Head, still furrowed by the defensive lines of the old Saxon invaders, stands out in the long, hollow range of coast reaching from Hurst Castle to Studland Bay, and on to Durlston Head, causing the shore line to present the plan of a double curve, somewhat similar to that marked in the air by the wings of a large bird. When the ordnance survey of this part of England was completed, in 1811, the area immediately to the north of the promontory presented a plan closely resembling that of a Roman post. A true lagoon then reached for the 1 mile of distance from the confluence of the two streams to the bar thrown up by the tide, which ran in a north-easterly direction from the end of Hengistbury Head to a promontory on the opposite mainland. In the middle of this bar was an opening, which looks on the survey exactly like an artificial entrance between two well-built walls. Within was a capacious basin, into which, however, thin lines of sand protruded from the mouths of the rivers, like the berges we have before described in the cases of the Rhine and of the Nile. But, by .1848, when the fourth sheet of the survey of the south coast of England was completed by Captain Sheringham, R. N., the appareil littoral had undergone a marked change. The greater part of the former lagoon had been transformed into marsh or into meadow, through which the confluent streams ran in a distinctly marked and curving channel. The central opening in the bar had disappeared, having been entirely choked by the action of the waves, and the escape of the water now takes place through a mouth more than a mile eastward, below Highcliff castle. Thus the formation of the berge, that of the cordon or bar, that of the lagoon, and that of the ultimate marsh and meadow, are illustrated in this beautiful spot by careful and exactly dated surveys. The face of the shore within half a century has undergone far more change than is apparent on the secular walls of the noble priory church that has looked down for 800 years on the activity of the rivers. So pure and dry is the air that the graceful decorations carved by the Roman masons on the pannelled walls of the sacred building are as sharp and clear as if they had been cut within the century. It is the work of man here that assumes permanence, while that of nature undergoes such comparatively sudden change.

An approximate estimate of the area of the gathering grounds of the Rhone and its affluents has been given by Professor Ansted, in a paper on Lagoons and Marshes, which was read at the Institute of Civil Engineers on February 16, 1869. This paper gave the fullest account of the Rhone delta that

VOL. CXLV. NO. CCXCVII.

I

we have met with before the publication of the work of M. Lenthéric, and the detailed account of the lagoons may still be read with interest. The watershed drained by the Rhone is stated in this paper at 37,000 square miles; but a note gives a correction to the effect that French geographers have lately given as the drainage area, in France alone, 45,884 square miles. A discrepancy of this amount in a special study of the subject is, at all events, a proof that the subject is not yet thoroughly mastered. If we may rely on Mr. Ansted's figures, the area covered by lagoons and marshes is in the proportion of a little more than two acres to every square mile of watershed basin; or in round numbers, about the three-hundredth part of the larger area. As to the rainfall, the information is but fragmentary. From 1857 to 1864, the mean rainfall at Montpellier was 36.58 inches. Over the Camargue the rainfall is said to be about one-fifth less. But what occurs in the upper part of the river's course is unknown. A long series of observations, carried on at properly distributed points, is necessary in order to arrive at clear information on a subject so deeply affecting the well-being of France. That showers and storms of great violence occasionally burst on the cradle of the Rhone and its affluents is well known. In October 1868, as much as 7 inches of rain is said to have fallen in 24 hours in the neighbourhood of Montpellier. If we assume the English average of 36 inches as that of the water-shed basin of the Rhone, we shall find that the annual rainfall over that area gives a total quantity of 150 milliards of tons, or 66 per cent. more than the measured volume poured into the Mediterranean by the Nile. The chief value of this comparison is the lesson which it points as to the need for ascertaining discharge, as well as rainfall. From its confluence with the Atbara the Nile runs for 24 degrees of latitude without receiving a single affluent. Its loss by evaporation in that distance materially reduces its volume. Were its course sufficiently prolonged, not a drop of its water would reach the sea, except in the season of flood. Yet no doubt can be entertained that the rainfall over the watershed of the Nile must be enormously greater than that of the basin of the Rhone.

Some valuable hydrometric observations on the River Tiber have been abstracted, in the Foreign Transactions of the Institution of Civil Engineers, from the Giornale del Genio Civile of 1875. From observations taken for a period of eight years, Signor Venturoli has calculated that the mean amount of the water brought down by the Tiber is 10,000 cubic feet per second. In 1870 the total average flood of

water in the valley of the Tiber was 213,900 feet per second; the flood water being calculated to be double that of the Po in relation to the area of its basin. The rainfall area of the Tiber is estimated at 6,455 square miles. The rainfall registered at Perugia is considered to be equal to the average fall over the whole basin of the Tiber. This is stated by Venturoli at 348 inches; one-fifth of which is deducted for loss by evaporation and otherwise, leaving an annual supply of 27.3 inches for feeding the river. The advance of the delta of the Tiber is measured by the obliteration of the ancient ports of Trajan and of Claudius. According to plans collected by Sir John Rennie, the retrogression of the sea here is at the rate of about two yards per annum. But this is not so much the advance of a projecting delta, as the gradual augmentation of a line of sea-board of undetermined length, lying within the great curve of 110 miles of coast, stretching from Capo Farnesio to Capo D'Anzo. The solid deposit of the Tiber is not estimated in the paper cited.

The action of the river that drains the great LombardoVenetian basin possesses an importance, not only from engineering considerations but from historic associations, scarcely inferior to that of the movement of the Nile itself. A frequent feature in river systems is the confluence of one stream with another, often at an obtuse angle to its course, and often near its mouth. Not unfrequently it is the case that the affluent stream drains a different description of country from that which feeds the principal river. In such cases, the junction is that of a torrent with a stream of permanent flow, as in the instances of the Parana and the Uruguay, and of the Nile and the Atbara. In the Lombard plain a somewhat different arrangement has been effected by the engineering of nature. The Tanaro, rising in the Maritime Alps; the Po, springing from Monte Viso; the Dora Riparia, reaching from Mont Cenis; the Dora Grossa, descending from Mont Blanc and Mont St. Bernard, and the Sesia, flowing from Monte Rosa, converge above the confluence, near Pavia, of the Ticino with the united streams. A fan-shaped network of water-ways is thus formed, extending over a circle, roughly measured, of some 80 miles' radius from a point near Vercelli, and draining a basin girded by the loftiest summits of the Alps, and covering more than 20,000 square geographical miles of ground. The lakes of Como, Iseo, and Garda send down their surplus waters from the north and north-west to swell the main stream of the Po. But below Mantua, and through the area of that ancient Eridanic delta within which the lake of Comacchio, as well

as the lagoons of Venice, were gradually walled off from the Adriatic, the Adige and the Brenta now find channels parallel with that of the Po; and the waters of these streams mingle only in the Adriatic. Thus, while Venice may be said, from a geological point of view, to be situated on the delta formation of the Po, it is the action of the Brenta which is now filling her lagoons, and threatening to convert the most picturesque of Italian cities into an inland town. The whole coast from Trieste to Ancona may be regarded as the actual boundary line of a geological delta, in the middle of the sweeping curve formed by the base of which the present mouths of the Po are protruding their active formation, far in advance of the cordons of the two lagoon systems before mentioned. The lido, or cordon of sand bounding the Venetian lagoons, is pierced with deep water openings, or foci, which have owed their maintenance, from the date of the foundation of Venice in 1104, mainly to the fact that the ebb of the faint tide of the Adriatic lasts for only about a sixth part of the time of the flow. Thus a force of scour is attained, to which we have no exact parallel in the cases of the Rhone, the Tiber, or the Nile. The rise of ordinary spring tides does not exceed 2.8 feet. When counteracted by a north wind the flow is less than 14 inches in rise; aided by the scirocco, it has been known to attain a height of 4·3 feet.

It is the more necessary to collect due materials for forming a clear opinion of the action of the rivers in the vicinity of Venice, from the fact that M. Lenthérie attempts to establish an exact parallel between the littoral apparatus of this portion of the Adriatic and that which exists at Port Said. The ability shown by this writer in his examination of the delta of the Rhone, and in his description of the dead cities of the Gulf of Lyons, is such as to give weight to his remarks on any similar district. It is therefore desirable to note the very contrary conditions which prevail in the Venetian lagoons and in the Nilotic basin.

The rainfall descending on the southward versant of the Carnic Alps makes its way into the upper portion of the Adriatic, and sweeps the Gulf of Venice with an appreciable southward current. The sands and mud brought down by the Tagliamento, the Piave, and other streams, are thus partially carried towards Ancona. During the scirocco, which blows with great fury in the gulf, the alluvial matter is thrown upon the cordon. So far, however, has this influence been from permanently widening the Lido (which is only 350 metres in mean breadth), that it has been found necessary to face this

outer barrier with stone, protected by groins or ribs run out to the sea, for a length of four miles between Lido and Malamocco. These two entrances, thus defended, and the two smaller foci or openings of Foggia, Tre Porti, and the Piave, admit the tide when raised by the scirocco; and being aided by dredging, maintain an ample scour. The waters of the Brenta, which are full of solid material, were diverted into a canal, furnished with sluices, and by this means the silting up of the lagoons was for a long time reduced to a minimum. Great contention, however, has arisen among the Italian engineers on this subject; and the importance of the first principle contended for by Rennie and his school, that of a catch water drain for arresting the deposits brought down by flood water, was departed from in 1840, when the Austrian engineers turned the Brenta into the lagoon. Professor Zanon, in the 'Rivista Maritima' for October 1875, argues in favour of this course. But the result has been that over the entire bed of the lagoon, an area of some 50 kilometres by 10, the bottom has been warped up 75 metre since 1840, while the delta has advanced 7 kilometres, and is now within 3 kilometres of Chioggia. The silting up of the bottom alone shows a. deposit of 11,000,000 of cubic yards per annum, independently of the growth of the delta.

By departing from the principle that prevention is better than cure, the Italian engineers have thrown away much of the special advantage with which nature had endowed the port of Venice. But even in its present condition it has no such menacing a foe to dread as exists in the case of Port Said, where a strong littoral current sweeps not from but towards the head of the Levant, bearing with it as much of the enormous mass of the annual deposit of the Nile as is not now disposed of in prolongation of the delta, or in raising the sea bottom between the Rosetta mouth and the Pelusiac Bay. This travelling mass, on reaching the shore of Syria, is partly blown from the margin of the sea, and advances, in desolating dunes, over the once celebrated gardens of Jaffa, as well as over the now barren isthmus. There is no scour from Lake Menzaleh or from Port Said, and all that can be done is to keep up a continuous dredging, the amount of which has risen from 161,000 cubic yards in 1871 to 937,000 cubic yards in 1875. Henry Rawlinson, in his address to the Royal Geographical Society on May 22, 1876, refers to the observations made. by Staff-Commander Millard, in February and March 1875. on the littoral between Port Said and the Damietta mouth of the Nile. He refers to the gradual shoaling of the Bay of

Sir