29. The regularity of the supply of fuel to the fire under the boilers of steam-engines is another mode of contributing to the uniformity of their rate, and also economizes the consumption of coal. Several patents have been taken out for methods of regulating this supply: the general principle being to make the engine supply the fire with small quantities of fuel at regular intervals by means of a hopper, and to make it diminish this supply when the engine works too quickly. One of the incidental advantages of this plan is, that by throwing on a very small quantity of coal at a time, the smoke is almost entirely consumed. The dampers of ashpits and chimneys are also, in some cases, connected with machines in order to regulate their speed. 30. Another contrivance for regulating the effect of machinery consists in a vane or fly, of little weight, but presenting a large surface. This revolves rapidly, and soon acquires a uniform rate, which it cannot greatly exceed, because any addition to its velocity produces a much greater addition to the resistance it meets with from the air. The interval between the strokes on the bell of a clock is regulated in this way, and the fly is so contrived, that the interval may be altered by presenting the arms of it more or less obliquely to the direction in which they move. This kind of fly, or vane, is generally used in the smaller kinds of mechanism, and, unlike the heavy fly, it is a destroyer instead of a preserver of force. It is the regulator used in musical boxes, and in almost all mechanical toys. 31. The action of a fly, or vane, suggests the principle of an instrument for measuring the altitude of mountains, which perhaps deserves a trial, since, if it succeed only tolerably, it will form a much more portable instrument than the barometer. It is well known that the barometer indicates the weight of a column of the atmosphere above it, whose base is equal to the bore of the tube. It is also known that the density of the air adjacent to the instrument will depend both on the weight of air above it, and on the heat of the air at that place. If, therefore, we can measure the density of the air, and its temperature, the height of a column of mercury which it would support in the barometer can be found by calculation. Now the thermometer gives information respecting the temperature of the air immediately; and its density might be ascertained by means of a watch and a small instrument, in which the number of turns made by a vane moved by a constant force, should be registered. The less dense the air in which the vane revolves, the greater will be the number of its revolutions in a given time: and tables could be formed from experiments in partially exhausted vessels, aided by calculation, from which, if the temperature of the air, and the number of revolutions of the vane are given, the corresponding height of the barometer might be found.(1*) NOTES: