The advent of the fine-Iined clippers saw the elimination of this square bottom and the introduction of a rounded heel and upper portion, the contour being half elliptical with its greatest width of body carried much higher, as there was a cleaner run of water behind the ship with a minimum of eddies. The rounded bottom to the rudder was also a protection if the ship grounded or touched her heel in shallow waters, and the slim tapered top above the waterline reduced the area liable to be struck by following or quartering seas.

  The fullest width of rudder of the old pattern was approximately onethirtieth of the ship's length, when the length was about four times the beam. As the length increased to five, six, or occasionally seven times the beam, the rudder width was based on a proportion of about one-eighth of the beam, which was the general average for a rudder shaped in an arc, as in the first example of drawing (53). In some cases the width was as little as one-tenth or one-eleventh of the beam, these types having much less curvature to the profile so as not to reduce the total area too much.

  Sternposts on large ships had been made with a rake from earliest times for a number of reasons. It was easier to procure naturally grown shaped knees to connect the stempost to the keelson if it was an obtuse angle rather than aright angle, a reason which was also influential in wooden ships being built with pronounced tumblehome at the deck levels for the sake of less angular beam knees. It was also found from long experience that a ship sailed better with a raking post as against a plumb one. The practice of moving weights aft to give trim by the stern also in effect increased the rake of the post.

  The rudder was slightly more effective if raked, but required more power to operate even with the ship upright as it was hanging from the pintles by gravity, as against the upright post where the weight was directly borne in line with the pintles. Some designers also sought to gain slight advantage in the tonnage when the length of keel was used in the computation. By raking the post excessively they increased the deck length and retained the same tonnage as if the post had been vertical. A well balanced ship was one in which the helmsman had little to do in all conditions of sailing, the chosen course being maintained by the balancing of the pressure on all sails against the lateral resistance of the hull. The centre oflateral resistance lay in a line approximately mid-Iength of the underwater profile, and the centre of effort of the sails theoretically with all sails set, up to 1/2O length aft of the CLR, but it could be moved  forward or aft by varying the sails set at each extremity. The designer could also vary the centre of lateral resistance, which also included the rudder area, by rounding off the forefoot or altering the rake of the sternpost, thus reducing the deadwood area at each extremity as required. Too much deadwood such as in a sharp square forefoot resisted the turning action of a ship although it also helped to resist leeway.

  The ultimate success of the balance, however, always rested with the master, who would study performance on each voyage until he understood the finer adjustments, and as the weather conditions worsened he could reduce sail progressively and still retain the balance without undue use of the helm. It required good perception, knowledge and intuition for success, and on taking over anew command a hitherto successful master could find himself with a ship which reacted quite differently from his previous experience. The action of the rudder is such that a resistance is set up on one side of the ship which in turn forces the ship to turn. Too frequent use of the rudder thus reduces the speed of the ship also, although obviously it is necessary to counteract the effect ofheavy seas striking the bow or stern. This is all apart from the use of the rudder in inland waterways, rivers, docks, etc. where balance of sail does not help as the ship may be under tow or gliding to an anchor. The narrow rudder of the clipper ships was not very effective under these conditions and sometimes caused disasters such as those on the Min River from Foochow.