A deciduous shrub 8 to 15 ft high (occasionally a small tree twice as high in the wild); young branches stellately downy at first. Leaves elliptic or obovate, 3 to 51⁄2 in. long, 11⁄2 to 3 in. wide, minutely and rather distantly toothed, abruptly pointed, wedge-shaped or rounded at the base, downy on both sides on first opening, but soon almost glabrous except on the midrib and veins beneath; stalk 1⁄2 to 3⁄4 in. long. Flowers pendulous, produced in May in clusters or short racemes from the joints of the year-old wood; stalks 1⁄2 to 3⁄4 in. long, slender, downy. Corolla bell-shaped, 3⁄4 in. long, deeply four-lobed, white; calyx very downy; stamens hairy. Fruits oblong, 11⁄2 to 2 in. long, 3⁄4 in. wide, with two longitudinal wings 1⁄4 to 3⁄8 in. wide, ending in a short spike.
Native of the south-eastern United States; introduced in 1758. This is far from being as good a garden plant as H. carolina or monticola; it is less hardy and is shy-flowering. It grows well, and is over 12 ft high at Kew, but never flowers as it does in France, especially south of Paris. Easily distinguished from H. carolina by the two-winged fruits.
The tree at Kew mentioned above was 20 ft high in 1946 but died a few years later, when about fifty years old. It bore flowers about 3⁄4 in. long, as described above, which seems to be about the upper limit for the normal state of H. diptera.
This species received an Award of Merit when shown from Borde Hill, Sussex, in 1948. This plant still exists, but does not thrive and rarely flowers. It was raised from seeds collected from cultivated trees on the Vanderbilt Estate at Biltmore, N. Carolina, by the New York Botanical Garden Expedition to the Appalachians (1933) under field number NY 157. The trees there were said to be variable in size of flower.
var. magniflora Godfrey – This is distinguished from the typical variety by its larger flowers, 3⁄4 to 11⁄4 in. long, and by its different habitat – upland woodland in contrast to the flood-plain woodland where the typical variety is found (Kurz and Godfrey, Trees of Northern Florida, pp. 263, 265 and fig. 168; and see also the discussion by R. de Belder in Journ. R.H.S., Vol. 94 (1969), pp. 90-91). M. de Belder states that a floriferous tree in the Kalmthout Arboretum, Belgium, belongs to this variety and also refers to it the tree in M. de Vilmorin’s collection at Verrières, suggesting that it is this tree that Mr Bean had in mind when he said that H. diptera flowered well in France ‘especially south of Paris’, which is almost certainly the case. It is possible that the variety magniflora would prove a greater success in Britain than typical H. diptera, coming as it does from upland locations.