or liaised Beaches in Scotland, ^c. 285 



the general parallelism of the line of terraces and the bed of 

 the stream. We would expect the one to be either more or 

 less inclined than the other. I do not assert that their incli- 

 nations strictly correspond, but they seemed to me to do so in 

 a general way. 



The different levels of the terraces might be accounted for 

 by assuming the existence of an ancient series of lakes in the 

 valley ; but I have already shewn that this hypothesis is inap- 

 plicable to those at Dunkeld. Neither can we resort to the 

 Glacial Theory for a solution of the problem ; for broad deep 

 masses of sand and gravel, with smooth and uniform surfaces, 

 like some of the terraces, cannot be lateral moraines. My be- 

 lief is, that they were deposited by the sea, and formed the 

 lateral portions of the bottom of an ancient firth like Loch 

 Long or Loch Etive. It is not improbable that a careful sur- 

 vey, with the levels accurately taken, would explain the ano- 

 malous facts I have adverted to. 



The principle laid down by Mr Darwin should be kept in 

 mind in investigating these phenomena, — ^that, supposing the 

 ocean to be the agent, it depends on a combination of cir- 

 cumstances, the form of the surface, the nature of the mate- 

 rials underneath, the presence or absence of currents, &c., 

 whether any terraces are formed, and if formed, whether they 

 are preserved. 



There is a pseudo-terrace accompanying the river, which 

 remains to be noticed. A border of flat meadow-land is foijnd 

 here, as in other Highland valleys, on one or both sides of the 

 river. It varies in height from one yard to five or six. It fol- 

 lows the ever-changing level of the Tay, and we can scarcely 

 avoid considering it as produced by the river itself, probably 

 in working its way down through the mass of gravel and sand 

 from its ancient and higher level to that which it now occu- 

 pies. Streams which meander, and have beds composed of 

 alluvial matter, level the bottoms of the valleys in which they 

 flow, by a slow but constant action on their banks. It consists 

 of a process of erosion and deposition going on simultaneous- 

 ly — sand, clay, and gravel, being constantly removed from the 

 upper sides of the peninsulas enclosed in the flexures of the 



VOL. XXXV. NO. LXX. OCTOBER 1843. U 



