But in spite of all prepossessions against it the screw had won a decisive victory over its rival. So striking were the results recorded by the _Archimedes_, that a decision was made in December, 1840, to change the _Great Britain_, an Atlantic liner then under construction, from paddlewheel to screw propulsion. In two ways she was a gigantic experiment: she was the first large ship to be built of iron, and it was now proposed to fit her with a screw. Mr. Brunel took all the responsibility for advising the adoption of both these revolutionary features; the result was a splendid testimony to his scientific judgment, boldness of enterprise, and “confident reliance on deductions from facts ascertained on a small scale.” Before the completion of the _Great Britain_ the Admiralty had initiated experiments which were to furnish important information as to the power and efficiency of the screw propeller in its various forms, and to settle beyond cavil the question of its superiority over the paddlewheel for the propulsion of warships. The sloop _Rattler_, 888 tons and 200 horsepower, was fitted with screw machinery. Several forms of screw were tried during the winter of 1843-4. First the screw as used in the _Archimedes_ was fitted: a screw of 9-foot diameter, 11-foot pitch, and of 5½ feet length, consisting of two half-convolutions of a blade upon its axis. Then a screw was tried of the same diameter and pitch but of only 4-foot length; and then the length was again reduced to 3 feet. The effect of cutting down the length was to give an increase of efficiency.[156] The screw was again shortened by 2 feet, and finally to 1 foot 3 inches; with each reduction in length the slip diminished and the propulsive efficiency increased. Various other forms of screws were tried, and it was shown that Pettit Smith’s short two-bladed propeller was on the whole the most efficient.