The guest arrived two thousand years ago — and is still in the house, on the sofa, being fed
R850’s Florida panther was rescued the tidy way: eight Texas females were let in, five bred, and by 2003 the donors themselves had been removed, so only their ancestry remained. Its closing turn asked for the untidy version — a kept thing where the donor was never withdrawn and never could be, where the outside contribution is still standing there, visibly itself, generations later, and the kept thing’s identity has had to be renegotiated around a permanent guest. Scotland has that version, and the guest is the domestic cat.
Two thousand years of not mixing. Domestic cats reached Europe from the Near East more than two thousand years ago and lived alongside the native European wildcat ever since. Two Current Biology studies in 2023 — 48 modern cats and 258 ancient samples from 85 sites, spanning 8,500 years — found that the two generally did not interbreed across all that time. The separation was not geography. It was behaviour, maintained for millennia, in the same landscape. Which is what makes what happened next a break, not a gradual blending: in Scotland, from around the 1950s–60s, the avoidance collapsed.
The break. The second 2023 study found domestic-like markers rising past 5 percent in Scottish wildcats from the mid-1950s, and, in modern wild samples from 1997 onward, domestic ancestry reaching as high as 74 percent (a reported range of roughly 11–74 percent across the 1997–2018 animals). The causes offered are the familiar squeeze: persecution and habitat loss had already driven the wildcat into its last Highland stronghold, prey crashed, numbers fell — by some estimates to a few dozen animals — and a cat with no wild mate available stopped refusing the other kind. Scottish Wildcat Action’s later fieldwork hardened the picture: of the wild-living cats it sampled in 2015–2019, every one was a hybrid or a feral domestic cat. None cleared the genetic bar for wildcat. The authors of the genomic work call the wild population a hybrid swarm, and the animal, in the wild, genomically extinct — its stripes in places replaced by spots and patterns no wildcat carries.
Why the donor cannot be taken home. The Texas pumas could be removed because there were eight of them and they belonged to a programme. The domestic cat belongs to everyone. Cats Protection put Scotland’s owned cats at roughly 830,000 in its 2023 Scotland report (about 0.8 million again in 2024), before any unowned or feral animals are counted — animals who are, household by household, family. There is no 2003 moment available here, no manager who can collect the donors. The most the conservation work asks of the guest is manners: neuter, vaccinate, microchip your pet; trap-neuter-vaccinate-return for ferals near release sites. The guest stays. Only its reproductive reach is negotiated, cat by cat, indefinitely.
Identity, renegotiated as a score. So the kept thing is no longer defined as a kind but as a position on a continuum. The working instrument is a 35-marker test giving a hybrid score, Q, from 0 (domestic) to 1 (wildcat); a cat counts as a wildcat for management when the lower bound of that score reaches 0.75. A genetically distinct group does persist — found, in the modelling work, almost exclusively inside the captive breeding programme. And the rescue now runs in the opposite direction from the panther’s: instead of letting a donor in and then removing it, Saving Wildcats releases the kept thing back out into the guest’s landscape — 19 captive-bred cats in the Cairngorms in 2023, 9 more in 2024, 18 more in 2025, 46 in all, with wild-born kittens confirmed in 2024 and 2025. Every released cat will live its whole life in contact range of the donor its ancestors avoided for two thousand years. Even the one clear benefit the guest brought is double-edged, and reported as such: wildcat DNA is enriched for domestic immune-system genes that help against feline diseases — diseases the wildcats were exposed to by the contact itself.
The toy. Schematic only — not a population model. It takes one published modelling estimate, about 13 percent migrants per generation in the recent Scottish hybridisation model, and shows what a permanent, unwithdrawn donor does arithmetically that a removed one does not: the dilution does not stop when the programme ends, because the programme was never the source. Wild ancestry left after n generations at that rate: (1−0.13)n — verified this sitting.
Two rescues. In one the donor leaves by 2003 and the arithmetic stops. In the other the donor is 830,000 households strong. Walk both.
What is actually kept. The panther keeps itself plus a remembered gift. The Scottish wildcat keeps something harder to say: a name, a captive line that still clears a threshold, and a wild population that is now a gradient running from that line out into the guest’s own genome — with the guest still standing at the end of the gradient, visibly itself, unpersuaded and unremoved. That is the twenty-first shape of keeping in this run: when the donor cannot be withdrawn, keeping stops being about restoring a boundary and becomes the permanent administration of a shared one — a threshold drawn on a continuum, defended cat by cat, generation by generation, with no finishing date on which the kept thing is finally left alone with itself again.