Four in the morning, and the two hens that came after

Part 1

The night itself

It was the fifth of October, somewhere around four in the morning, and what woke me was not noise from the garden but a neighbour's security light throwing a rectangle on the bedroom ceiling. By the time I got down the stairs and out of the back door with a torch there was nothing to see but feathers on the path and the run door standing at an angle. I lost two birds, both of the older ones, and a third was in the corner of the shed with a wound across her back that healed over about three weeks.

The point of failure was the door, and specifically the bottom hinge and the gap under it. I had built the run with twelve-millimetre welded mesh on every panel, which is the right stuff, and then hung a door with a five-centimetre clearance at ground level so that it would swing over uneven paving. A fox does not need to chew through mesh when there is a gap it can get its head and shoulders into and lift. Everything I had worried about was fine. The thing I had made a convenience decision about at seven in the evening two years earlier was the thing that failed.

Part 2

What changed in the fortnight after

The door got rehung on a level threshold with a timber sill that it closes against, and two bolts instead of a turn-button, top and bottom. That was the essential fix and it took an afternoon. Then I went round the whole run on my knees with a torch looking for the same class of mistake somewhere else, and found three: a panel joint where two frames met with a finger-width gap, a section where the ground had sunk about eight centimetres under the mesh, and the pop hole, which had been left open at night for two years because the run was fox-proof.

The skirt was the bigger job. Forty centimetres of mesh laid flat on the ground all the way round the outside of the run, pinned down with long staples and then covered with soil and turf, so that anything digging at the base hits wire before it hits earth. Foxes dig at the line of the fence rather than starting a metre out, so a flat skirt works better than burying wire deep. That took two weekends and a bag of pegs, and I should have done it when I built the run.

Part 3

Bringing the number back to six

I waited until the end of October, partly because the injured bird needed the space quiet and partly because I did not want to introduce new birds into a flock that was still unsettled. Two point-of-lay pullets came from a smallholding out past Loxley in the first week of November, both about eighteen weeks, and they went into the shed in a wire-fronted crate rather than straight in with the others.

Everything I have read about integration is some version of the same advice, which is to let them see each other before they can touch each other. That worked out at nine days in my case: the crate in the corner of the run through the day where everyone could look, and the two of them shut in a partitioned end of the shed at night. Nobody was happy. The oldest remaining hen stood at the crate mesh for most of the first two days doing a slow, unpleasant sort of patrolling.

Part 4

Three weeks of genuinely poor behaviour

The first evening they were loose together was the twelfth of November and it was worse than I expected. Two of the established birds went straight at the newcomers, not once but repeatedly, and one of the pullets spent that night on the run frame rather than go through the pop hole. I sat on an upturned bucket in the dark and put her in by hand at about nine, which I then did for six nights running.

What got us through it was space and obstacles. I put two pallets on edge in the run and a length of plank across two blocks so there were places to get out of sight, and moved the second feeder and a water bucket to the far end so that the newcomers were not obliged to walk past the established birds to eat. Line of sight is most of the problem. A run that is a bare rectangle gives a bullied bird nowhere to be, and a run with three obstacles in it gives her somewhere to wait until the moment passes.

It took about three weeks for the shoving to stop and closer to six before the two pullets stopped roosting at their own end of the perch. The first egg from either of them came on the ninth of January, which is exactly the wrong time of year and was therefore a small event in this house.

Part 5

Where the flock stands now

Six again, four of them from the original line and two from November 2025, and the one with the scar across her back laid normally through the following spring. The run has had no further visits, although I saw a fox on the back lane twice in January and once, in daylight, crossing the end of the garden in February. It is not that there is no fox. It is that the door now closes against a sill and the pop hole is shut before I go to bed.

The part I keep coming back to is that the failure was two years old and I walked past it every day. I had inspected the mesh, which I was proud of, and never looked at the gap under a door I opened four times a day. When I check the run now I make myself start at ground level with a torch and work up, because the eye goes to the thing you built deliberately and slides straight over the thing you settled for.