The straightening of the river need not be considered, as it finds no supporters in the present day. The expenditure would be out of all proportion to the results obtained, which would be an addition of a foot, perhaps, or a foot and a half to the fall, and a slight increase in the velocity.

DIVERSION OF A PORTION OF ITS WATERS.

The most popular scheme in every age for moderating the inundations of the Tiber has been to divert a portion of its waters above Rome, and either restore them to the river at a lower level, or convey them to the sea by an independent channel. Some have supposed that the canal referred to by Pliny the younger in the following passage: "the Tiber, though relieved by the canal which the most wise of Emperors has made, inundates the valleys and floods the plains,”* commenced at some point above the Ponte Molle, and after being carried by a cutting through the Val del Inferno, was discharged into the country

surface of the river, and let AC be the fall in the distance AB. It is self-evident that, however much we may widen the river between A and B, we can never lower its level by a height equal to AC, because there must always be a fall from A to B, otherwise the river would cease to flow. Let AB be a mile and CA a foot, which exceeds the average fall of the Tiber in that distance in and below Rome, then by widening the river for a mile we could only lower its level at A by something less than a foot.

Nor would the level of the river in time of floods be lowered by more than this small difference.

For, since by supposition the breadth of the river at B remains unchanged, no more water can pass through it than before, and, consequently, no more water can be discharged by the widened part. Its level would not, therefore, be lowered relatively to the river at other points. On the contrary, the difference of level would be lessened, as I have shewn, by a rise in the river, and a portion of the advantage gained would be lost. If the river be widened, not, as I have assumed, for a mile, but only for two or three hundred yards, the lowering of the level in time of floods would be reduced to something trifling in the extreme.

* Tiberis quanquam fossa quam providissimus Imperator fecit exhaustus, premi valles, innatat campis (Plin. VIII. Ep. 17 to Macrinus).-The most probable supposition, however, is that the canal in question was the present Fiumicino arm, which is considered by many to be artificial, and to have been dug by the Emperor, to connect the Tiber with his newly-created port.

below, and left to find its way to the sea; a monstrous supposition, as if there were no rights of private property in those days, or as if even an Emperor could turn a river through the lands of his subjects without any regard to the damage it might cause. It would not be surprising if the Tiber had avenged itself for the unceremonious manner in which it was turned out of a back door by laying waste the country, and tainting the air.

A plan for diverting the Anio from the Tiber was brought forward in the time of Clement VIII. The canal was to commence at the Ponte Mammolo, and after passing close to the walls of Rome on the eastern side, so as to serve as a wet ditch for the protection of the city, was to rejoin the river at the church of Santo Paolo without the walls. It was thought, that not only would the Anio be cut off from the Tiber, but in time of floods, a portion of the waters of the latter would discharge themselves through the Anio into the canal. In the opinion of Bonini very little good would be effected by such a diversion, because there are often floods of the Tiber without any increase of the Anio. The Anio, besides, drains but a small portion of the basin of the Tiber, as will be seen by inspection of the map.

All these proposals for drawing off a portion of the Tiber in time of floods, and restoring it to the river at a lower point, remind me of the ingenious device of the Irishman for lengthening a piece of cloth, by cutting off a portion from the top and sewing it on to the bottom. Every one would be sensible of the folly of trying to empty a reservoir by taking a pailful out of one end, and discharging it into the other. Yet the plans proposed are only a degree less absurd, when we consider the slight fall of the Tiber between Ponte Molle and the sea, and the great expense which would be incurred to lower the level of the floods a foot or two; if indeed it were possible to lower them as much.

Let AB (fig. 5) represent the surface of the river between Ponte Molle and San Paolo, the inclination being greatly exaggerated to make it sensible to the eye, and let CB be

the horizontal line. It is evident that whatever quantity of water we abstract at A, and restore at B, we can never lower the surface at A below the point C, otherwise the water would run back again from B to C. According to Chiesa and Gambarini, the fall of the river from the Ripetta to Ripa Grande is four and eleven-twelfths palms, or about three and a half English feet. Let us assume the difference of level between Ponte Molle and Santo Paolo to be eight feet. The effect of a rise in a river, as will be hereafter shown, is to obliterate differences of level, and render the declivity of the surface more gentle along the whole course of the stream. The difference of level, therefore, will be considerably reduced, say to five feet, and since the level of the flood at B will not be lowered at all, the difference at the Ripetta would be the mean between AC and O, or two feet and a half. But the whole, even of this difference, would not be gained. As the velocity of a river is always greatest in the middle, and the canal at its origin would have but a slight inclination, the same as that of the river itself, it would be extremely difficult to divert the current of the river into the canal,† and to make it draw off the required quantity of water. Again, the current from the canal when it joined the river again, by striking the stream of the river obliquely, would retard its velocity, raise its level above the point of junction, and in some degree neutralise the advantage which had been gained.

Altogether, the flood would not be lowered more, perhaps, than a few inches at the point where it is most important that

That is to say, if the difference of level is caused by an obstruction over which the river falls. In the case of a stone bridge, the difference of level increases as the river rises. In this calculation I leave out of consideration the effect produced by the Ponte St. Angelo, in keeping up a head of water, because such an act of vandalism as the removal of that bridge is not contemplated.

It seems to be assumed that the Tiber would rush into the canal, as if a flood-gate were opened in a reservoir. But there is no reason why, the levels being the same, the water should elect to enter the canal, rather than to continue its course in the main channel. Without a barrage, to which there would be objections which it would occupy too much space to discuss, it would be impossible, I believe, to lower the level of the river one foot.

I

it should be lowered; in the neighbourhood, namely, of the Ripetta and the Corso.

But not to speak of their inefficiency, canals through which the water was to discharge itself only in time of floods, whether they entered the river at a lower level, or were continued to the sea, would soon become an unmitigated nuisance. To say nothing of the first cost, a considerable expense would be incurred in maintaining these new channels, and keeping them open; otherwise, as the waters sank, they would become full of sandbanks, between which pools of stagnant water would collect; a rank growth of weeds would choke their beds; unwholesome vapours would be evolved from them during the heats of summer; and musquitoes would find in them the breeding places which they most affect. The comparative immunity of Rome from musquitoes is doubtless owing to the absence of stagnant water. Every one, who is acquainted with the natural history of that delectable insect, is aware that it passes the first stage of its existence under water, and that, when it is about to undergo its transformation, it rises to the surface. But, unless the water is perfectly tranquil, the winged insect, as it emerges from its larva case, is overset and drowned. The stagnant pools in these canals would, therefore, supply what the musquitoes would doubtless consider a great desideratum to their race.

When a fire breaks out in a theatre, the tank that should extinguish the fire is usually found to be empty, and the pipes that should distribute the water to be out of order. So it would be with the canal. When a great flood came, it would be found that the banks had fallen in, that the bed was choked, and that the canal was incapable of performing its office.

PROPOSED EMBANKMENT OF THE TIBER.

Among the many schemes which have been proposed to the present government of Italy, in consequence of the late inunundation, is that of embanking the Tiber. I shall, therefore, consider, first, the subject of embankments in general, and,

second, the expediency of embanking the Tiber under the actual conditions of that river.

The Tiber does not appear to have been embanked by the Romans. The idea did not occur to the commissioners appointed by Tiberius to consider the means of moderating its inundations, though far more feasible than the plan proposed by them of diverting its affluents; a plan which would be considered impracticable by engineers of the present day. passage "Tiberis ripas extruxi," in the letter of Aurelian already quoted, has been translated by some, "I have raised embankments along the Tiber." It means, I think, merely: "I have built up the banks of the Tiber," that is, I have repaired the banks, faced them with stone, or supported them by piles. The whole tenor of the letter shows that the object of Aurelian was not to prevent inundations, but to remove obstructions to the navigation of the river, in order to facilitate the supply of grain, the great object of anxiety with the Roman Emperors. There are no traces of such a work as is supposed, though portions of it must necessarily have survived the lapse of ages.

In mountainous regions torrents are confined by mounds of earth and stone to prevent them from tearing up the ground, and converting the adjoining fields into a sandy and stony desert; but rivers in the plains were originally embanked, not for the purpose of preventing their occasional inundations, which are by no means an unmixed evil,* but to reclaim land which was under water the whole or greater part of the year; for in those cases the declivity of the rivers was so slight that it was impossible to drain the country by lowering the level of their beds. High mounds, therefore, were erected along their course, sufficient to confine the river during the greatest floods,

* Torricelli used to say that "les limons," or muddy deposits of rivers, were more precious than sands of gold.

During an inundation of the Loire in 1846 the plains of Forez above Roanne, where the river had liberty to expand itself, were overflowed, and many houses destroyed, yet the engineer in chief, Boulangé, declared that far more good had been done by the fertilizing deposits left by the flood than harm by the destruction of the houses.