’s goals was to collect farther north in Hubei than Wilson ever had, with the hope of bringing hardier material into cultivation. The trip’s plant explorers, including former Arboretum Senior Research Scientist Peter Del Tredici, describe their discovery of a witch-hazel grove in fruit: “A little way beyond the Zelkova shrine, we found several plants of Chinese witch hazel … loaded with unopened seed capsules. We were particularly pleased to collect this winter-blooming species, which has been gaining popularity in American gardens. After seeing so many plants without seed, it was a treat to find one in fruit, and we greedily collected every seed capsule we could find. The plants were growing on a dry, shady hillside near another plant in the witch hazel family, Sinowilsonia henryi …” Two individuals from this collection (697-94-A and 698-94-A) can be found growing on either side of Meadow Road adjacent to the maple collection.
These and other Hamamelis mollis specimens are currently part of an investigation by Jessica Savage, a Putnam Fellow at the Arboretum, examining what allows Chinese witchhazel and other precocious flowering plants to produce flowers early in the year before they develop new leaves. Plants that flower later in the season can use resources provided by their leaves to support their floral displays, but precocious flowering plants depend on nutrients stored in their stems. Her research will help us understand how plants like witch-hazel access the resources required for blooming and overcome the challenges of flowering, in some cases, while the ground is still frozen.
As I write this passage, Michael Dosmann, the Arboretum’s Curator of Living Collections, is on an expedition in the Ozarks with several botanical colleagues. Hamamelis vernalis is on the group’s list of targeted species for collection. The seeds and stories he brings back from his journey will most certainly add to the rich history of witch-hazel at the Arboretum and deepen our understanding of this exceptional genus.
Sidebar | All in the Family
The Arnold Arboretum currently has living specimens representing these genera within Hamamelidaceae:
Arnold Arboretum. 1913. Bulletin of Popular Information. No. 37.
Bean, W. J. 1973. Trees and shrubs hardy in the British Isles, 8th ed. Vol. 2. London: John Murray, Ltd.
Bush, B. F. 1895. A list of the trees, shrubs and vines of Missouri. Jefferson City, Missouri: Tribune Print. Co.
Connor, S. 1994. New England natives: A celebration of people and trees. Cambridge: Harvard University Press.
Del Tredici, P., P. Meyer, H. Riming, M. Cailiang, K. Conrad, and R. W. Thomas. 1995. Plant collecting on Wudang Shan. Arnoldia 55(1): 12–20.
Leonard, S. W. 2006. A new species of witch-hazel (Hamamelis: Hamamelidaceae) apparently endemic to Southern Missouri. Sida, Contributions to Botany 22: 849–856.
Meyer, F. G. Hamamelidaceae. 1997. In: Flora of North America Editorial Committee, eds. 1997. Flora of North America North of Mexico. New York and Oxford: Oxford University Press. Vol. 3: 362–367.
Rehder, A. 1945. Notes on some cultivated trees and shrubs. Journal of the Arnold Arboretum 26: 67–78.
Rehder, A. 1920. New species, varieties and combinations from the herbarium and the collections of the Arnold. Journal of the Arnold Arboretum 1: 254–263.
Rehder, A. 1928. New species, varieties and combinations from the herbarium and the collections of the Arnold Arboretum. Journal of the Arnold Arboretum 4: 29–31.
Sargent, C. S. 1894. Forest flora of Japan: Notes on the forest flora of Japan. Boston: Houghton, Mifflin and Company.
Sargent, C. S. 1890. The Silva of North America, Vol. 5. New York: Peter Smith (reprint 1947).
Sargent, C. S. 1913. Trees and shrubs: Illustrations of new or little know ligneous plants. Vol. 2. Boston: Houghton Mifflin Company.
Weaver, R. E. Jr. 1981. Hamamelis ‘Arnold Promise’. Arnoldia 41(1): 30–31.
Weaver, R. E. Jr. 1976. The witch hazel family (Hamamelidaceae). Arnoldia 36(3): 69–109.
Wilson, E. H. 1927. Plant hunting. Vol. 2. Boston: The Stratford Company.
Wyman, D. 1963. New plants registered. Arnoldia 23(9): 111–118.
Andrew T. Gapinski is Manager of Horticulture at the Arnold Arboretum.