212 FRAGMENTS OF SCIENCE 



sand and gravel of the island, tlie same fluviatile shells 

 as are now found in the Niagara Eiver higher up. It 

 would also account for those deposits along the sides of 

 the river, the discovery of which enabled Lyell, Hall, and 

 Eamsay to reduce to demonstration the popular belief that 

 the Niagara once flowed through a shallow valley. 



The phj^sics of the problem of excavation, which I made 

 clear to my mind before quitting Niagara, are revealed by 

 a close inspection of the present Horseshoe Fall. We see 

 evidently that the greatest weight of water bends over the 

 very apex of the Horseshoe. In a passage in his excellent 

 chapter on Niagara Falls, Mr. Hall alludes to this fact. 

 Here we have the most copious and the most violent whirl- 

 ing of the shattered liquid; here the most powerful eddies 

 recoil against the shale. From this portion of the fall, in- 

 deed, the spray sometimes rises without solution of conti- 

 nuity to the region of clouds, becoming gradually more at- 

 tenuated, and passing finally through the, condition of true 

 cloud into invisible vapor, which is sometimes reprecipitated 

 higher up. All the phenomena point distinctly to the cen- 

 tre of the river as the place of greatest mechanical energy, 

 and from the centre the vigor of the fall gradually dies 

 away toward the sides. The Horseshoe form, with the 

 concavity facing downward, is an obvious and necessary 

 consequence of this action. Bight along the middle of the 

 river the apex of the curve pushes its way backward, cut- 

 ting along the centre a deep and comparatively narrow 

 groove, and draining the sides as it passes them.' Hence 

 the remarkable discrepancy between the widths of the 



* In the discourse the excavation of the centre and drainage of the sides 

 action was iUustrated by a model devised by my assistant, Mr. John CottrelL 



