luctuosa, as they in fact belong to the genus Carpophaga. Fourthly, Columba guinea, which ranges from Guinea (6/8. Temminck 'Hist. Nat. Gen. des Pigeons' tome 1; also 'Les Pigeons' par Mme. Knip and Temminck. Bonaparte, however, in his 'Coup- d'oeil' believes that two closely allied species are confounded together under this name. The C. leucocephala of the West Indies is stated by Temminck to be a rock-pigeon; but I am informed by Mr. Gosse that this is an error.) to the Cape of Good Hope, and roosts either on trees or rocks, according to the nature of the country. This species belongs to the genus Strictoenas of Reichenbach, but is closely allied to Columba; it is to some extent coloured like certain domestic races, and has been said to be domesticated in Abyssinia; but Mr. Mansfield Parkyns, who collected the birds of that country and knows the species, informs me that this is a mistake. Moreover, the C. guinea is characterised by the feathers of the neck having peculiar notched tips,--a character not observed in any domestic race. Fifthly, the Columba oenas of Europe, which roosts on trees, and builds its nest in holes, either in trees or the ground; this species, as far as external characters go, might be the parent of several domestic races; but, though it crosses readily with the true rock-pigeon, the offspring, as we shall presently see, are sterile hybrids, and of such sterility there is not a trace when the domestic races are intercrossed. It should also be observed that if we were to admit, against all probability, that any of the foregoing five or six species were the parents of some of our domestic pigeons, not the least light would be thrown on the chief differences between the eleven most strongly-marked races.
We now come to the best known rock-pigeon, the Columba livia, which is often designated in Europe pre-eminently as the Rock-pigeon, and which naturalists believe to be the parent of all the domesticated breeds. This bird agrees in every essential character with the breeds which have been only slightly modified. It differs from all other species in being of a slaty-blue colour, with two black bars on the wings, and with the croup (or loins) white. Occasionally birds are seen in Faroe and the Hebrides with the black bars replaced by two or three black spots; this form has been named by Brehm (6/9. 'Handbuch der Naturgesch. Vogel Deutschlands.') C. amaliae, but this species has not been admitted as distinct by other ornithologists. Graba (6/10. 'Tagebuch, Reise nach Faro' 1830 s. 62.) even found a difference in the bars on the right and left wings of the same bird in Faroe. Another and rather more distinct form is either truly wild or has become feral on the cliffs of England and was doubtfully named by Mr. Blyth (6/11. 'Annals and Mag. of Nat. Hist.' volume 19 1847 page 102. This excellent paper on pigeons is well worth consulting.) as C. affinis, but is now no longer considered by him as a distinct species. C. affinis is rather smaller than the rock-pigeon of the Scottish islands, and has a very different appearance owing to the wing-coverts being chequered with black, with similar marks often extending over the back. The chequering consists of a large black spot on the two sides, but chiefly on the outer side, of each feather. The wing-bars in the true rock-pigeon and in the chequered variety are, in fact, due to similar though larger spots symmetrically crossing the secondary wing-feather and the larger coverts. Hence the chequering arises merely from an extension of these marks to other parts of the plumage. Chequered birds are not confined to the coasts of England; for they were found by Graba at Faroe; and W. Thompson (6/12. 'Natural History of Ireland' Birds volume 2 1850 page 11. For Graba see previous reference.) says that at Islay fully half the wild rock-pigeons were chequered. Colonel King, of Hythe, stocked his dovecote with young wild birds which he himself procured from nests at the Orkney Islands; and several specimens, kindly sent to me by him, were all plainly chequered.