In the Eddystone the radius of the base, at the level of high water of spring tides, is somewhat less than _one-fifth_ of the height of the Tower above that level; while in the Bell Rock, at the same level, it is little more than _one-seventh_ of the height. If, again, we suppose the curve of the Eddystone to be continued downwards to the level of low water, the radius (in so far as we may judge from sketching the continuation of a curve undefined by any geometrical property) would be rather more than _one-fourth_ of the whole height above that level; while in the Bell Rock the proportion, in reference to the same level, is a little more than _one-fifth_. Viewing the whole height of the Skerryvore Tower above _high water_ of spring tides as equal to 142 feet, and finding that, in the cases of the Eddystone and the Bell Rock, the radius of the horizontal section at that level is respectively _one-fifth_ and _one-seventh_ of the whole height; and again, viewing the extreme height of the Skerryvore Tower above _low water_ of spring tides as equal to about 155 feet, and considering the proportionate radii of the Bell Rock and Eddystone (in so far as the latter is ascertainable) as respectively _one-fifth_ and _one-fourth_ of the heights of the top of the masonry above the level of low water, I finally decided upon giving the Tower at the Skerryvore such dimensions as would not be widely discordant with these general proportions. In this view, I determined that the radius of the base should not exceed 22 feet, on the level of about 4 feet above the high water mark, where I expected to obtain a solid foundation--a base which bears to the whole height of the Tower a proportion somewhat _less_ than that of the Bell Rock, which is _one-fifth_. It so happens, that the diameter adopted is nearly the greatest which the Rock affords; for, although a glance at the accompanying plan of the Rock at high water (Plate, No. III.) would lead one to suppose that a more extended base might have been obtained, I found, after many careful examinations of the gullies and fissures which intersect it, that some of the concealed fissures run much farther into the Rock than might at first be imagined. The adoption of a much larger base, even had it been otherwise advisable, would therefore have involved some risk of the external ring of stones of the lowest course giving way by the yielding of an unsound part of the outer portion of the Rock to the pressure of the superincumbent mass, and might eventually have led to the destruction of the Tower. [Illustration: No. 2.]