230 



SCIENCE. 



[Vol. XII. No. 302 



farther. Second, degree of comminution. If the particle is larger, 

 it will fall sooner ; if the particle is smaller, it will be carried far- 

 ther, for the smaller the particle the greater the supporting surface 

 in proportion to its volume. Third, the velocity of the water. If 

 the velocity is decreased, the excursion of the particle will be short- 

 ened ; if the velocity is increased, the excursion of the particle will 

 be lengthened. Fourth, the depth of the water. If the water is 

 shallow, the floating particle will sooner reach the bottom ; if the 

 water is deeper, the particle will be carried farther before it strikes 

 the bottom. In this subject, therefore, we have to consider the 

 specific gravity of the load, the comminution of the load, the veloc- 

 ity of the water, and the depth of the water. 



As the water runs down the channel, it may roll sediment along 

 the bottom. This is the driven load. Such sediment is moved by 

 the impact of the water from above. But in order to do this the 

 materials on the bottom of the water must present up-stream sur- 

 faces on which impact may act ; that is, the bottom of the channel 

 must present heterogeneity of surface. This heterogeneity may be 

 of such a nature that the passing water may by impact lift the parti- 

 cles from the bottom so that they will be transported in the vehicle 

 by their own gravity. To the extent that materials are rolled along 

 the bottom by impact, the energy of the water is utilized in trans- 

 portation ; but to the extent that transportation is accomplished by 

 flotation, the gravity of the particles themselves is the entire force 

 of transportation. Whatever is driven is transported by the energy 

 of the water ; whatever floats is transported by its own inherent 

 gravity. This statement is made fully, because it is fundamental, 

 and because the principles involved have been neglected and seri- 

 ous error has arisen therefrom. 



The particles floating in a stream collide, and there arises there- 

 from inter-particle friction, but if in the collision between two par- 

 ticles one is retarded, the other must be accelerated. If the 

 particles are broken or ground by the process, work is done, and 

 the energy involved must be derived from the total energy of the 

 moving water and load. Some energy, therefore, must be lost by it, 

 but the disintegration arising therefrom promotes transportation, as 

 the smaller particles make longer excursions. Heat is also de- 

 veloped and dissipated, but perhaps the quantities involved are 

 unworthy of consideration. 



There is a degree of comminution that so approximates molec- 

 ular disintegration that some geologists and chemists believe that 

 a qtiasi o'c pseiido combination between the water and the load re- 

 sults therefrom : if this be the case the character of the fluid is 

 changed and the degree of fluidity is diminished. Here, again, it 

 may be possible that the quantities involved are so small that they 

 may be neglected. 



The volume of water remaining the same, if the velocity of the 

 water is increased, the depth of the water is diminished. There- 

 fore the excursion of the particle will be lengthened by the increase 

 of velocity, but shortened by the decrease of depth, and the one 

 compensates the other. On the other hand, to increase the veloc- 

 ity of a stream enables it to drive larger particles ; and this ability 

 increases with the sixth power of the velocity. 



The load increases the volume of the stream to the amount 

 measured by its own volume, and the load increases the mass of 

 the stream to the amount of its own mass ; and as the load is of 

 higher specific gravity than the water, the mass is increased at 

 a higher rate than the volume. As load increases volume, it there- 

 by increases velocity ; and as load still further increases mass, it 

 still further increases the effective energy of gravity. 



In order that transportation by flotation may begin, the detritus 

 must be loaded in the water, and when it sinks to the bottom it 

 must be reloaded that flotation may be continued. In erosion, 

 loading is primarily effected by the impact of raindrops. This load- 

 ing is continued, and reloading is accomplished by the flow of the 

 water over the surface through its impact against obstructions, and 

 thus the wash of the surface is carried into the stream. But the 

 load in the water sinks when, if the declivity is sufficient, it will be 

 driven, but if the declivity is insufficient, conditions for its reload- 

 ing must be produced. The driven load often becomes floating 

 load when the water plunges over great declivities, but the chief 

 method of reloading is by lateral corrasion : this arises in the case 

 of deposits which are built up until they become portions of the 



banks of a stream and are subsequently attacked by the stream 

 and carried away. Reloading is therefore chiefly accomplished by 

 the process of lateral corrasion. 



As the load is of greater specific gravity than the water, all load 

 is over-load in the sense that the load must be deposited because 

 the water is unable to permanently hold it in suspension. An ex- 

 treme condition of load may be reached, which is sometimes ex- 

 hibited in nature, in which the particles are so crowded that they 

 cannot move freely in suspension ; that is, they are in part held in 

 suspension by the water and in part supported in position by one 

 another, and still the particles may be so fine that they will slowly 

 move down stream : in this case the water moves faster than the 

 particles, and is strained through them at varying degrees. Thus 

 partial hydraulic suspension may exist. At the one extremity the 

 suspension may be so nearly perfect that the load is scarcely re- 

 tarded thereby. At the other extremity the movement of the load 

 may be wholly stopped and the water will be strained through it. 

 Under such conditions streams may disappear from the surface 

 and run wholly underground, and may re-appear when the ac- 

 cumulated sands have been passed. The accumulated sands may 

 be of such an extent as to absorb all the water and hold it until it 

 evaporates. Streams that thus empty into dry valleys and sand 

 plains are abundant in arid regions. 



The friction of pure water is so slight that where the formations 

 are hard, corrasion cannot be accomplished thereby (all processes 

 of solution are here neglected) ; but where the formation is inco- 

 herent, corrasion may progress through the impact of the water 

 against the more or less disintegrated particles lying at the surface 

 of the bottom and banks or on the ' wetted perimeter.' In this 

 case the particles must present surfaces up stream, against which 

 the flowing waters may act. The surface of the bottom and banks 

 must be heterogeneous. Such disintegration must be accomplished 

 by some instrument, and this is the load which passes along the 

 channel in course of transportation ; and it may be affirmed, that, 

 other things being equal, the greater the load the greater the cor- 

 rasion, and the less the load the less the corrasion. Again, the 

 banks of the stream may be disintegrated by sapping, and loaded 

 on the water by gravity, and the rate of lateral corrasion will be 

 greatly increased thereby. But corrasion furnishes additional load, 

 and it may be further affirmed that the greater the corrasion the 

 greater the load, and the less the corrasion the less the load. 



Rain is discharged from the surface of the land by flowing in the 

 direction of greatest declivity, and as multiform heterogeneous and 

 opposing declivities meet, the line of junction becomes the channel. 

 A stream may thus have a line of maximum depth. If the channel 

 is straight and homogeneous the line of maximum depth is the cen- 

 tral line of the stream, which in many cases may be a broad zone ; 

 but it is deflected, now to one side and now to the other, by curva- 

 ture and by a multiplicity of other conditions. The instrument of 

 corrasion is the load, and chiefly the driven load which is drawn 

 toward the line of maximum depth down the opposite declivities of 

 the wetted perimeter by gravity, and thus corrasion is more or less 

 concentrated along the line of maximum depth. Finally, the line 

 of maximum depth is the line of greatest velocity, where impact is 

 at a maximum. Hence it may be affirmed that in vertical corrasion 

 the line of maximum depth is the line of maximum corrasion. 



The geological formations into which channels are cut are greatly 

 varied in constitution : they may be granite, basalt, limestone, sand- 

 stone, clay, alluvium, etc. Many degrees of hardness and cohe- 

 rence are presented in these varying materials. The beds them- 

 selves are not co-extensive with the land, but are always limited by 

 the conditions of their production. Every such formation is a 

 comparatively small bed of sand, gravel, clay, limestone, granite, 

 etc., as the case ma> be, and the bed itself is variable in structure, 

 as in thickness, hardness, and general constitution, through many 



Geological formations are primarily horizontal, but subsequently 

 they may be tilted by diastrophic agencies, so that some beds are 

 horizontal, some are inclined at varying degrees, and others are 

 vertical. Where in its course a stream passes from bed to bed, the 

 conditions of corrasion are changed ; the harder bed corrades with 

 more difficulty, the softer bed with less. Where the beds are ver- 

 tical, this change along the course is at a maximum ; where the 



