174 The American Geologist. ScptLMiiin-, 1^591; 
owing to the gradual iii)lift of tho eonntry on the north and 
east, including- the Rome outlet, tlie western i)art of lake Jro- 
(|Uois was caused to rise ujion its coast to a iiight of about 
200 feet, finally holding that level during many years, witli 
the formation of the well developed Irocpiois beach. The dif- 
ferential epeirogenic movement of the land and the conse(iuent 
westward rise of the lake, conlined by the waning ice-sheet on 
its northeast side, are known by the northeastward ascent of 
the old beaches and by drift sections which show two read- 
vances of the ice-front in the vicinity of Toronto during the 
gradual rise of the lake there, to which reference will be again 
made on a following page. 
Beginning of Niagara Hivek and its Ekosion of thk 
Gorge below the Falls. 
The departure of the ice-slieet, as we have seen, uncovered 
the northeastern part of the continent with so low an altitude 
that the drainage of the Laurentian lakes region then, for the 
first time, as I believe, began to outflow eastward, earliest by 
the Hudson, and later, with the farther glacial retreat, by the 
St. Lawrence. From the day when the Rome avenue of out- 
flow drew down the waters below the Niagara escarpment, 
which runs between the basins of lakes P^rie and Ontario, the 
Niagara river has been flowing over its falls and has been 
cutting its gorge backward from the escari)ment. At first this 
river was scarcely separated from the river Erie b}'' any lake 
in the Erie basin; but gradually the differential uplifting of 
the land raised the eastern rim of the basin, at Buffalo, to 
such hight that the extensive but shallow lake Erie covered 
the former river Erie, and the great level plain through which 
that river flowed is now the lake bed. 
Variations of the Vollme of Niagara River. 
Gilbert, in his able papers on the history of Niagara river, 
well reviews its probable changes of volume due to variations 
of the average yearly rainfall and snowfall, the melting of the 
ice-sheet during its departure, and other varying conditions. 
Most of these causes of variation of the river's volume seem 
to have been not widely different from those of tlie past thou- 
sand years or even of the past century, during which *its flow 
and the rate of erosion of the falls and gorge have been ob- 
.served. Some of the variations have doubtless tended to 
