430 



NATURE 



{Feb. 28, 1889 



theory. The elastic theory, however, is regarded as somewhat 

 less convenient as a working hypothesis than the electric. — In 

 this number appears Part i of an exhaustive monograph, with 

 numerous illustrations, on the geology of the volcanic island of 

 Fernando de Noronha, South- West Atlantic, by John C. 

 Branner. 



The last volume (xviii.) of the Memoirs of the Kazati Society 

 of Naturalists contains an elaborate inquiry into the distribution 

 of solanin (an alkaloid discovered by Desfosses in vazxiy Solanacece) 

 in plants, by E. Wotschall ; short reports on geological ex- 

 ploration in the Governments of Vyatka and Ufa, by A. Netschaeff 

 and A. Lavrsky ; and a description of the flora of the neigh- 

 bourhood of Ufa, by A. Gordyaghin. 



SOCIETIES AND ACADEMIES. 

 London. 



Royal Society, February 7.— "The Principles of Training 

 Rivers through Tidal Estuaries, as illustrated by Investigations 

 into the Methods of improving the Navigation Channels of the 

 Estuary of the Seine." By Leveson Francis Vernon-Harcourt, 

 M.A., M.Inst.C.E. 



After stating the principles upon which the training of the 

 non-tidal portions of rivers are carried out, the undefined and 

 unsatisfactory condition of the principles followed in training 

 rivers through wide tidal estuaries, and the discordant views of 

 engineers on the subject, were pointed out. The absence of 

 definite piinciples, and the divergence of opinion amongst 

 engineers, have received a remarkable illustration in the great 

 variety of schemes proposed for extending the training walls 

 in the Seine estuary beyond Berville, where the works were 

 stopped, in 1870, owing to the unexpected changes the works 

 had already produced. It occurred to the author in August 

 1886, that if it should be possible to reproduce, in a workino^ 

 model, the original state of the Seine estuary before the training 

 works were commenced, and next the present state of the Seine, 

 as modified by these works, could be obtained, then the suc- 

 cessive introduction in the model of the several schemes, pro- 

 posed for the extension of the training walls, might furnish 

 results indicating approximately in miniature the changes which 

 the works would actually produce if carried out in the estuary, 

 and also afford a basis for the establishment of general principles 

 for training rivers through wide estuaries. A model, accord- 

 ingly, was made of the tidal Seine, to the scale of 1/40,000 

 horizontal, and 1/400 vertical ; the bed was formed of fine sand, 

 so that it could be moulded by the current ; the fresh-water 

 discharge was produced, at the upper end, by the flow of water 

 from a small cistern ; and the tidal ebb and flow were effected 

 by the tipping of a tray, placed at a suitable angle at the lower 

 end, representing the open sea. The model was first worked in 

 November 1886, and the experiments were continued at inter- 

 vals up to 1889. Silver sand was first used for forming the bed 

 of the miniature estuary ; and some of the phenomena of the 

 actual estuary, such as the bore, the '' verhaiiVe" or reverse 

 current, and the shifting channels, were reproduced in the 

 model ; but when the training walls were introduced into the 

 model, on the lines of the existing training walls, it was found 

 that the silver sand could not be adequately carried in suspen- 

 sion by the small currents in the model to reproduce the accre- 

 tion which has occurred in the estuary behind and beyond the 

 trammg walls. A variety of fine powders, of low specific gravity, 

 were consequently experimented on in the model, but they mostly 

 proved too sticky, or pasty, or otherwise unsuitable. At last 

 a fine sand from Chobham Common, belonging to the Bagshot 

 beds, containing an admixture of peat, offered better results, and 

 was employed for the subsequent experiments. 



After working the model for some time with a bed formed of 

 this Bagshot sand, the channels assumed a form very closely 

 resembling in general outline the chart of the Seine of 1834. 

 This result, by reproducing a former condition of the estuary, con- 

 firms the previous results obtained by Prof. Osborne Reynolds 

 with a model of the Upper Mersey estuary, showing that it is 

 quite practicable to reproduce in a model the main tidal channels 

 in an estuary. 



The second stage of the investigation involved the quite novel 

 condition of introducing training walls in a model, and producing 

 the resulting accretion. This most essential stage was the 

 subject of a long series of experiments, but was at last satisfac- 



torily accomplished with Bagshot sand. The existing training 

 walls were inserted in the model, and the resulting deepening of 

 the trained channel and the accretion outside and beyond were 

 reproduced in the model, and also the shifting channel between 

 the termination of the training walls and the sea. 



The third stage of the investigation was then entered upon, 

 consisting in the successive introduction in the model of the 

 lines of the five principal schemes at present advocated, observing 

 the changes they respectively produced in the model of the 

 estuary, and recording them in the form of charts of the estuary, 

 which are appended as plates to the paper. A final experiment 

 was also made with an arrangement of training walls forming as 

 gradually expanding a channel as practicable, without restricting 

 the width of the outlet. The lines designed for the extension of 

 the training walls in each scheme are briefly described in the 

 paper, as well as indicated in the charts, and also the channels 

 and accretion which they each produced. 



The probability of the results obtained really representing in 

 miniature the results which corresponding works in the estuary 

 of the Seine would actually produce was then considered ; for if 

 the effects of any training works could be foreshadowed by 

 experiments in a model, the value of such experiments, in guid- 

 ing engineers towards the selection of the most suitable design, 

 could not be over-estimated. Though the effects of winds and 

 waves, and the actual rate of accretion, cannot be reproduced in 

 a model, it is evident, from the first stage of the investigation, 

 that the main forces at work, in the comparative shelter of an 

 estuary, are the tidal ebb and flow and the fresh-water discharge, 

 which are the forces which can be produced in a model. More- 

 over, the correspondence of the second stage of the investigation 

 with the existing state of the Seine estuary confirms the accord- 

 ance between the results in the model and the condition of the 

 estuary. The extension of training walls decreases the influence 

 of winds and waves ; and therefore the results of the third stage of 

 the investigation are more likely to correspond with the changes 

 which such works would actually produce in the estuary, than 

 those of the first and second stages. Also the results obtained 

 in the model with the two earlier schemes are precisely those 

 which the author predicted would occur, before the experiments 

 were commenced, if the schemes were actually carried out in the 

 estuary. 



The paper concludes with a classification of the experiments, 

 with the view of deducing general principles for guidance in train- 

 ing rivers through tidal estuaries. The three classes are, (i) outlet 

 of estuary considerably restricted, and channel trained inside 

 towards outlet ; (2) channel trained in sinuous line, expanding 

 towards outlet, but kept somewhat narrow at changes of curva- 

 ture ; (3) channel trained in as direct a course as practicable, 

 and expanding regularly to outlet. 



The experiments of the first class exhibited a deep outlet, and 

 a fairly continuous channel inside where the training.works were 

 prolonged to the outlet. The channel, however, was irregular 

 in depth near the outlet ; and a bar appeared in front of the 

 outlet outside. The breakwater, also, extending across part of 

 the original outlet, occasioned deposits both inside and outside 

 the estuary, by producing slack water in the sheltered recesses. 



The second class of trained channel was designed to profit by 

 the well-known scour at the concave face of bends in non-tidal 

 rivers, and to continue the depth thus obtained by restricting the 

 width between the bends. Experiment, however, did not bear 

 out the advantages of this system, probably owing to the vari- 

 able direction of the flood tide at different heights of tide, 

 its being checked in its progress by the winding course, and not 

 acting in unison with the ebb, from the difference in its direction 

 and the width of the trained channel near the outlet. The third 

 class of trained channel afforded a wide channel, tolerably 

 uniform in depth, in the experiments ; the flood tide was less 

 impeded in its progress than with the other forms of training 

 walls, and appeared to act more in concert with the ebb. 



The experiments accordingly indicate that the only satisfactory 

 principle for training rivers through wide estuaries with silt- 

 bearing currents is to give the trained channel a gradually ex- 

 panding form, with as direct a course as practicable to the outlet. 

 The rate of increase in width between the training walls must be 

 determined by the special conditions of the estuary, 



February 14. — " On a Series of Salts of a Base containing 

 Chromium and Urea," No. 2. By W. J. Sell and Prof. W. J. 

 Lewis. 



The paper is a continuation of that published by one of the 

 authors (Proc, Roy, Soc, vol. xxxiii. p. 267). It is here shown 



