"Upon the 9th of August, 1 marked out the entry and staircase ; and having unloaded the Eddystone boat, which was loaded with the first pieces of Course XV., we imme diately proceeded with it ; and from this time were blessed with such an uninterrupted continuance of line weather, that upon the 20th of August, Course XVIII. was completed, which reunites the building into a complete circle, by cover ing the passage to the staircase : the external face of the stone of that course, which makes the cover or head of the entry-door having the figures 1758, denoting the year in which this part of the work was accomplished, cut in deep characters upon it.
On the 24th of' August, the fine weather. and in conse quence the works, were interrupted, Course XX. being then in hand ; and it was not till the 24th of September, that, with every possible exertion, Course XXIV. was finished, which completed the solid, and composed the floor of the store-room.
"The 25th and 26th of September, Course XXV., being the first course of the superstructure, was successfully com pleted in its place ; but, as the mode of construction now became entirely different from the former, it is necessary to give an account thereof, as also of the reasons for the change. The building was carried up solid, as high as there was any reason to suppose it exposed to the heavy stroke of the sea ; that is, to thirty-five feet four inches above its base, and twenty-seven feet above the top of the rock, or common spring-tide high-water mark. At this height, as it was reduced to sixteen feet eight inches in diameter, it became necessary to make the best use of this space, and make all the room and convenience therein that was possible, con sistent with the still necessary strength. The rooms being made of twelve feet four inches diameter, this would leave twenty-six inches for the thickness of the walls. These being made with single blocks in the thickness so that sixteen pieces might compose the circle, would, from its figure, compose a stout wall ; yet moor-stone, as has been observed, being a tender kind of stone, in respect to the union of its component parts, any method of dovetailing the blocks together, at this thickness, appeared to me. imprac ticable to any good purpose. What seemed to be the most effectual method of bonding the work together, was that of cramping with iron, which would confine each single piece to its neighbouring piece in the same circle : and if to this be added, that every piece should, at each end of it, lay hold of an inlaid piece, or joggle in the same nature as the cubes, then not only all the pieces in the same course would be united to each other by the cramps, but steadied from moving upon the under course by the joggles, and of conse quence would be fastened at thirty-two points : for in each course there being sixteen joggle-stones, as each end of each principal piece, at its base, took hold of half a joggle, there would be thirty-two points of confinement in the circle above ; that is, the joggles being made to occupy the middle of the upper bed of each block, in that situation they would cross the joints of the course above. These joggles, as well
as the rest, were of sawn marble, and made eight inches long, four inches broad, and three inches thick : each end of each block, therefore, would occupy four inches in length, four in breadth, and one inch and a half in the height of each joggle; and this I judged quite sufficient to keep every course in its place, at the height that this kind of work was begun, and so as to constitute a piece of solid masonry. There was, however, another matter, that it seemed quite material also to attend to ; and that was, to render the habit able. rooms contained within those shells of walls, perfectly dry and comfortable in all weathers ; and this seemed to merit very particular attention ; for the seas that are said to rise up against, and in a manner to bury the house, in time of storm, would make effectual trial of every joint.
"The level joints being pressed together by the incumbent weight of the building, would keep firm and sound that cohesion of parts pro'dueed by the mortar; so that once being made water-tight. there was no doubt that they would so remain : but with respect to the upright joints, the least degree of shrinking, either of the stone or of the mortar between, tended to open the joint, so that it might always remain leaky, in a greater or a less degree ; for we know Of no degree of separation of parts, however minute, short of absolute contact, which will stop or prevent the percola tion of water. For this purpose I conceived that if flat stones were introduced into each upright joint, so as to be lodged partly in one stone, and partly in its neighbour, (much upon the same idea that Dutch laths were formerly introduced into the joints of chamber floors, to !tinder the passage of wet,) the water might be prevented from making its way through the upright joints of the walls.