Along a disused railway in the west of the country there are apple trees at irregular intervals, and every one of them was planted by somebody throwing a core out of a window. The pattern is legible if you know what to look for. Trees cluster near where stations once stood, thin out along the straight sections, and reappear before curves, where trains slowed. There is a dense group at a point that corresponds to nothing on the modern map; the archive shows a signal box there until 1961, and signal boxes meant waiting. A botanist who surveyed the line described the result as an inadvertent record of human impatience. People eat when they are bored, and they are bored when they are stationary. The orchard is not a record of what passengers ate but of where they had nothing else to do. None of the trees is of commercial value. Apples do not breed true from seed; a pip from a cultivated variety produces something unpredictable and usually inedible. The line's trees are, botanically, a collection of unique individuals, each one a genetic experiment that no orchardist would have permitted and that nobody has ever evaluated. This is where the survey becomes more than a curiosity. Commercial apple growing rests on a very narrow genetic base, propagated by grafting, and is correspondingly vulnerable. The material that might one day be needed against some new disease is unlikely to be found in orchards, which contain the same few varieties everywhere. It is more likely to be found in exactly this sort of accidental population: unmanaged, unselected, and consequently diverse. Several such lines have now been catalogued. The work is slow and unglamorous, funded intermittently, and conducted largely by people who do it in addition to something else. Its practitioners are notably reluctant to overstate the case. Most of the trees are, by any ordinary standard, bad apples. The point is not that any particular one is valuable but that the population contains variation which no deliberate programme produced and none could easily replicate. There is an obvious irony in a conservation resource created by littering. The botanist noted it and then, characteristically, complicated it. The trees survive because the line was abandoned. Had the railway remained in use, the embankments would have been cleared on a regular cycle and nothing would have reached maturity. The orchard exists at the intersection of two kinds of neglect: people who could not be bothered to keep their rubbish, and an institution that could not be bothered to maintain the land. The line is now a footpath. A local group has begun labelling the trees, which the botanist opposes on the grounds that labels invite management, and management is precisely what produced the problem the trees might one day solve. The argument has not been resolved. In the meantime the trees continue to do what they have done since the 1890s, which is to fruit heavily, unevenly, and to nobody's specification.