A moose - loose . . .

The latest in a long line of hail-squalls rears up above the islands, fills the sky, and hurls itself upon the house with a ferocity that even doubleglazing cannot muffle. You'd think such an icy hammering would strike birds out of the air and litter the hillside with fallen rooks and gulls. Perhaps, like the sheep, they manage to turn their backs and tuck their heads in, the hailstones bouncing off their wings.

The turn in the weather has, however, produced other casualties, announced by sudden sharp clicks in the recesses of the hot press and the cupboard under the sink. Mus musculus, finding life suddenly too chilly in the hedgerow for all its abundance of seeds and berries, has met an abrupt end while seeking a nibble of dog-biscuit.

Ethna sets the mouse-traps, having a delicate touch and stronger nerves, and I dispose of the corpses (perhaps a dozen in a winter). Besides, I have to check that the Thallabawn house mouse is of regulation mouse-colour, with four paws and proportionate tail, before flipping it over the hedge for a kestrel to find.

But why kill the poor little things at all: why not capture them in a live trap and let them go at the end of the acre?

I have a book which offers many kinds of traps for mice, including one from America called the Ketch-All which bats the mice into a hopper with a kind of revolving paddle (there is an accessory drowning-bath as an optional extra). Biologists, kinder folk, prefer to sample small mammals with the Longworth trap, which ushers its prisoner into a nest-box lined with dry grass or sheep's wool for inspection, measurement and subsequent release.

The trouble is that Mus musculus domesticus does like its warmth in winter. In its ancestral setting, nibbling grass seeds on the steppes of Turkestan, it no doubt coped with bitter weather, but it was quick to take advantage of the earliest growing of grain and to develop commensal habits which carried it along human trade routes. It could have arrived in Ireland stowed away among the sacks in a Bronze Age currach.

It remains a tough little creature when it needs to be, especially on islands. On South Georgia, for example, it lives a truly feral lifestyle in near-Antarctic conditions (no wonder some mice can spend their whole lives in butchers' cold stores). On islands around Ireland, too, it commonly leads a largely outdoor life, at a distance from houses, under cover of bracken, brambles and dry stone walls. But when the people leave and houses fall empty, the survival of Mus musculus seems to depend on how far it has to compete with the island population of the field mouse, Apodemus, the most common of all the Irish mammals. On the Great Blasket and the Inishkea Islands, for example, where the long-tailed Apodemus is plentiful, the house mouse is extinct. On St Kilda, off Scotland, similarly, it died out within two years of human evacuation.

On the Irish mainland, however, Mus is rarely found more than 100 metres away from buildings. Studying trapping records from a variety of countryside habitats - grassland, hedgerows, woods and dry stone walls - mammalogists James Fairley and Chris Smal found only a dozen or so house mice caught along with more than 1,100 field mice.

One Irish island, though scarcely "offshore", has a famous population of house mice. Exactly 100 years ago, a Louth-born naturalist called Henry Lyster Jameson discovered that many of the house mice living on the sand-dunes and salt-marsh of North Bull Island in Dublin Bay had sandy fur on their backs. This was taken to be a special protective camouflage against predation by the kestrels and short-eared owls that haunt the island in winter.

If this was, in fact, a product of natural selection, in which the birds killed more of the darker mice and left the sandy ones to breed, it was a remarkably rapid effect on an island which, at that time, was scarcely a century old. But later trapping has confirmed the sandy mice, and James Fairley, in his Irish Beast Book, concedes they are "an interesting phenomenon, worthy of further study".

Given the ingrained commensality, or table-sharing, of Ireland's mainland mice, it would make little sense, I fear, to livetrap our autumn invaders and dump them down the boreen. The countryperson's instinctive habits of storing the flour in a tin box and folding down the top of the Cornflakes bag between breakfasts are not enough to discourage them from the cosy labyrinth within our walls.

Even without the 50 droppings a day per mouse and an average 6.2 litter at low density breeding, what I really cannot take about Mus musculus is the incredible racket they kick up in their parallel universe behind the plasterboard. Mice, we are told, have this incredibly subtle system of scent-marking (with urine) that allows them to negotiate every object in the dark. They can tippy-toe across narrow bridges, sense precipices, recognise each other to the nearest molecule. And yet here is this clattering, banging and crashing right behind my bed-head.

"There's a moose loose, loose aroond the hoose . . ." Thus began one of our marching songs on the road to Aldermaston (I was with a Scottish contingent). But pacifist and nature-lover or not, what is my ignoble cry as the trap snaps to in the kitchen?

"Gotcha!"