TRANSACTIONS OF SECTION A. 575 



estuary at a time when a considerable depth of water (such as a few feet depth 

 along its mid-channel) was maintained in it by the height of the sea-water outside, 

 and when the slow rising and sinking of the ocean-waves was producing in the 

 estuary a tiux and reflux on a small scale like that of the tidal flow in large 

 estuaries.* The motions of the water being indicated by numerous little pieces of 

 sea-weed carried in suspension, I noticed that the water at or very close to the 

 channel-bed reversed its landward or seaward flow always much earlier than did 

 the main body of the water in the channel, less affected by contiguity to the bed. 

 The phenomenon being noticed, the reason at once became "apparent. The lamina 

 contiguous to the bed, or channel -face, would be always hindered, by the frictional 

 resistance of that face, from getting into so great a velocity, either seaward or land- 

 ward, as that which would be attained to by the main body of the water. Then, 

 when the water at the sea-end of the estuary was raised in level, by the arrival of 

 an ocean-wave, so as to give a gravitational propulsive influence tending to cause 

 the water to flow landward along the estuary, the main body of the water, in virtue 

 of its inertia with seaward momentum, would continue to flow for some time sea- 

 ward, flowing as it were uphill ; - while the frictionally restrained lamina at the 

 channel-face, being nearly devoid of inertial tendency seaward, would readily yield 

 to the landward gravitational propulsive influence due to the landward surface 

 declivity of the water in the estuary. 



Exactly a like explanation, mutatis mutandis, is applicable to the case of reversal 

 of the flow from having been landward to its becoming seaward. The channel-face 

 lamina makes its reversal of flow, just as in the other case, earlier than does the 

 main body of the water, and for like reason. 



It may now further be noticed that precisely corresponding phenomena would 

 present themselves in the flux and reflux of water in a pipe, if, for instance, the 

 pipe were connecting two cisterns, and a plunger were kept oscillating upwards and 

 downwards in one of them so as to cause the alternating flow through the pipe. 

 The phenomena might be very interestingly manifested in an open trough connect- 

 ing two cisterns, arrangements being made, by a plunger or otherwise, for causing 

 flux and reflux along the trough, and the motions of the water being indicated by 

 small visible particles in suspension in the water or by the dropping in of granules 

 of aniline dye. 



It may now be worthy of remark that the hydraulic principle brought into notice 

 in the present paper, in respect to flux and reflux along channels, is closely allied 

 to, and is in some respects identical with, the leading principle set forth in previous 

 papers by myself on the flow of water round bends in rivers, &c. In that case the 

 frictionally resisted and retarded lamina in contiguity with the channel-face, or bed, 

 flows transversely (or rather obliquely) across the channel towards the inner bank 

 of the bend, impelled inwards by gravitational pi'opulsive influence (that is, down- 

 hill as it were), while the main iDody of the stream, flowing quicker in the bend, 

 exerts centrifugal force outwards, or tends inertially out towards the outer bank. 

 The papers here referred to on Flow of Water round Bends in Rivers, &c., are to be 

 found in the ' Proceedings of the Royal Society 'for May 1876; in the British 

 Association Report for the Glasgow Meeting, 1876, Section A, page 31 of Trans- 

 actions of the Sections ; in the ' Proceedings of the Royal Society,' 1877, No. 182, 

 page 356 ; and in the ' Proceedings of the Institution of Mechanical Engineers,' 

 August 1879, p. 456. Also some other important cases in which like principles of 

 fluid motion come into play (in Whirlwinds &c.) are adduced in a paper by myself 

 in the British Association Report for the Montreal Meeting, 1884, Section A. 

 p. 641. 



' The period of these oscillations may be about from ten to twenty seconds, as I 

 have been informed that Professor Stokes has found, by observations on that coast, 

 that the period from one wave to the next, in the large Atlantic waves there, is at 

 most about seventeen seconds. 



- Or, in more precise terms, flowing from a place of lower to a place of higher 

 free-level. 



