The model was accordingly dragged through the tank at a speed of nearly 5 feet per second, and, after deducting from the total observed pull the resistance due to the calculated skin friction of the model, it was found that the resistance to the motion of the model at this speed due to wave-making was 1·08 lb. Hence, by Froude’s second law, the wave-making resistance of the ship was predetermined to be— 1·08 × (240/12)^3 × 40/39 = 8850 lbs. The last fraction 40/39 is a correcting factor in passing from fresh water to salt water. The surface of the proposed ship was 10,280 square feet, and the skin friction was known to be 1·01 lb. per square foot at a speed of 13·5 knots. Hence the total skin resistance of the ship would be— 10,280 × 1·01 × 40/39 = 10,620 lbs. Adding to this, the 8850 lbs. for wave-making resistance, we have a total resistance of 19,470 lbs. predetermined as the total pressure required to be overcome in moving the ship at a speed of 13·5 knots. Hence, since 1 horse-power is defined to be a power which overcomes a resistance of 33,000 lbs. moved 1 foot per minute, it is easy to see that 19,470 lbs. overcome at a rate of 13·5 knots represents a power of— (19,470 × 13·5 × 6080)/(33,000 × 60) = 810 horse-power But now, in the case of a screw-driven steamer, a part of the power is lost in merely churning up the water, and a part in internal frictional losses in the engine and screw-shaft. It is not far from the truth to say that 50 per cent. of the applied engine-power is lost in useless water-churning. Hence, for the above steamer, an actual power of at least 1600 H.P. would have to be applied to the screw-shaft. To allow, however, for the loss of power in friction, and to allow a margin for emergencies, it would be usual to provide for such a steamer engines of at least 3000 _indicated_ horse-power. Each shipbuilder has, however, at call a mass of data which enable him, from actual measured mile trials, to determine the rates between the calculated driving horse-power and the indicated horse-power of the engines, and so enable him, in the light of experience, to provide in any new ship the exact amount of steam-power necessary to produce the required speed. As an instance of how accurately this can be done by the aid of the tank experiments, Mr. A. Denny gives an example drawn from experience in building the well-known paddle-steamers _Princess Josephine_ and _Princess Henriette_ for the Belgian Government Dover to Ostend fast mail-steamer service. The speed guaranteed before the boats were built was 20¹⁄₂ knots. The estimate was made for 21 knots, and the actual results of trials on the measured mile, when the ships were built, showed that each did 21·1 knots on prolonged and severe test. The reader, therefore, cannot fail to see how important are these methods, laws, and researches of Mr. Froude.