It also seemed important, in designing that Tower, with reference to the rise of tide, to give its lower part a sloping form, as the least likely to obstruct the free passage of the waves. The outer stones of the lower courses were also selected of unusual length _inwards_, so as to bring them more under the influence of the vertical pressure of the upper wall. Before leaving this subject, I may remark, that it is quite possible to construct a Tower of a curved form, in such a manner, that the pressure of the upper part of the pillar shall be distributed to the greatest advantage on every stone, by building the outer walls as inverted arches, so that the section of each stone shall be that of a voussoir, with joints perpendicular to the successive tangents of the curve. This arrangement of the stones is, in fact, practised in sea walls of various kinds, and has even been recommended for circular Towers in an ingenious paper in the Transactions of the Royal Scottish Society of Arts. But in many situations, and at Skerryvore in particular, this mode of transmitting the pressure, so as to throw it perpendicular to the beds of the stones, is inadmissible, as conducing to or involving a greater evil. The evil has already been noticed, and consists in the thrust of the lowest stone (which is of course inclined to the horizon) having a tendency to push out the sides of the Rock on which the Tower is built. This fear, where the Towers are to be placed on small steep rocks or pinnacles, and more especially when these Rocks are traversed by veins nearly vertical, is by no means visionary; and there is good reason to apprehend, that the pressure thus resulting in a line considerably inclined to the plane of _cleavage_, might throw outwards a thin portion of rock, which, under the more conservative influence of a vertical pressure, might continue to retain its connection with the rest of the Rock unimpaired for ages. Another method of, in some degree, increasing the resistance of a Sea Tower to a horizontal thrust, if such aid be required, is to give the line of courses a continuous spiral form, instead of building them in successive horizontal layers. Were there reason to fear that the entire dislocation of the building might take place in a plane nearly horizontal, this method seems more calculated to counteract the danger than the use of dowels or joggles passing from the course below to the course above; but, as this is one of the accidents least to be apprehended, there does not seem any good ground for resorting to a mode of structure which would lead to considerable intricacy of workmanship, and would, in practice, be attended with difficulty in obtaining a proper vertical bond or union among the several stones.