The relation of life to the earth's area is a fundamental question of bio-geography. The amount of that area available for terrestrial life, the proportion of land and water, the reduction or enlargement of the available surface by the operation of great cosmic forces, all enter into this problem, which changes from one geologic period to another. The present limited plant life of the Arctic regions is the impoverished successor of a vegetation abundant enough at the eighty-third parallel to produce coal. That was in the Genial Period, when the northern hemisphere with its broad land-masses presented a far larger area for the support of life than to-day. Then the Glacial Period spread an ice-sheet from the North Pole to approximately the fiftieth parallel, forced back life to the lower latitudes, and confined the bio-sphere to the smaller land-masses of the southern hemisphere and a girdle north of the equator. The sum total of life on the globe was greatly reduced at the height of glaciation, and since the retreat of the ice has probably never regained the abundance of the Middle Tertiary; so that our period is probably one of relative impoverishment and faulty adjustment both of life to life and of life to physical environment.[292] The continent of North America contained a small vital area during the Later Cretaceous Period, when a notable encroachment of the sea submerged the Atlantic coastal plain, large sections of the Pacific coast, the Great Plains, Texas and the adjacent Gulf plain up the Mississippi Valley to the mouth of the Ohio.[293] The task of estimating the area supporting terrestrial life which the earth presented at any given time is an important one, not only because the amount of life depends upon this area, but because every increase of available area tends to multiply conditions favorable to variation. Darwin shows that largeness of area, more than anything else, affords the best conditions for rapid and improved variation through natural selection; because a large area supports a larger number of individuals in whom chance variations, advantageous in the struggle for existence, appear oftener than in a small group. This position is maintained also by the most recent evolutionists.[294]