Compensation Pendulum.—As the length of a rod or bar of any material depends on its temperature (see EAT), a clock with an ordinary pendulum goes faster in cold, and slower in hut, weather. Various contrivances have been devised for the purpose of diminishing, if not destroying, these effects. The most perfect in theory, though perhaps not the most available in practice, is that of sir D. Brewster (q.v.), founded upon the tal discovery of Nlitscherlich, that some crystals apand by heat in one direction, while contracting in the perpendicular_ one; and therefore that a rod may be cut out of the crystal in such a direction as not to alter in length by any change of temperature. In the method of correction usually employed, and advantage is taken of the fact that different substances have different coefficients of linear dilatation; so that if the bob of the pendulum be so suspended as to be raised by the expansion of one substance, and depressed by the expansion of another, the lengths of the effective portions of these substances may be so adjusted that the raising and depression, taking place simultaneously, may leave the position of the hob unaffected. There are two common methods of effecting this, differing a little in construction, but ultimately depending on the same principle. Of these, the mercurial dulum is the more easily described. The rod AC, and the framework CI3, are of steel. Inside the framework is placed a cylindrical glass jar, nearly full of mercury, which can be raised or depressed by turning a nut at B. By increase of temperature, the steel portion AB is lengthened by an amount proportioned to its length, its coefficient of linear dilatation, and the charge of temperature, conjointly—and thus the jar of mercury is removed from the axis of.suspension. But the expansion of the glass, which is very small, the mercury rises in the jar by an amount proportional to its bulk, its coefficient of cublical dilatation, and the change of temperature conjointly. Now, by increasing or diminishing the quantity of mercury, it is obvious that we may so adjust the instrument that the length (-) of the equivalent simple pendulum shall be unaltered by the change of temperature, whatever be its amount, so long as it is not great enough to sensibly change the coefficients of dilatation of the two metals. The screw at B has nothing to do with the compensation, its use is to Idjust the length of the pendulum so that it shall vibrate in one second.
The construction of the gridiron, pendulum will be easily understood from the cut. The black bars are steel, the shaded ones are brass, copper, or some substance whose co-efficient of linear dilatation is more than double that of steel. It is obvious from the figure that the horizontal bars are merely connectors, and that their expansion has nothing to do with the vibratiou of the pendulum, so they may be made of any substance. It is easily seen that an increase of temperature lowers the bob by expanding the steel rods, whose effective length consists of the sum of the lengths of Aa, BC, the steel bar to which the bob is attached; while it raises the bob by expanding the brass bars, whose effective length is that of one of them only; the other, as well as the steel rod be, being added to the instrument for the sake of symmetry, strength, and stiffness only. if the effective lengths of steel and brass be inversely as their respective dilatation co-efficients, the position of the bob is unaltered by temperature; and there fore the pendulum will vibrate in the same period as before heating. This is on the supposition that the weight of the frame-wdrk may be neglected in comparison with that of the bob; if this weight must be taken into account, the requisite adjustments, though possible, are greatly more complex, and can only be alluded to here. Practically, it is found that a strip of dry fir-wood, carefully varnished to prevent the absorption of moisture, and consequent hygrometric alterations of its length, is very little affected by change of tem perature; and, in unary excellent clocks, this is used as a very effective substitute for the more elaborate forms just described. To give an idea of the nicety which modern astronomy requires in the construction of an observ ing-clock, we may mention that the Russian astronomers find the gridiron superior to the mercurial pendulum; because 'differences of temperature at different parts of the clock case (though almost imperceptible in a properly protected instrument), may heat the steel or the mercury unduly in the latter: while, in the former, the steel and brass bars run side by side through the greater part of the length of the pendulum, and are thus simultaneously affected by any such alterations of temperature.
It would lead us into details of a character far too abstruse' for the present work to treat of the effects of the hydrostatic pressure and viscosity of the air upon the motion of a pendulum.