A hybrid intermediate between its parents, most likely A. forrestii and A. homolepis, or perhaps a near ally of A. forrestii such as A. georgei. (See below).
Growing just to the north of the Royal Botanic Garden Edinburgh’s former west gate (now the John Hope Gateway building) is a curious fir bearing the accession number 19715734*A. It recalls Abies homolepis in several features, most especially the bark and the crown, so too something about its foliage, but its branches are much more widespreading and on closer inspection the foliage is wrong, too: the needles are longer and broader, and not the typical minty green colour common to almost all forms of A. homolepis. Most years this tree produces masses of male and female cones; again both are too long to be A. homolepis. Few trees at the RBG Edinburgh have had their labels changed as often as this, such has been the diversity of theories as to its identity. Its origins remain a mystery, but the database gives the original name as A. delavayi var. georgei (here A. georgei), which it is not. At some point it was reappraised and for a long while it was labelled A. homolepis var. umbellata (here A. × umbellata), again in error for it is not that taxon either, and over time this fueled misconceptions about A. × umbellata in cultivation, at least in the UK.
Michael Frankis has proposed a theory that this, and other trees like it, are hybrids between A. forrestii and A. homolepis. Frankis first noticed trees of this type during a visit to Foxhill Arboretum, Cheshire, in c. 1980. ‘There was a group of three young trees there (4–5 m tall, ~15–20 years old?), which had cones just about exactly intermediate between A. forrestii and A. homolepis; the foliage was also intermediate’ (M. Frankis pers. comm. 2020). A tree growing in the pinetum at Gosford Castle, Co Armagh, Northern Ireland, was later identified as the same hybrid (Tree Register 2020) and it has cropped up occasionally in other collections, too: ‘there seem to be enough to suggest that some nursery…had raised a batch of seedlings and sold them to a number of gardens, so there may be more yet waiting to be found’ (M. Frankis pers. comm. 2020).
Whilst the influence of A. homolepis is clear to see, the only documentary clue we have to the other parent is that RBGE originally received their tree as A. georgei (as treated here), so presumably seed was obtained from a cultivated tree bearing that name, but which had hybridised with a nearby planting of A. homolepis. This is not as helpful a clue as it may seem, as the name A. georgei (or any permutation of it) is inherent in a perplexing group of closely related firs distributed in southwest China, which includes A. forrestii, and which are notoriously difficult to identify regardless of the taxonomy one lays over them. There is no guarantee that the parent tree was correctly identified by the original supplier; it surely belongs to that group of taxa, but this uncertainly would more accurately be conveyed by referrring to these hybrids as A. forrestii agg. × homolepis.
The Edinburgh tree has been propagated over the years and there are now scions in various locations – usually bearing the erroneous name A. homolepis var. umbellata. It is a distinctive form and should probably be named, for the original is ‘an aesthetically beautiful tree’, extremely hardy and endowed with ‘phenomenal vigour and resilience’ (C. Pirnie pers. comm. 2020). Scions were distributed by Callum Pirnie to some very cold collections, including to Ashmore House in Highland Perthshire and to Crathes Castle in Aberdeenshire, both characterised by cold winters and cool summers; it is thriving in both locations. It was also propagated and distributed by the late Derek Spicer (C. Pirnie pers. comm. 2020). Other forms of this hybrid are grown in several collections, for example at the Yorkshire Arboretum, though none have yet been found that can rival the beauty of the Edinburgh form.