It was during his residence at Oxford that he invented the air-pump, which was afterwards improved for him by Hooke, and with which he conducted most of those experiments on the "spring" and weight of the air, which led up to the investigations that have rendered his name inseparably connected with "the gaseous laws." The experiments of Galileo and of Torricelli had shown that the pressure of the air was capable of supporting a column of water about thirty-four feet in height, or a column of mercury nearly thirty inches high. The younger Pascal, at the request of Torricelli, had carried a barometer to the summit of the Puy de Dome, and demonstrated that the height of the column of mercury supported by the air diminishes as the altitude is increased. Otto von Guericke had constructed the Magdeburg hemispheres, and shown that, when exhausted, they could not be separated by sixteen horses, eight pulling one way and eight the other. He was aware that the same traction could have been produced by eight horses if one of the hemispheres had been attached to a fixed obstacle; but, with the instincts of a popular lecturer, he considered that the spectacle would thus be rendered less striking, and it was prepared for the king's entertainment. Boyle wished for an air-pump with an aperture in the receiver sufficiently large for the introduction of various objects, and an arrangement for exhausting it without filling the receiver with water or otherwise interfering with the objects placed therein. His apparatus consisted of a large glass globe capable of containing about three gallons or thereabouts, terminating in an open tube below, and with an aperture of about four inches diameter at the top. Around this aperture was cemented a turned brass ring, the inner surface being conical, and into this conical seat was fitted a brass plate with a thick rim, but drilled with a small hole in the centre. To this hole, which was also conical, was fitted a brass stopper, which could be turned round when the receiver was exhausted. By attaching a string to this stopper, which was so long as to enter the receiver to the depth of two or three inches, and turning the stopper in its seat, the string could be wound up, and thus objects could be moved within the receiver. The tube at the bottom of the receiver communicated with a stop-cock, and this with the upper end of the pumpbarrel, which was inverted, so that this stop-cock, which was at the top of the barrel, took the place of the foot-valve. The piston was solid, made of wood, and surrounded with sole leather, which was kept well greased. There being no valve in the piston, it was necessary to place an exhaust-valve in the upper end of the cylinder. This consisted of a small brass plug closing a conical hole so that it could be removed at pleasure. The construction of the cylinder was, therefore, similar to that of an ordinary force-pump, except that the valves had to be moved by hand (as in the early forms