TRAUSTI EINARSSONwith Europe or America in the Lower Tertiary. On the contrary, the flora suggests lack of such a land connection.There is, in my opinion, one main evidence which suggests rather great Tertiary changes in the North Atlantic area. I am referring to the sea-bottom topography. This looks relatively fresh and is suggestive of considerable tectonic changes which might have taken place in the Upper Tertiary. There is one important tectonic event of which we know: the uphft of Iceland, Greenland, Scandinavia, and Britain from low lands to the present mountainous countries took place in the Upper Tertiary. At the same time it is possible that the ocean floor changed considerably.The geophysicists are today inquiring into the origin and history of ocean basins by way of seismic studies and through the examination of bottom sediments. It is advisable to await more such investigations in the North Atlantic and Arctic area before drawing further conclusions regarding the geographic development in this area in the Tertiary. But palynological studies also seem quite promising.As matters stand today, we must face the possibility that the North Atlantic Ocean existed in the Lowest Tertiary and that the various remnants of plateau basalts indicate as many separate and distinct volcanic regions. To see, for instance, how Iceland might have originated in a deep ocean, the following hypothesis may be considered (cf. Einarsson, 1960):1(1) A Deep-Lower Tertiary flora of warm character.(2) A mixed flora with warm and cool elements.(3) A Tertiary flora of Hoifell and Tjornes type, ranging perhaps from Upper Miocene to Upper Pliocene.(4) An Upper Tertiary and Pleistocene type flora.(5) A Pleistocene type flora.This scheme represents in a very broad sense a chronological order. But it has been revealed that rather early there were considerable fluctuations between "warm" and "cool" floras, whether or not this is a reflection of climatic alternations or of the repeated and rapid changes in the other external conditions which 1 have mentioned earlier.In western Iceland the localities of Skardstrond and the Northwestern Peninsula belong to Pflug's first type. Hredavatn and Stafholt in Borgarfjordur still belong to this group, but Sleggjulaekur, a little higher than Hredavatn, pertains to the second type. In the topmost plateau group we then have the horizon of Litlisandur (Hvalfjordur) with a leaf bed in loess above varve-clay and a moraine on a clearly glacier-striated floor (Fig. 3). This is the lowest known glacial horizon in western Iceland. Above it are certainly three reverse magnetic periods and three normal ones, and more probably the total number of magnetic periods above this horizon is closer to 10. Pollen has