Let us briefly sum up the last two chapters on the pigeon. We may conclude with confidence that all the domestic races, notwithstanding their great amount of difference, are descended from the _Columba livia,_ including under this name certain wild races. But the differences between the latter throw no light whatever on the characters which distinguish the domestic races. In each breed or sub-breed the individual birds are more variable than birds in a state of nature; and occasionally they vary in a sudden and strongly-marked manner. This plasticity of organisation apparently results from changed conditions of life. Disuse has reduced certain parts of the body. Correlation of growth so ties the organisation together, that when one part varies other parts vary at the same time. When several breeds have once been formed, their intercrossing aids the progress of modification, and has even produced new sub-breeds. But as, in the construction of a building, mere stones or bricks are of little avail without the builder’s art, so, in the production of new races, selection has been the presiding power. Fanciers can act by selection on excessively slight individual differences, as well as on those greater differences which are called sports. Selection is followed methodically when the fancier tries to improve and modify a breed according to a prefixed standard of excellence; or he acts unmethodically and unconsciously, by merely trying to rear as good birds as he can, without any wish or intention to alter the breed. The progress of selection almost inevitably leads to the neglect and ultimate extinction of the earlier and less improved forms, as well as of many intermediate links in each long line of descent. Thus it has come to pass that most of our present races are so marvellously distinct from each other, and from the aboriginal rock-pigeon. REFERENCES [1] Temminck ‘Hist. Nat. Gén. des Pigeons,’ etc., tom. i. p. 191. [2] I have heard through Sir C. Lyell from Miss Buckley, that some half-bred Carriers kept during many years near London regularly settled by day on some adjoining trees, and, after being disturbed in their loft by their young being taken, roosted on them at night. [3] ‘Annals and Mag. of Nat. Hist.,’ 2nd ser., vol. xx., 1857, p. 509; and in a late volume of the Journal of the Asiatic Society. [4] In works written on the pigeon by fanciers I have sometimes observed the mistaken belief expressed that the species which naturalists called ground-pigeons (in contradistinction to arboreal pigeons) do not perch and build on trees. In these same works by fanciers wild species resembling the chief domestic races are often said to exist in various parts of the world; but such species are quite unknown to naturalists. [5] Sir R. Schomburgk in ‘Journal R. Geograph. Soc.,’ vol. xiii., 1844, p. 32. [6] Rev. E. S. Dixon ‘Ornamental Poultry,’ 1848, pp. 63, 66. [7] ‘Proc. Zoolog. Soc.,’ 1859, p. 400.