no observers, but Mr Dixon and M. Quinet made their routes follow rivers and coast lines, whether there was evidence to support this idea or not.
Only to a certain extent can it be safely contended that the present route of a species is an indication of its earlier journeys, or that the direction of original dispersal is recapitulated in the present line of migration. Heredity, experience, and imitation would certainly tend to preserve and confirm the general direction; the shortest and easiest passage from food-base to food-base would become an hereditary route, unless circumstances arose which caused a change. Mr Cooke shows how there has probably been evolution of the route as well as of everything else concerned with a mutable animal. The fly-line across an arm of the sea may be lengthened if this lengthening means a corresponding advantage in reaching the desired haven. Thus the birds which now cross the Gulf of Mexico at its widest part, at one time probably coasted round the Gulf, as many do still, by the land-bridge of Mexico and Central America. The gradual straightening of this curve would shorten the journey both in time and distance, though lengthening the actual single flight across a portion of the sea. We can imagine a bird arriving in autumn at the mouth of the Mississippi, at first passing from Louisiana to Mexico, so as to save the time of travel through Texas. Generations
