Sorbus ellipsoidalis McAll.

As with several members of section Discolores described recently by Hugh McAllister, Sorbus ellipsoidalis was recognised in cultivation. In this case the trees were found growing at Bodnant, Conwy, in 1982, and were confirmed to have come from the collection McLaren C288. When George Forrest died in 1932 his experienced collecting team was kept on for another season through the sponsorship of H.D. McLaren (later Lord Aberconway), and seed gathered in that year is usually attributed to ‘McLaren’s collectors’. Among the seed and specimens they gathered was this tree, and the type specimen at Edinburgh bears the note ‘half way up Sung Kuei Mountain’ (Yunnan). It remains the only collection known, though as a tetraploid apomictic microspecies S. ellipsoidalis is easily propagated from seed. It is closely related to S. glabriuscula (see p. 803), but differs in being more slender in appearance, with narrower leaflets and an elongated (ellipsoid) fruit. A more technical distinction is the presence of a white patch around the micropyle of the seed, about one-quarter the length of the seed (McAllister 2005a). In other regards, with its almost perfectly white ripe fruits, it is very similar to S. glabriuscula – and apparently its equal for garden performance, though with distinctly more delicate foliage. Howick 1960, collected at c.3990 m on the southwestern side of the Wuming Pass, Sichuan in 1994, is comparable and may become common in cultivation as it too has almost pure white fruits, and similarly delicate foliage.