_February 2, 1853._--Several drafts of ships have been made and much consideration given to the subject, and frequent discussions with various parties. The result of all is that my present views are as follows:-- The ship, all iron, double bottom, and sides up to water line, with ribs longitudinal like the Britannia tube. I have not been able to devise any good mode of determining the relative amount of friction of a copper and an iron surface; and, although I believe in copper, it would not do to act on mere belief. I therefore at present settle iron, the surface being carefully made smooth. Doubts have come across me also as to whether with a very long surface the difference between the smoothness will so much affect the total resistance. Is not a film of water, after a certain distance, carried with the body? and, if so, its greater or less roughness, if not producing currents, is almost unimportant. Would there be any difference in the resistance of a fine file or a rough one drawn through tallow, if they both covered themselves with grease? Is there any similarity? As to size, if we are to go round the world,[148] I do not think we can do with less than--length, 730; beam, 85; draught deep, 34; and I assume a nominal horse-power of engines equal to 1¼ of the sectional area at 30 feet; but, taking consumption as a better measure, and assuming that every possible economy is practised, and every refinement introduced that can produce economy, I shall assume 7½ lbs. per hour per nominal horse-power, or say 0·08 ton per day per horse-power; and as I assume the horse-power to be 1¼ sectional area, it makes the consumption =0·1 ton per day per foot of sectional area. And this is a very large allowance and ought to ensure a very high speed. In order to effect the utmost economy, I should work up to 20 lbs. steam (calling it 16 lbs.), cutting off certainly at ⅓ the stroke, and adopting every precaution to keep the steam hot and the condenser cool. The latter depends, I believe, solely upon the perfect dispersion of the injection water, so that the condensation of the steam may take place suddenly, otherwise the same amount of water may condense the steam _in time_, the same amount of heat be given off, the same quantity of injection water used, and yet the condenser be always full of steam at a good pressure. It might be well worth the experiment to try the effect of a large injection at the moment of the exhaust port being opened; but above all things I believe the heating of the steam to be important; and for this purpose I should jacket the steam pipes and cylinders top and bottom, and heat with high pressure steam, say at 60 lbs.--I have increased this pressure the more I think of it; 60 lbs. would be