Land birds
To the Island in these first few weeks of spring there had come, from the surrounding seas and from the Atlantic ocean, from lands and continents to the south, from the neighbouring mainland of Britain, some hundreds of thousands of birds. Of this multitude many had passed on towards their breeding grounds elsewhere, in Britain and in Ireland, in Iceland, Greenland and Scandinavia; but we estimated that nearly two hundred thousand birds - more than ninety-one thousand pairs - remained on Skomer's seven hundred and twenty acres to breed. By far the most of these were the seabirds, and yet, if one stood on the highest rock on the island, close to the house, one might scan the plateau of the island in vain, looking for this great throng. It was only at night that the shearwaters could be seen here in their thousands, sitting outside the burrows on the turf in all the drier parts of the island. But by day only the lesser black- backed gulls could be seen in any numbers, for the herring gulls were mostly near the cliffs, or on the cliffs, and the great black-backs were dotted about on high points of the rocks here and there. But the cliffs were noisy all day with the crowded jostling birds: the puffins, most numerous of all, and outnumbering all the other birds on the island, were along the cliff-tops, where they had their burrows, and below them were the dainty screaming kittiwakes on their isolated ledges, and the guillemots and razorbills packed tightly along the more open, long rock-shelves.
The surface of the island plateau, apart from the colonies of lesser black-backs, was scarcely more thickly populated than similar clifftop moorland on remote parts of the mainland coast, and the birds breeding there were much the same. Of the thirty-six species breeding on the island, eleven were seabirds properly so-called: the four gulls, three auks, Manx shearwater and storm-petrel, cormorant and shag. The other twenty-five species totalled rather less than two hundred and fifty pairs between them, and thus provided an almost insignificant proportion of the whole population, for all their variety.

Rock-pipits
These 'land-birds' included two species, the rock-pipit and the oyster-catcher which have, perhaps, as much claim to be called seabirds as any, except that neither of them possesses webbed feet, and neither of them normally swims, though the oyster-catcher can swim well enough on occasion. Both these birds forage much at the edge of the sea, the rock-pipit among the spray-splashed rocks and crevices, and among the wrack left by the ebbing tide, and the oyster- catcher also among the weed, where it knocks limpets and other shell-fish off the rocks with a sharp blow of its powerful bill. But the oystercatchers live all over the island - there were thirty-six pairs of them and probably on Skomer, as on the much smaller island of Skokholm, those in the centre of the island seldom or never went to the sea's edge for their food, but found all they needed in the marshy ground in the centre of the plateau. The rock- pipits on the other hand confined themselves to the cliffs during the breeding-season, and there were about thirty pairs of them more or less evenly distributed round the perimeter of the island. Only after the young were fledged would we see them on top of the island, trespassing on to the ground occupied throughout the summer, and indeed throughout the year, by the meadow- pipits, of which about sixty pairs nested among bracken and heather. These three species were the commonest of the land-birds, and between them were spread over the whole island. During the breeding season the two pipits seldom trespassed on each other's regions of the island: the rock-pipits remained on the cliffs, while the meadow-pipits shared the turf of the island with the wheatears. Here too there was some division between the habitats, for the wheatears frequented the shorter rabbit- grazed turf, and the stone walls, in which, as well as in the rabbit burrows, they made their nests, but the meadow-pipits tended to forage more often among the taller grasses and in the belts of heather and bracken. No hard- and-fast line could, of course, be drawn between these two habitats, and wheatears were often to be seen on tumps of dry heather, and meadow-pipits on the walls; but for all that there was some tendency to keep apart in the way described.

Larks and Wagtails
Closely related to the pipits are the larks and wagtails, but of these there were only three pairs of skylarks, and a single pair of pied wagtails. The wagtails raised two broods in one of the ruined outbuildings, and were often to be seen fly-  catching on the roof of the house. There was also a pair of pied wagtails on Skokholm and one day we caught in the trap by the house a chick which had been bred on Skokholm a few weeks earlier. This was the only record of bird traffic between the two islands which we obtained during this season, though since 1946 there have been a few recoveries of seabirds marked on Skomer and found on Skokholm. The skylarks sang over the island, as they sing every year over Skokholm, and they bred near the marshy centre of the island.

Dunnocks, Blackbirds and Whitethroats
There, where the stream flowed out of the pond, was a tangle of brambles, and some other bushes further down this stream: in these dunnocks and blackbirds and whitethroats-the only warbler to breed on the island in 1946 - were to be found. There were about five pairs of whitethroats, two pairs of blackbirds, and one solitary pair of dunnocks. Before the war dunnocks used to breed regularly on Skokholm, sometimes as many as seven pairs, and their cheerful little song was almost always to be heard from the bracken near the big trap. Now, since 1940, they have not bred and in 1946 and 1947 they were not even seen on the island. These small isolated populations are liable to be wiped out quite suddenly, after a bad season, and it may then be some years before a new colony is established. In 1948 a dunnock was seen again on Skokholm, and this may be the forerunner of a new colony. The blackbird also used to breed on Skokholm, but has not done so since 1935, though it is a regular visitor outside the breeding season, and there seems no reason for its ceasing to breed.
Wrens
Another bird which visits Skokholm regularly in winter, but which has never yet been known to breed, is the wren: but on Skomer we had some thirteen pairs, mostly about the cliffs, or the ivy- and honeysuckle-covered outcrops above the cliffs. Wrens are found breeding on most islands, even on the remotest groups, where, as on St. Kilda, local races may develop in centuries of isolation. Skomer is of course too near the mainland for any difference to be noticeable among the island wrens, which no doubt receive recruits from the mainland each autumn. But why they never stay to breed on Skokholm remains a mystery.
Jackdaws, Crows and Choughs
Not less strange is the absence of jackdaws from Skokholm, for they are always to be seen about the cliffs of Skomer where about twenty pairs bred in old rabbit and puffin burrows made in the loose earth at the cliff-tops. About the cliffs three other members of the crow family were to be found throughout the summer: three pairs of ravens, eighteen pairs of carrion-crows, and two or possibly three pairs of choughs. The ravens nest, very early in the year, in the same inaccessible sites year after year, and they are to be seen on almost all the islands, even on remote Grassholm. The carrion-crows also nest on the cliffs, in default of trees, and make a great harvest of the eggs of razorbill, guillemot and gulls - also of oyster-catcher or any other bird that leaves its eggs briefly unguarded. (But the oyster-catcher is very bold and effective in the defence of its nest, and suffers far less than the bewildered and hopelessly inefficient guillemots.) The choughs were most elusive, but someone usually saw them on most days - often six of them together. They breed for preference within some high cave or recess, where the nest is protected from the weather and from sight, and they come up on to the fine turf at the top of the cliff to probe for insects with their slender curved red bills. It has been suggested that the spread of jackdaws has led to a decline in the number of choughs breeding about our coasts, and certainly they have declined in the past fifty or sixty years, though of late protection has enabled this most handsome and svelte of birds to hold its own along the coasts of Pembrokeshire. The case against the jackdaw is not proven: the case against the egg-collector on the other hand is all too clear. Perhaps one day the buoyant flight of the chough swooping and playing in the currents of air above the cliffs, and the cheerful squealing call will again be familiar to many, and at least the visitors to the Pembrokeshire National Park may have some chance of enjoying them. B. H. Ryves, studying the chough in Cornwall, has found that there are many non-breeding adults present in the nesting season. The behaviour of the six adults on Skomer suggested to us that they were nonbreeders. At least no mating or nesting activities were certainly recorded.
Buzzards and Peregrines

The buzzards also bred about the cliffs, and it was a magnificent sight to see sometimes as many as fifteen of these great birds soaring together towards the summer sky. Seven pairs bred on the island, but others came over at times from the mainland, as the presence of the remains of a mole in the cast of a buzzard on Middleholm proved. For the most part they lived on the rabbits of the island, though they no doubt took some small birds, mice and voles, and perhaps frogs, beetles and grasshoppers. The whole island was theirs, though they made their nests in the safety of the cliffs.
Above the screaming, growling, groaning proletariat of sea- fowl on the Wick a pair of peregrines had their eyrie, and when not stooping after puffin or razorbill, jackdaw or meadow-pipit they would stand motionless for hours, surveying the multitudinous life of the cliff with cold indifference. We seldom saw them over the centre of the island: the cliffs provided all their needs, for they did not need to prey on the small flightless creatures such as voles, shrews and lizards which the slow buzzards and kestrels must content themselves with. Sometimes no doubt the peregrines would fly across to Skokholm, and perhaps they brought back from there a puffin, whose leg, with a ring put on on Skokholm, we found one day at the foot of a cliff. The peregrines were watched from a hide. On June 8th one parent brought a ringed puffin to the nest, and tore off the leg with the ring. This tough morsel was offered to the chicks which refused it, whereupon the adult swallowed leg, ring and claw. The eyrie was littered with the remains of rabbits, puffins and shearwaters; this pair seemed to feed more than usual on prey taken on the ground - it is almost certain that the shearwaters would be taken as they rested on the ground, probably early in the morning. R. M. L. in fact surprised one of the adults feeding on a freshly killed shearwater dragged from under a slab of rock where, judging by the litter of feathers there, it had been first killed. When James Fisher climbed down to the eyrie on June 2nd to ring the two eyasses he found the remains of eleven shearwaters and two puffins. Joan Keighley saw a peregrine on August 10th feeding on a freshly killed kestrel, which she brought into the house for identification; it appeared to be an immature bird, quite probably not expert enough on the wing to escape a determined peregrine.
Short-eared owls
The short-eared owls nested in the bracken near the north stream. In April the striking 'song-flight' was watched. It usually took place in the morning, but was recorded several times in the evening also. One bird would rise until it was estimated to be as much as seven hundred feet up (other estimates on various occasions varied between two and five hundred feet). It would circle, then suddenly clap its wings several times beneath its body, at once falling vertically; the clap of the wings was likened to a ratchet by one watcher. Usually the bird rose again, circled, clapped its wings again, then uttered the hooting 'boo-boo- boo' as many as twelve times in succession. The mate of the displaying bird meanwhile remained below actively quartering the ground from a low level. Mating frequently followed the return of the aerial exhibitionist to earth, even when in May the pair were known to have well-grown young. We did not find the eggs, but but the young birds - three of them - were estimated to be three weeks old on May 19th, when they were ringed. This is very early, and places egg- laying at the end of March, a month before the usual time in Britain. The three young owls were killed by gulls as soon as they left the nest and before they were able to fly; they had probably blundered into the nesting territory of the lesser black-backed gulls in the same area of bracken. Early in June the adults, deprived of their brood, resumed display and song-flight for about a week, but were much interfered with by the gulls which attacked each soaring owl as it circled over their nesting colonies.

We found no second nest of this owl, although in good vole years this species will rear a second brood. The short-eared owls on Skomer lived mainly on the Skomer voles and the wood- or longtailed field-mice; it would be interesting to discover if the numbers of this owl fluctuate with fluctuations in the numbers of the Skomer voles, as they do on the mainland with the numbers of the mainland vole, and, especially, on the Continent. It seems likely that at one time this happened on Skomer also, for there are records of as many as seven pairs of short-eared owls on Skomer about fifty years ago, although in 1946 there was only one pair. It is certain, however, that with the great increase of the breeding gulls, the numbers of the island vole are kept down to a fairly low level, since the gulls occupy so much of the same ground as the voles, and also the voles are partly diurnal, affording the diurnal gulls many opportunities of preying upon them. There are no recent records of more than one pair of short-eared owls breeding on Skomer. When the gulls have left the island in late autumn and winter, barn owls take up residence on Skomer and may remain late into the spring (April 1946; May 1947); they have sometimes bred there - in 1897 Drane found a pair breeding in a small cave on the Mew Stone, a most improbable place to find this owl, one would have supposed, among the cormorants and gulls.
Drane, quoted above, mentions the storm-petrel killed at night by, as he seems to suggest, an owl of some kind. The short-eared owls of Skomer sometimes visited Skokholm: to an owl which soared so well this was only a short journey, two miles across the sea to the lower island plateau so plainly to be seen from Skomer. One day a visitor to Skokholm picked up a pellet from the ground beside a roosting short-eared owl which he disturbed. This pellet contained the remains of a stormpetrel, a prey never before authentically recorded for the short-   eared owl. We had had much trouble before the war with little owls on Skokholm, and in the end we had had to destroy or deport the owls in order to preserve the petrels. On Skomer there seem to be fewer storm-petrels than in Drane's time, probably now not more than thirty pairs, and these were persecuted by the little owls, as we shall describe in the chapter on petrels.
Little Owls
The little owl is a new arrival in Pembrokeshire, first breeding in 1920, and was numerous on the islands seven years later. One pair (possibly two) bred in 1946. Though only one nest was found, and the young ones in this did not survive, a juvenile bird just able to fly was seen on July 20th - but this may possibly have flown over Jack Sound from the mainland. The curious feature of the nesting of this pair of little owls was the finding, on June 17th, of one owlet dead and the other dying with a damaged wing; the nest was under a huge slab of rock on the south cliffs, and in a recess at the back, beyond the owlets' home, a puffin was brooding. A puffin, we believe, is more than a match for a grown little owl, but we never discovered whether or not it had attacked the young owls.

Curlews and Lapwings
On the marsh, not far from the house, there were curlew and lapwing breeding: the latter we had expected, for they breed regularly on Skokholm, but the finding of five pairs of curlew on Skomer was a most delightful surprise, for surely there is no lovelier sound made by birds than the wild, thrilling song of the curlew as he flies above his mate, quickly shivering his wings in time to the bubbling cry. That song had been the favourite of one of the finest men who had ever been to Skokholm, but, after winning the D.S.O. and the M.C. (twice over) Geoffrey had been killed in the Mediterranean at the age of twenty-six. And so he could not share in this new pioneering on the islands, nor lie and listen to the curlews with us any more.
One pair of curlew, with a nest near the south pond, had seven eggs to incubate. This nest was found on May 5th, and closely watched from a distance on several occasions, but only two birds were ever seen near the nest, and of these one was observed to sit for varying periods. On May 27th one egg was chipped and the beak showing through. On the 29th two chicks had hatched out and were successfully piloted away to food and cover by the parents, which then abandoned the five remaining eggs, which were found to be addled.

Oyster-catchers
But of all the land-birds on Skomer the most conspicuous in voice and in plumage was the oyster-catcher or sea-pie (as it used to be called, with more aptness) - the bird sacred in Celtic legend to St. Bride, over whose bay we looked out every day. In 1939 we had begun a study of this striking bird on Skokholm (Buxton 1939) but the war had interrupted this, and now on Skomer we proposed to continue this study so far as we might. Skomer is less well suited to the oystercatcher than Skokholm, which has a larger number of pairs breeding. The reason for this may be the greater density of bramble and tangled vegetation on Skomer, for the oyster- catcher likes to run up to its nest and also likes some sort of hillock from which to keep watch near at hand. This season we gathered what data we could about the incubation period of the species, but these eventually proved less precise than we required and this part of the work was continued with great patience and skill by Joan Keighley on Skokholm in 1947 (Keighley and Buxton, 1948). In 1946 the first clutch found on Skomer was already complete on April 28th and, since the incubation period is normally slightly under twenty-six days, the first chicks must have hatched not later than May 24th - a little earlier than we have found to be usual on Skokholm. The last chick hatched on July 19th, much later than any of which we have record from Skokholm, but probably the clutch here was a replacement of eggs taken by gulls or crows. (In 1946 on both islands and in 1947 on Skokholm over 30 per cent. of the eggs were lost in this way.)
Late in the summer, when the young birds, distinguishable by their browner plumage and dull bills, are on the wing, the oyster-catchers tend to congregate together, when they are not feeding, on some rock or boulder - on Skomer a favourite place was on the west side of South Haven. Here they stand and snooze peacefully side by side, and the excited piping is seldom heard after July is out. With the moult they get a white collar about the neck, and we tried to discover, by counting the proportion of birds with this collar, how soon the winter plumage was assumed. To our surprise the proportion seemed to remain about the same throughout the last two months of our stay on the island, and this probably suggests a constant-shift in the population seemed to remain about the same throughout the last two months of our stay on the island. And indeed, since then, we have had a number of oyster-catchers from Skokholm recovered in France, showing that there is a rather more marked migration in autumn than had been supposed. But our studies of this bird continue and are likely to continue for some years to come.
The chicks may take as much as three days to emerge from the egg after the shell has first been chipped. As soon as they are dry they run from the nest, and thereafter the territory of the parents is mobile, centred upon the chicks, which, in the weeks that elapse before they are able to fly, may run hundreds of yards from the nest.
The elaborate strange 'piping' ceremony of the oyster- catcher which is used in many different situations has been too often and too well described for us to describe it again. It is perhaps enough to say that the familiar vibrant trilling which accompanies this performance was a frequent but always delightful sound, by day and by night, during the first part of the summer. It is a communal performance and the voices of the piping birds blend perfectly as they run along, he and open bill pointing to the ground; and (as Selous said) they have an absent- minded appearance, as if they were wholly absorbed in what they are doing, as, no doubt, they are.
Mallards and Moorhens
On the north pond, or in the flags and sedge beside it there were four pairs of mallard and two most elusive pairs of moorhens. It was strange that these birds should be so wild and skulking here, like waterrails, when they are so tame and confiding on the concreted pools of most of our great cities, but it was so, and this problem of tameness is one which has only recently received the attention it deserves.

Linnets and Starlings
Two other very familiar species bred on the island, at either end of it - about four pairs of linnets on the Neck, and at the most seaward point of the island an adventurous pair of starlings. This may well have been the first occasion on which starlings bred, though linnets were known to be more or less regular on Skomer, for it was only in 1940 that starlings first bred, in the eaves of a shed belonging to the lighthouse, on Skokholm. The starling is one of the most successful birds of recent times and it has spread considerably within the last century - in Pembrokeshire more recently still, for when Mathew wrote his account of the birds of Pembrokeshire in 1894 it was still quite scarce as a breeding bird. It is so vulgar and plebeian a creature with its cockney whistlings and mutter that one is not nowadays surprised to find it anywhere, and after all it is a cheerful bird and, in a good light, a handsome one. The difficulties of colonising an island full of predatory birds were too much for this solitary pair, which nested within sight of the peregrines' eyrie. The pair were busy at the rock cranny near the Basin, where a tell-tale splash of white mutings marked a nest inaccessible to human climbers, until May 19th, after which only one adult was recorded. The widow, or widower, disappeared after June 6th, and no young birds were fledged. It may be significant that a migrating merlin was seen near the Basin on May 20th, but peregrine or owl might have been responsible, too. We saw no more starlings until the first pack of twenty-nine brown- plumaged juveniles arrived from the mainland on June 22nd, the forerunners of the autumn migration.

Pheasants
One other bird remains, the pheasant, though whether it bred on the island or not we never discovered. There were a cock and three hens, but no chicks or young birds were seen. Some of us rather expected that the skipper's rabbit- snares, set in paths known to be frequented by the cock pheasant and his harem, would catch one. But instead the snares unexpectedly caught once an oyster- catcher and once (August 9th) an adult raven.
Presumably the pheasant was introduced on to the island at some date, though we have seen pheasants flying across Jack Sound to the mainland, and it might perhaps have colonised the island without human assistance. Skomer is so utterly unlike the usual haunt of the pheasant that it always seemed odd to see them on the island, where there were no neatly ploughed land, no stubbles, no carefully preserved coverts. But the pheasant is a hardy bird, and it seems well able to thrive even in such conditions as Skomer offers, and if it chose to desert the island it is able to do so.