The Palæontological evidence in favour of destruction of ancient continental areas and their replacement by the sea is more satisfactory than that which is based on physical grounds. The distribution of the Glossopteris flora of the Permo-Carboniferous period points to the former existence of a great southern continent, including the sites of Australia, India, South Africa and South America,--the Gondwanaland of Prof. E. Suess[124]. [Footnote 124: On this question and that of the other destroyed continental areas noted here, see W. T. Blanford's _Presidential Address_, _loc. cit._] Again, a study of Jurassic and Cretaceous faunas has led palæontologists to conclude that there was a connexion betwixt S. Africa and India in Mesozoic times across a portion of the area now occupied by the Indian Ocean, and also between S. Africa and S. America, and these inferences are supported by study of the distribution of existing forms. The sudden appearance of the Dicotyledonous Angiosperms in Upper Cretaceous rocks has also been used as evidence of destruction of considerable tracts of land subsequently to Upper Cretaceous times, and there is a certain amount of evidence in favour of the existence of this land in the north polar region, in an area now largely occupied by water, though relics of it are left, as the Faroe Isles, Spitsbergen, Novaya Zembla and Franz Josef Land. I cannot conclude the consideration of the question of permanence of oceans and continents more fitly than by quoting from Dr Blanford's address. He says, "There is no evidence whatever in favour of the extreme view accepted by some physicists and geologists that every ocean-bed now more than 1000 fathoms deep has always been ocean, and that no part of the continental area has ever been beneath the deep sea. Not only is there clear proof that some land-areas lying within continental limits have at a comparatively recent date been submerged over 1000 fathoms, whilst sea-bottoms now over 1000 fathoms deep must have been land in part of the Tertiary era, but there are a mass of facts both geological and biological in favour of land-connexion having formerly existed in certain cases across what are now broad and deep ocean[125]." [Footnote 125: _Loc. cit._, _Proc._ p. 107.]