bodies, is exceedingly nearly (for small velocities sensibly) parallel to the sides, yet the variation of the speed from the side to the centre of the stream occasions a _ripple_, or molecular discontinuity, which inclines forwards from the sides to the centre of the stream at an angle with the axis depending on the ratio of the central and lateral velocity. The veined structure of the ice corresponds to the ripple of the water, a molecular discontinuity whose measure is not comparable to the actual velocity of the ice; and therefore the general movement of the glacier, as indicated by the moraines, remains sensibly parallel to the sides." This theory opens up to us a series of interesting and novel considerations which I think will repay the reader's attention. If the ripples in the water and the veins in the ice be due to the same mechanical cause, when we develop clearly the origin of the former we are led directly to the explanation of the latter. I shall now endeavour to reduce the ripples to their mechanical elements. The Messrs. Weber have described in their 'Wellenlehre' an effect of wave-motion which it is very easy to obtain. When a boat moves through perfectly smooth water, and the rower raises his oar out of the water, drops trickle from its blade, and each drop where it falls produces a system of concentric rings. The circular waves as they widen become depressed, and, if the drops succeed each other with sufficient speed, the rings cross each other at innumerable points. The effect of this is to blot out more or less completely all the circles, and to leave behind two straight divergent ripple-lines, which are tangents to all the external rings; being in fact formed by the intersections of the latter, as a caustic in optics is formed by the intersection of luminous rays. Fig. 48, which is virtually copied from M. Weber, will render this description at once intelligible. The boat is supposed to move in the direction of the arrow, and as it does so the rings which it leaves behind widen, and produce the divergence of the two straight resultant lines of ripple. [Sidenote: RIPPLES DEDUCED FROM RINGS.] [Illustration: Fig. 48. Diagram explanatory of the formation of Ripples.] The more quickly the drops succeed each other, the more frequent will be the intersections of the rings; but as the speed of succession augments we approach the case of _a continuous vein_ of liquid; and if we suppose the continuity to be perfectly established, the ripples will still be produced with a smooth space between them as before. This experiment may indeed be made with a well-wetted oar, which on its first emergence from the water sends into it a continuous liquid vein. The same effect is produced when we substitute for the stream of liquid a solid rod--a common walking-stick for example. A water-fowl swimming in calm water produces two divergent lines of ripples of a similar kind.