of the steam-engine). The piston was raised and depressed by means of a rack and pinion. The pumps could be used either for exhausting the receiver or for forcing air into it, according to the order in which the "valves" were opened. If the stop-cock communicating with the receiver were open while the piston was being drawn down, and the brass plug removed so as to open the exhaust-valve when the piston was being forced up, the receiver would gradually be exhausted. If the brass plug were removed during the descent of the piston, and the stop-cock opened during its ascent, air would be forced into the receiver. In the latter case it was necessary to take special precautions to prevent the brass plate at the top of the receiver being raised from its seat. All joints were made air-tight with "diachylon," and when, through the bursting of a glass bulb within it, the receiver became cracked, the crack was rendered air-tight by the same means. Other receivers of smaller capacity were also provided, on account of the greater readiness with which they could be exhausted. With this apparatus Boyle carried out a long series of experiments. He could reduce the pressure in the large receiver to somewhat less than that corresponding to an inch of mercury, or about a foot of water. Squeezing a bladder so as to expel nearly all the air, tying the neck, and then introducing it into the receiver, he found, on working the pump, that the bladder swelled so that at length it became completely distended. In order to account for this great expansibility, Boyle pictured the constitution of the air in the following way. He supposed the air to consist of separate particles, each resembling a spiral spring, which became tightly wound when exposed to great pressure, but which expanded so as to occupy a larger circle when the pressure was diminished. Each of these little spirals he supposed to rotate about a diameter so as to exclude every other body from the sphere in which it moved. Increasing the length of the diameter tenfold would increase the volume of one of these spheres, and therefore the volume of the gas, a thousandfold. Possibly this was only intended as a mental illustration, exhibiting a mechanism by which very great expansion might conceivably be produced, and scarcely pretending to be considered a _theory_ of the constitution of the air. Boyle's first idea seems to have been derived from a lock of wool in which the elasticity of each fibre caused the lock to expand after it had been compressed in the hand. In another passage he speaks of the air as consisting of a number of bodies capable of striking against a surface exposed to them. He demonstrated the weight of the air by placing a delicate balance within the receiver, suspending from one arm a bladder half filled with water, and balancing it with brass weights. On exhausting the air, the bladder preponderated, and, by repeating the experiment with additional weights on the other arm until a