ISSI.] '^'"'^ [Spencer. 



sands. But along some of tbc hillsides excavated so deeply in the drift, 

 we find old beaches resting nnconformably on boulder clay. 



Near the centre of the city of Hamilton, in the wider portion of the 

 Dundas valley, a well was sunk to the depth of over 1000 feet. This 

 well revealed a most interesting fact. Though known to me several 

 years ago, I did not apply it until recently to its true bearing, 

 since discovering the origin of the Dundas valley, Mr. J. M. Williams 

 sunk this well, at the Royal Hotel, in Hamilton. He told me several 

 years ago that he had to sink through 290 feet of boulders, before 

 coming to hard rock, thus causing the outlay of a large sum of money in 

 excess of his calculations. Unfortunately, this well-record has been lost 

 by fire. At that time, the fact was so fresh in his memory (improved by the 

 extraordinary cost of the well) that his statement could be relied on, being 

 experienced in well-borings. The mouth of this well is 63 feet above Lake 

 Ontario, and therefore the hard rocks are absent for a depth of 327 feet be- 

 low the lake surface. See section. Fig. 2. 



As the valley is five miles wide at this place, and as the well is only about 

 one mile distant from its southern side, it becomes apparent that the valley 

 in the centre must have been much deeper. Moreover, if we produce the 

 southern side of that portion of the valley, which is over two miles wide, 

 we find that the well is less than a quarter of a mile away from it. Now if 

 we connect the top of the Medina shales (240 feet above Lake Ontario) 

 with the base of the drift in the well, and produce it to the centre of the 

 valley, it would indicate a central depth of over 500 feet. At the base of 

 the drift there are nearly fifty feet of Medina shales, below which are the 

 Hudson River rocks (more or less calcareous and arenaceous, mixed with the 

 shales). This harder formation along the bed of a river would be less exten- 

 sively removed by aqueous action than the overlying Medina shales, espe- 

 cially as the pitch of the waters would be much lessened. Thisgraphic method 

 of calculation seems as perfectly admissible here as it does in determin- 

 ing other constants of nature. However, I have placed the estimated depth 

 in the section at about 70 fathoms below the lake surface, which depth is 

 perfectly compatible with the soundings of the lake of no very great dis- 

 tance to the eastward. Even this depth gives only very gentle slopes from 

 the sides of the river valley. It should be remarked that Burlington baj' 

 is excavated from stratified clays in places to a depth of 78 feel. But this 

 water is silting up comparatively quickly. 



Now we have seen that the deep excavation in the Dundas valley and 

 westward is cut through more than 250 feet of Niagara and Clinton rocks, 

 mostly of limestone, and to a depth in the IMedina shales, so that the total 

 known depth of the canon is 743 feet, but with a calculated depth in the 

 middle of the cliannel of about 1000 feet. This depth for a ccmon is not 

 extraordinary for Eastern America. In Tennessee there are river valleys 

 excavated to a depth of 1600 feet, and in Pennsylvania Mr. Carll reports 

 others to be equally deep. 



Again, this Preglacial river explains the cause of the present topography 



Pr.OC. AMER. PHILOS. SOC. XIX. 108. 20. PRINTED APRIL 11, 1881. 



