Of course, I am not contending in the above remarks upon food that there is no physiological difference between a dog and a sheep in the matter of their digestive organs, and that the one is not by the nature of its body more fitted for one kind of food than the other; but rather that we should not neglect the importance of mere habit in such matters. Custom changed first; the change of physiological structure followed slowly after. What happened was probably something like this. Some time in the far back past a group of animals, driven perhaps by necessity, took to hunting in packs in the woods; it developed a modified physical structure in consequence, and special habits which in the course of time became deeply fixed in the race. Another group saved its life by taking to grazing. Grass is poor food; but it was the only chance this group had, and in time it got so accustomed to eating grass that it could not imagine any other form of diet, and at first would refuse even oysters when placed in its way! Another group saw an opening in trees; it developed a long neck and became the giraffe. But the fact that the giraffe lives on leaves, and the sheep on grass, and the wolf on animal matter, and that custom is in each so strong that at first the creature will refuse any other kind of diet, does not of itself prove that that diet is the best or most physiologically suitable for it. In other words, it is an assumption to suppose that "adaptation to environment" is the sole or even the main factor in the constitution of well-marked varieties or genera; for this is to neglect (among other things) the force of mere use or wont, which has about the same import in race-growth that momentum has in dynamics; and causes the race, once started in any direction, to maintain its line of movement--and often in despite of its environment--even for thousands of years.