the country every 5,000 years or so, it becomes perfectly obvious that the close of the glacial epoch must be of comparatively recent date. There is another circumstance which shows that the glacial epoch must be referred to the latest period of great eccentricity. If we refer the glacial epoch to the penultimate period of extreme eccentricity, and place its commencement one million of years back, then we must also lengthen out to a corresponding extent the entire geological history of the globe. Sir Charles Lyell, who is inclined to assign the glacial epoch to this penultimate period, considers that when we go back as far as the Lower Miocene formations, we arrive at a period when the marine shells differed as a whole from those now existing. But only 5 per cent. of the shells existing at the commencement of the glacial epoch have since died out. Hence, assuming the rate at which the species change to be uniform, it follows that the Lower Miocene period must be twenty times as remote as the commencement of the glacial epoch. Consequently, if it be one million of years since the commencement of the glacial epoch, 20 millions of years, Sir Charles concludes, must have elapsed since the time of the Lower Miocene period, and 60 millions of years since the beginning of the Eocene period, and about 160 millions of years since the Carboniferous period, and about 240 millions of years must be the time which has elapsed since the beginning of the Cambrian period. But, on the other hand, if we refer the glacial epoch to the latest period of great eccentricity, and take 250,000 years ago as the beginning of that period, then, according to the same mode of calculation, we have 15 millions of years since the beginning of the Eocene period, and 40 millions of years since the Carboniferous period, and 60 millions of years in all since the beginning of the Cambrian period.