35 



example of this than in the case of the Niagara river. Here a deep channel has 

 evidently been cut out by the water constantly falling over the rocks, for a distance 

 of about seven miles, that is, from Lake Ontario to the present Falls. Sir Charles 

 Lyell states that after careful inquiries and observation on the spot, he came to 

 the conclusion that the Falls probably vi^ere receding at the rate of about one 

 foot a year, and he remarks that if this rate was uniform it would have taken 

 the river more than 37,000 years to have cut its way back from the cliffs on Lake 

 Ontario to the present position of the Falls. But for the rate of denudation to 

 be uniform, it would be necessary not only that the velocity of the river remained 

 constant, but it also requires the rocks over which the Falls have descended to be 

 of uniform hardness throughout. Now, this is a very improbable supposition, 

 and one which is rarely found to hold true for any great length of rock ; and in 

 the particular case of the ravine cut out by the Niagara river it certainly is not 

 the case. The uppermost rock over which the Falls now descend is a hard lime- 

 atone of the Silurian period, and underneath it is a soft shale which would be quickly 

 worn away compared to the hard limestone ; and there is evidence that about 

 four miles from the present Falls at some remote period a barrier of very hard 

 rock must have existed which retarded the action of the Falls for ages, and 

 allowed of the fluviatile deposit now found there to accumulate. This single 

 instance must suffice to show the necessarily unsatisfactory nature of the evidence 

 afforded as to time even by the most favourable case, and I must now hasten on 

 to make a few remarks upon glaciers. Glaciers, like rivers, wear away and carry 

 off quantities of material from the rocks through which they pass, and in 

 cases where the glacier melts and runs into the sea during the summer, to some 

 extent this material is often borne to a great distance ; but in all glaciers, to a 

 greater or less extent, there is at their termination what is called a terminal 

 morraine, consisting of fragments of rock and debris of all kinds brought down 

 by the glaciers. Now, it can, to some extent, be ascertained what amount of 

 accumulation is annually deposited by a glacier at its terminal morraine, and 

 therefore by measuring the length and depth of the morraine, some notion of the 

 time during which the glacier may have existed in its present form may be 

 obtained. So far, this method of estimating geological time is perhaps fairly 

 satisfactory, but it must be remembered that any subsidence of the earth's surface 

 at that place where the glacier is, would raise the temperature and would probably 

 destroy the glacier, as would likewise a rise of temperature from any other cause, 

 and its depositing action would for the time cease, to be resumed again upon a 

 subsequent elevation of level or depression of temperature. It is somewhat 

 beyond the scope of this paper to enter into any discussion as to the causes of 

 glacial periods, but it may not be quite out of place if I point out that there are 

 two chief causes for glacial periods, one of which will account for a general or 

 universal glacial epoch over the higher latitudes of one hemisphere, and the other 

 for local and isolated glacial periods, and it is very necessary that geologists 

 should carefully distinguish between these two causes of a glacial action. The 

 one cause for a general glacial epoch in the northern hemisphere may be found in 

 astronomical considerations connected with the eccentricity of the earth's orbit, 



