Laws of River Action. 443 



which is unfortunate. There is no a priori reasoning possible in physical science. 

 No theory has ever yet been established, except based on careful experiment, in 

 any branch of science. I therefore submit even my imperfect experiments as 

 much better than mere theoretical views. 



The results of observation have, no doubt, corrected many erroneous views 

 about the flow of water ; and when similar labours are expended upon ice, we 

 shall obtain more accurate knowledge than at present. 



(8) There are many separate works on the theory of the action of glaciers, 

 which are physical agencies of far less importance than rivers. The modern 

 glacier theory turns on the experiment of regelation, that is of pieces of ice at 

 32 degrees joining together or freezing. This action I do not think has been 

 sufficiently limited or investigated. Snow evaporates at very low temperatures, 

 and this must be a source by which heat is consumed, and the surrounding snow 

 cooled. Regelation only occurs when the quantity of water to be frozen is so 

 small that the part that evaporates into the pores of the ice, cools down the rest 

 of the water below freezing point by the heat abstracted in evaporating the first 

 portion of the water. Regelation does not occur when two dry surfaces of ice 

 are rubbed together, because heat is then produced by friction. 



The bubbles or cavities in ice, if they still contain air, all appear to contain 

 air at a low pressure, or in an attenuated form. The instant two dry surfaces 

 of ice are rubbed together, part passes into water by the friction producing heat. 



RIVERS. 



(9) There is no special and exhaustive book written on rivers, but there are 

 many reports of engineers on different rivers, such as Humphreys and Abbott 

 on the Mississippi, Sir C. Hartley on the Mouths of the Danube, Revy on the 

 La Plata, and chapters about rivers are to be found, in books of physical 

 geography and general mechanics. Sir Proby Cauttley has published a most 

 valuable work on the Ganges Canal. 



The absence of a special book on rivers describing the physical geography 

 of their basins is surprising, considering the importance of the subject, scienti- 

 fically and commercially, and also in an agricultural and sanitary point of view. 



In the Rhine, the hard rocks near Bingen have raised the flood level 14 ft. 



I shall now try to prove to you that rivers cut out their beds and valleys to 

 such a slope as is shown in Fig. 5, so as to attain such a curve and cross 

 section as would give uniform velocity to the stream. Also that water generally 

 travels down stream, in a navigable river, every day atone rate, not overtaking 

 the water before it, and that glaciers have, from the same cause, the same ten- 

 dency to uniform mean motion. Floods and avalanches are both exceptional 

 occurrences. In Figs. 6 and 7 I give a diagram explaining uniform mean motion. 

 I believe that a river is a machine for inducing uniform motion in the water 

 flowing in it. This is its real function, but it has escaped notice. So little 

 was known by the public about the proper vise and function of rivers, that it 

 was supposed they were only intended for navigation, and for use as drains, 

 and not worthy of a book to themselves. Their strncture is, however, I think, 

 one of the most beautiful adaptations of simple mechanical means to perform a 

 most complicated office in nature. 



It was not until Mr. Smee, one of the Managers of this Institution, stopped 

 the Croydon Board of Health from turning a sewer into the River Wandle 

 that it was ascertained that the common law of England was sufficient to pre- 

 vent persons throwing what substances they pleased into rivers. 



The mechanism of rivers has not, I think, yet been understood, for everyone 

 has appeared to have the full expectation that any river would dispose of any- 

 thing that was thrown into it without further care. 



I am not going to attempt to-night anything more than to establish some 

 general propositions about the action and the motion of water in rivers, and of 

 ice in glaciers, and the causes it depends upon ; and to show how the surface 



