may be presumed that they set out with them on their migration. On a visit to the sand-pit, on the 10th of September, 1840, not a bird was to be seen, but less than a mile distant I saw a few associated with house-martins and swallows, the latter of which especially were abundant.*

On the 2nd of September, 1842, I observed several of these birds fly into their burrows in a layer of hard sand about two feet in thickness, surmounting the high gravelly bank of a wild mountain stream, before the shooting-lodge at Aberarder, in Inverness-shire.

During a tour to the south of Europe and the Levant, made by the author in the spring and summer of 1841, this bird was only observed:-On the 9th of April, about the Rhone, between Lyons and Avignon, where very few appeared; on the 17th, when a number were seen about Valetta in Malta, in company with many swallows, house-martins, and swifts; on the 30th, when several were observed between Navarino and Modon. Aristotle having mentioned the sand-martin as frequenting the valleys of Greece, I was much gratified by thus meeting with it in the first valley, or rather defile, of the Morea that I visited.

Macgillivray's British Birds (vol. iii.) contains a very good description of this species by the author, enriched by valuable contributions from Mr. Weir and Mr. Duncan. Audubon gives a very full and interesting account of it as an American species. He remarks:-"The sand-swallow is a rather hardy bird; for I observed, that the transient cold weather that at times occurs in the Floridas at night, seldom forces them to remove farther south. On one occasion, however, when the ice was about the thickness of a dollar, many were found dead along the shores, as well as floating on the water." Orn. Biog., vol. iv. p. 587.

* Mr. Poole, with reference to the departure of the sand-martin from the county of Wexford, mentions its having been seen on the 31st of October, but he has not observed its arrival there so early as we have in the north. He remarks, that "they frequent our fields towards the latter end of autumn, though we never see them at any other time," and describes nests which have come under his notice, as being "from twelve to twenty-four inches from the entrance of the cavity;-composed of fine sea-grass, with a few picces of dry grass-stalks; the burrow, as nearly as may be, horizontal, and of almost uniform diameter so far as the nest, where it is enlarged to a globular form. Eggs taken in the middle of May, and first week of June; five in number."

THE PURPLE MARTIN.

Hirundo purpurea, Linn.

Is said to have been once obtained in Ireland.

Mr. YARRELL states in his British Birds, that the species is "included, in consequence of a letter received from Mr. Frederick M'Coy of Dublin, informing me that a female example of the species had been shot near Kingstown, in the county of Dublin, which had been sent for dissection to Dr. Scouler a few hours afterwards, and when preserved was placed in the Museum of the Royal Dublin Society." vol. ii. p. 257, 2nd edit. The date of its occurrence is not mentioned: the communication respecting the bird was published about March, 1840.

We are further informed by the same author that "during the first week of September, 1842, two other examples of this same species were shot by Mr. John Calvert of Paddington, at the Kingsbury reservoir." One of these is a young bird of the year, with the outside tail-feathers not fully grown, and the other an old male, which circumstance, taken in connection with the fact that two or three days intervened between those on which they were killed, inclines Mr. Yarrell to believe that a brood of them may have been reared there. These additional instances of an American species occurring in the British Islands;-see observations on Carolina cuckoo, belted kingfisher, American bittern, &c., —and in no more easterly part of Europe, strengthen the opinion that such birds crossed the Atlantic Ocean.

Audubon, in the first volume of his Ornithological Biography (p. 115), gives a very full and interesting description of the habits &c. of this species. He remarks that," the circumstance of their leaving the United States so early in autumn [they leave Boston, &c., about the 20th of August] has inclined me to think that they must go farther from them than any of our migratory land birds," p. 120. p. 120. He adds in vol. v. p. 408, that "although this beautiful swallow reaches the vicinity of the Arctic Circle earlier than others, it is said to migrate far within the tropics, as, according to Mr. Swainson, it was observed in numbers around Pernambuco 81° south of the line."

Is a regular summer visitant.

ALTHOUGH this bird is common in favourite localities, the species must be noticed with reference to Ireland generally, as but partially distributed. Along much of the wild western range of the island it is rather scarce, and in some extensive districts is never to be met with. It is stated by Mr. J. V. Stewart to be rare in Donegal. During a week of the fine summer weather spent in the north-west of that county in June, 1832, not a bird of this species came under my notice. The swift, unknown in Connemara (M'Calla); is very rare about Tralee, where it never breeds. (Neligan); and in the county of Kerry one pair only was ever seen by Mr. R. Chute. They remained but for a few days frequenting an old windmill in the village of Blennerville.

The swift is more regular as to the time of its appearance around Belfast than any species of the allied genus Hirundo, or, indeed, than any migratory bird. It may generally be seen during the first week of May, and frequently on the second day of that month :—even in 1842, when the arrival of birds generally was remarkably late, it was noticed on the 10th of May. Mr. Selby, in his Illustrations of British Ornithology, observes: "It has been remarked that these birds delight in sultry weather, with approaching thunder-storms, at such times flying in small parties with peculiar violence: and as they pass near steeples, towers, or corners of buildings, uttering loud screams, which White, in his Natural History of Selborne, supposes to be a sort of serenade to their respective families. This is fanciful and pretty; but I should rather be inclined to reason the opposite way, and to consider this action and cry as the consequence of irritability, excited by the highly electrical state of the atmosphere at such times." I agree with Mr. Selby in considering the remark of White respecting the scream of the swift to be merely "fanciful and pretty," as I have heard these birds scream in the manner

described, so soon after their arrival, as to afford sufficient proof that the cry did not proceed from the "males serenading their sitting hens," as incubation had not then commenced.* Nay, from the time that it has ceased until that of the bird's departure, the screaming is continued to the same degree as at any other period. But I cannot coincide in the opinion that "this action and cry are the consequences of irritability excited by the highly electrical state of the atmosphere at such times." This idea differing from my own, previous to the publication of the admirable work in which it appeared, I gave some attention to the subject for two summers, to see how far my preconceived opinion was justified. In the years 1832 and 1833, from the 7th and 9th of May, the days on which the swifts first came under my observation about Belfast, until the 1st and 3rd of June (when I left home), they daily kept flying about in small parties, screaming loudly, in dull and gloomy, as well as in bright and cloudless weather.

The following particular notices on this subject are abbreviated from my Journal :—

May 24th, 1832.--For the last eight or ten days the swift's scream has been daily heard; and when present this evening at the meeeting of an Historic Debating Society, the swifts obtruded themselves on my attention by flying, “in small parties,” closely past the windows, screaming most furiously. Though amusing to the ornithologist, it must have been very annoying to the assembled company, to be "serenaded" by their ill-timed scream, which not only jarred most discordantly with the 'eloquent music" discoursed within, but for the time being entirely drowned the voices of the speakers; indeed almost seemed to be intended as a mockery of what was passing there. During these ten days the weather has been rather dark and cloudy; the barometer remarkably stationary, and very high. With the exception of a few showers on one day, no rain has fallen.

May 27th and June 3rd, 1832.-Weather remarkably fine and warm: sky almost cloudless. The screaming of swifts heard above every other sound, about the localities frequented by them.

* In the last two years,-1847 and 1848,-my attention was directed to the earliest swifts of the season which I saw at Belfast by their loud screams. The date was the 9th of May in both years. In the former year, they had been seen over that town by another person on the 7th of the month. On the 4th, they were observed between Newry and Portadown. I have often remarked what doubtless led White to conjecture that the cry of the swift is the serenade of the males to "their sitting hens," as, at the season of incubation, these birds (but of which sex I cannot say) may often be observed flying about in the neighbourhood of their nests, and heard screaming only "when they come close to the walls or eaves."

May 22nd, 1833. After eight this evening, which was very warm and the sky cloudless, swifts, as noisy as usual, were flying about in little parties of three or four: two of these parties would occasionally join, and continue together for a short time, screaming vociferously. Their evolutions with that accompaniment, have always seemed to me manifestations of pure enjoyment. When these parties were about to meet, and when just separating, their power of screaming was exerted to the utmost.* Evolutions, in which a much greater number of these birds participated, were witnessed on the 24th inst., the weather being similar to that on the 22nd. The state of the barometer and weather has been mentioned, that some idea may be formed whether or not the atmosphere could have been "highly electrical " throughout the varied weather described, or indeed daily throughout that of any two months in this climate.+

Swifts particularly delight in flying about the squares and large open spaces in towns. Lofty edifices, especially when in a state of dilapidation, are preferred by them for building in; but in the north of Ireland, where these do not often occur, they content themselves with more humble dwellings. In many of our northern towns (where swifts are as plentiful as in any country) they select as their domicile the eaves of the oldest houses, or those from which the fast encroaching spirit of improvement, has not yet banished the thatched roofs. On the 8th of July, 1833, I observed many of these birds flying under the eaves and clinging to the walls of occupied two-story houses of this kind, in the town of Antrim. Although they and the martins appeared an indiscriminate multitude when flying about the streets, their places of nidification were quite distinct, the martins building on the south, and the swifts confining themselves to the north side: on a house just opposite the chief abode of the latter, I reckoned

*Mr. Macgillivray remarks, "that the loudest and most frequent cries are heard when birds are evidently in active and successful pursuit." At the times above alluded to, they certainly were not feeding.

+ Mr. Macgillivray, in his British Birds, vol. iii. pp. 619 and 622, enters fully into the subject of the swift's screaming. His observations of 1837 very generally agree with mine, made a few years before. Dr. J. D. Marshall, in his memoir on the Statistics and Natural History of the island of Rathlin, where swifts are plentiful, states that the result of his observations is opposed to the views of White and Selby. He believes the loud screaming of these birds to be particularly induced by fine weather and an abundance of food.

When on Ram's Island, in Lough Neagh, in the month of June, 1833, I remarked several of these birds flying in the vicinity of the ancient round tower, whose "rents of ruin " were most probably their breeding place.