[10] This, however, is not an absolute test of the fineness of the _water-lines_ of a vessel, and it can only be used as such on the assumption that the midship sections of ships are of similar form. The best test of the fineness of water-lines is made by taking the displacement as a percentage of the prism whose length is that of the ship and whose section is the same as the midship section of a ship, assuming, however, that the midship section of all ships is approximately that found in general practice to-day; in speaking of coefficients it will mean the percentage of the rectangular block above named. [11] More than thirty years ago this matter had been observed by the officers of the British navy, and experiments were ordered to be tried with H. B. M. S. Flying Fish, a 1,100-ton cruiser, her length being 200 feet, breadth 30 feet 4 inches, and her draft of water 10 feet 6 inches forward and 13 feet aft. With 1,290 I. H.-P. her speed was only 11.64 knots, whereas with 577 I. H.-P. it was 9.923 knots, and a speed of 11.201 was obtained with but 878 I. H.-P. A false bow 18 feet long was then fitted, so as to give finer lines forward, or, as sailors describe it, “a better entrance,” when it was found that with 1,285 I. H.-P. a speed of 12-1/2 knots was attained, and with 1,345 very nearly 12-3/4 knots. There is also every reason to suppose that could the stern have been altered in a similar way, the speed would have been still higher, in spite of the ship being larger and with a consequent increase of immersed surface to cause resistance. It has, besides, been observed on many occasions that when steamers have been cut in two and lengthened there has been no diminution of the speed, but, on the contrary, in some cases there has actually been a gain; so that in these two instances there is an apparent anomaly, viz., that with the same power the larger ship is propelled at a quicker speed.