On Monday morning Watt proceeded to make an experimental test of his new invention, using for his steam-cylinder and piston a large brass surgeon's-syringe, 1-3/4-inch diameter and 10 inches long. At each end was a pipe leading steam from the boiler, and fitted with a cock to act as a steam-valve. A pipe led also from the top of the cylinder to the condenser, the syringe being inverted and the piston-rod hanging downward for convenience. The condenser was made of two pipes of thin tin plate, 10 or 12 inches long, and about one-sixth of an inch in diameter, standing vertically, and having a connection at the top with a horizontal pipe of larger size, and fitted with a "snifting-valve." Another vertical pipe, about an inch in diameter, was connected to the condenser, and was fitted with a piston, with a view to using it as an "air-pump." The whole was set in a cistern of cold water. The piston-rod of the little steam-cylinder was drilled from end to end to permit the water to be removed from the cylinder. This little model (Fig. 25) worked very satisfactorily, and the perfection of the vacuum was such that the machine lifted a weight of 18 pounds hung upon the piston-rod, as in the sketch. A larger model was immediately afterward constructed, and the result of its test confirmed fully the anticipations which had been awakened by the first experiment. [Illustration: FIG. 25.--Watt's Experiment.] Having taken this first step and made such a radical improvement, the success of this invention was no sooner determined than others followed in rapid succession, as consequences of the exigencies arising from the first change in the old Newcomen engine. But in the working out of the forms and proportions of the details of the new engine, even Watt's powerful mind, stored as it was with happily-combined scientific and practical information, was occupied for years. In attaching the separate condenser, he first attempted surface-condensation; but this not succeeding well, he substituted the jet. Some provision became at once necessary for preventing the filling of the condenser with water.