are demanded of it. Its faculties have been developed and specialized in the direction sought by the breeder. Man does not create in selecting; he exerts no immediate influence on the production of variability. He contents himself with exposing organized beings, for a special purpose, to new conditions of existence. Nature then acts on the organization and causes it to vary. Man chooses the variations which Nature furnishes and accumulates them. This is the principle, the application of which has given us races of pigeons with very different aptitudes. For example, French and Belgian breeders select with a view to success in the races, and often specialize the instinct of their pigeons. Birds from the same stock will, for instance, be trained for generation after generation to the east-to-west direction; and if we take a pigeon without being acquainted with the special aptitude of its ascendants, and try to train it to the north-to-south direction, we shall probably meet with mishaps. In England, where much fog prevails, the breeders keep only the birds that can fly through a misty atmosphere. The English breeds have consequently a capacity for finding their way in weather which would often baffle the pigeons of other countries. For like reasons pigeons raised in Sweden and Norway are able to return to their homes in the face of snow, which often puts the instinct of French pigeons to fault. The training of pigeons at sea requires special aptitudes, which a rational breeding will develop by selection.
We read in books on pigeon culture that the carrier pigeon is hardly ever white. The reason for this is very simple: pigeons on their journey are selected by birds of prey, which most readily pick out those of conspicuous colors; consequently these birds disappear without having opportunity to found a stock. This observation does not apply so much to the common pigeon, which, never straying far from habitations, is less frequently struck by the hawk. So pigeons flying near the ground are certain to fall sooner or later under the shot of the hunter, and usually leave very few descendants. This circumstance, independent of our will, often intervenes to play an important part in the transformation of a domestic species.
Selection permits us to adapt our races to any sort of service. We might, for example, create a stock of birds that would retain the recollection of their home for a very long period; we might develop the aptitude for traveling back and forth. We have sometimes asked ourselves what limit could be fixed to the utilization of the carrier pigeon. To fix a limit would be to deny the principle of transformability of species, which is a law of evolution. Our races are continuously undergoing modification, and are consequently capable of indefinite improvement. Instead of looking for limits to the employment of the pigeon, we should point out
