Pomology and the identification of apples

A grafted fruit variety is a clone. Every tree of a cultivar is genetically the same individual, propagated by scion wood, and that is what makes a variety name meaningful in the first place: the name is supposed to designate one particular tree, indefinitely multiplied.

Naming practice never lived up to that. Varieties were named locally, after the farm, the parish or the person who raised them, and the same fruit acquired different names a few miles apart while different fruit acquired the same name in different counties. Nurseries propagated under whatever name their customers used, mislabelled stock circulated, and a variety could be lost and renamed within a generation.

The result by the early nineteenth century was a nomenclature in which nobody could be confident that two trees sold under one name were the same thing. That is not a cosmetic problem: a grower planting on the strength of a description needs the tree to be the variety described, and a cider maker choosing fruit for acid and tannin needs the same assurance.

The response was systematic description supported by illustration. A pomona is a work that describes named fruit varieties in a fixed order — origin, tree habit, fruit shape and size, skin, eye, basin, stalk, cavity, flesh, season, use — and provides a plate accurate enough to compare a real fruit against. The genre developed across Europe from the late eighteenth century and reached its most elaborate form in the second half of the nineteenth.

The English works that matter for cider are the fruit manuals and county pomonas of the middle and later nineteenth century, of which the illustrated Herefordshire work produced between the 1870s and 1880s under the direction of a local horticultural society is the most important for cider fruit and perry pears specifically. It combined descriptive text with coloured plates and treated cider varieties as worth the same attention as dessert fruit, which was not the norm.

The Continental tradition ran in parallel and on a larger scale. Multi-volume French dictionaries of pomology from the same decades attempted to describe every variety in cultivation, with synonymy lists that are still consulted, and German and Belgian works did comparable jobs for their own regions. The genre’s ambition was total coverage, and its besetting problem was that no author could see every tree.

Description alone cannot settle synonymy, because two descriptions written by different people from different fruit will differ even when the variety is the same. What was needed was fruit in one place, at one time, side by side.

That is what the national fruit congresses of the later nineteenth century provided. Exhibitions to which growers sent named fruit from across a country allowed committees to compare specimens directly, reconcile names, and publish agreed lists. The large English apple congress held at the beginning of the 1880s is the best-known instance and had a lasting effect on English variety nomenclature.

The permanent version of the same idea is the reference collection: a planted collection holding trees of every named variety, from which fruit and wood can be taken for comparison in any season. National fruit collections were established in several countries through the twentieth century, and they are the physical basis of everything said about variety identity since. A description in a book is an assertion; a tree in a collection is a specimen.

The descriptive method is more disciplined than it appears. An identification looks at the whole fruit and the tree, in a defined sequence, using terms with fixed meanings: the shape in profile and in cross-section, the ratio of height to width, whether the eye is open or closed and how the sepals sit, the depth and shape of the basin and the cavity, the length and thickness of the stalk, the pattern of russet and lenticels, the colour and texture of the flesh, the shape of the core lines and the pips, and the season of ripening and keeping.

Its limits are equally well understood by the people who use it. Fruit varies enormously with season, rootstock, crop load, position on the tree and the weather of a particular summer, so a single fruit is rarely diagnostic. A good identification uses several fruit, prefers typical specimens from a well-cropped tree, and treats the tree’s habit and leaf as evidence alongside the fruit.

For cider fruit the difficulty is compounded, because cider apples were the least documented category. Many were never described in any pomona, local names were used for populations rather than for clones, and the reference plates that exist for dessert fruit have no counterpart for most bittersweets.

From the 1990s, marker-based fingerprinting — first microsatellite markers, later dense arrays of single-nucleotide markers — made it possible to compare a sample directly against a reference set. The method does not tell you what a tree is; it tells you which reference accession it matches, and how closely.

Applied to national collections, it produced two uncomfortable findings that are now well established. A significant proportion of accessions held under distinct names turned out to be genetically identical to one another, meaning that historic synonymy was worse than anyone had assumed. And a proportion of accessions did not match the variety they were labelled as, meaning that some material had been mislabelled at some point in a chain of propagation stretching back a century or more.

The consequence is a change in what an identification claims. A molecular identification is a statement about the reference database: this tree matches that accession. If the accession is itself misnamed, or if the true variety is not in the database at all, the method returns a confident answer to a question that was not asked. Both the descriptive and the molecular approaches remain necessary, and the strongest identifications use them together with the documentary history of the tree.