Sir John Lubbock, in his dissent from some calculations made by Mr. Geikie on the general effect produced by rivers in excavating valleys and lowering the general level of the country, says, "As regards the higher districts, however, his data are perhaps not far wrong, and if we apply them to the valley of the Somme, where the excavation is about two hundred feet in depth, they would indicate an antiquity for the palæolithic epoch of from one hundred thousand to two hundred and forty thousand years."[7] Dana, in his chapter on the length of geological time, says, in speaking of the time required to excavate the gorge of Niagara River, that "on both sides of the gorge near the whirlpool, and also at Goat Island, there are beds of recent lake shells ... the same kinds that live in still water near the entrance to the lake, and which are not found in the rapids. The lake, therefore, spread its still waters, when these beds were formed, over the gorge above the whirlpool. A tooth of a mastodon (_M. giganteus_) has been found in the same beds. This locates the time in the Champlain epoch.... Six miles of the gorge have been excavated since that mastodon was alive.... "There is a lateral valley leading from the whirlpool through the Queenstown precipice at a point a few miles west of Lewiston. This valley is filled with drift of the glacial epoch, and this blocking up of the channel may have compelled it to open a new passage. "If, then, the falls have been receding six miles, and we can ascertain the probable rate of progress, we may approximate to the length of time it required. Hall and Lyell estimated the average rate at one foot a year,--which is certainly large. Mr. Desor concluded, after his study of the falls, that it was 'more nearly three feet a century than three feet a year.' Taking the rate at one foot a year, the six miles will have required over thirty-one thousand years; if at one inch a year--which is eight and one third feet a century--three hundred and eighty thousand years."[8] The calculation made by Dana is for the Champlain epoch. As this epoch was subsequent to the glacial, the time must be either thrown still farther back, or else allow the calculations to begin with the end of the glacial. _Probable Duration._--Lyell has attempted to form an estimate of the duration of the glacial epoch by considering "the most simple series of changes in physical geography which can possibly account for the phenomena of the glacial period," and enumerates as follows: "First, a continental period, toward the close of which the forest of Cromer flourished; when the land was at least five hundred feet above its present level, perhaps much higher, and its extent probably greater than that given in the map, Fig. 41." (In this map the whole of the British Isles are connected with one another, and with the continent--the German Ocean and the English Channel constituting dry land).