For this reason therefore we approve highly of the practice, which we believe is very general among arti ficers, we mean that of backing up the arch with solid masonry, lor several courses above the springing. For granting that the friction on the horizontal bed be fully equal to the thrust, yet as the tendency to slide off is greatest there, it is well that it should be effectually re sisted. This is readily done by the solid backing, which increases the mass of friction; and in the ease of a bridge of several arenes, enables us to set the thrusts of adjoining arches in opposition to each other. The materials, therefore, in that part, ought to be laid close up to the spring courses, and also bonded into the inferior part of the abutment or pier, which will act as a sort of dowelling, and does not preclude the employ ment of that means also. If great security is thought necessary, cement, being a compressible substance, ought to be sparingly employed in the vertical joints at the back of the archstones.
The friction of the sections of the arch, as it permits a considerable variation to take place in the position of the joints, will also admit a considerable deviation from the load, which is necessary for equilibrium over any point of the curve.
It would not be difficult to investigate the extent to which this variation of weight might be cat ried. But we shall at present only remind the reader, that as we find a variation of 20° practicable in the position of the joints, he may conclude, that each section will admit of its load being altered to that which would suit a point in the curve 20' on either side of it.
But in speaking of this alteration of weight, it must be observed, that we consider it only so far as it is like ly to cause the sections of the arch to slide on each other. lithe overload be considerable, or if the arch be thin, and of course flexible, a deviation from equilibrium may produce an effect equally destructive with that which would arise by the sliding of the sections. Sup pose this deviation to consist in overloading the crown; a case very likely to occur, for even the. variable pres sure of a loaded waggon bears, in some cases, a very sensible proportion to the weight at the vertex of the arch, this overload will be equivalent to a greater thick ness of crown, and of course will sensibly increase the horizontal thrust, while the total weight of the arch, or vertical pressure, is much the same as before. This is a still stronger reason for backing up the spring courses with masonry; for unless the stones, where their joints become nearly horizontal, be sufficiently steady, they may, by the increasing thrust, be pushed along their beds ; for they are already near the limit of steadiness from friction : a very short slide will open the joints to wards the crown. The same thing will also be produced by the compression of the cement in the arch. The
vertical sections will descend. They already tend strong ly to do so. The separation of the sections on each side of the vertex is equivalent to drawing their joints to a lower point than before. This will again increase the horizontal thrust.
We have seen, that these joints are already too low for equilibration in a common circular arch. The motion of the lower sections, therefore, if once begun, will go on increasing, until the arch falls to pieces. But what is perhaps of more importance, as the joints towards the crown will now open below, and the sections, not being in entire contact, will hang by their upper corners only, these may chip and crumble away, thereby form ing them into more acute wedges, and giving us a new cause of destruction. For though the cohesion of the Inatux of the archstone may effectually resist the tan gential pressure, when distributed over a Joint of con siderable superficies ; yet when the whole of that pres sure is condensed into a small compass, ur a mere point, and that near the edge, and acting perhaps in a very unia%ourable direction, since the friction permits it to act t u great obliquity, its uestructivc tendency may be irresistible.
Suppose a motion of this kind actually going on in an area, as is generally the case when the centre or scaffolding is taken from oelow it, How is it to be pre vented ? We answer, not easily : For though the motion be exceeding slow, or almost imperceptible, yet the quantity of matter is so enormous, that its momentum is great. Nothing, therefore, but the most solid work could resist it. Accordingly, in striking the centres of an arch, the whole is not taken away at once. It is not likely that any arch could withstand that treatment ; but the centre is gradually let down, stopping now and then until the work settles. It does so first at the springing and haunches, and the crown of the arch is the last part that departs From the centre. It appears, therefore, that e \ cry arch is a segment of a greater circle, after it is finished, than before. Allowance should be made for this in the design, and in calculating the weights necessary for equilibrium from the horizontal thrust. Au attempt is sometimes made to remedy this change of figure, by driving down the key stones. And though it is far better to ender such expedient unnecessary by careful workmansnip, yet this method is not to be de spised. It seems to have been the common practice of the ancient architects. An overdriven key-stone, or console, as it is termed, is one of the most usual orna ments of the archivolt. Nevertheless, even when this expedient is thought necessary, it should be employed with great caution. Wedges of small taper have great power. And the horizontal thrust should be in no cast increased, without very weighty reasons.