[6] The principle which underlies the experiment is this:—If any one body forming part of a system of heavy bodies be moved from one position in the system to another, the weight of the body moved multiplied into the distance through which it is moved, is precisely equal to the weight of the whole system of bodies multiplied into the distance through which the common centre of gravity of the whole has moved. If in a ship, therefore, a movable weight of known amount is moved across the deck through a given known distance, the centre of gravity of the ship itself, with all on board, has been moved in a line parallel to that through which the small weight has been transferred, and through a distance inversely proportioned to the weight of the whole ship to the weight moved. If, for instance, a weight of five tons should be moved through a distance of twenty feet, then multiplying this weight into this distance and dividing by the total weight of the ship, the distance through which the ship’s centre of gravity has travelled parallel to the deck is obtained. If, at the same time, an exact measure of the angle through which the ship has been inclined by moving the five tons through the distance named has been taken, and the position of the ship’s metacentre has been obtained, then the elements of a triangle are known—namely, the degrees in each of its angles, and the length of one of the sides—and from these the length of the remaining sides of the triangle is easily deduced. One of these sides will be the distance between the metacentre of the ship and its centre of gravity, and, consequently, the metacentre being known from calculation, the position of the centre of gravity becomes known also. [7] The classification of strains here given is as contained in White’s “Manual of Naval Architecture.” To this authoritative source readers must turn who wish a full exposition of the several problems go shortly dealt with in these pages. [8] This will be more fully referred to further on, but it may be stated here that the need for independent calculation is largely obviated, owing to the existence of “co-efficients,” deduced from investigations made by experts. Further, the existence and influence of the Registration Societies are such that the codes of scantling and the structural supervision instituted by them together constitute the only guarantee of structural strength generally desiderated.