PENDULUM. Under IlomoLocr will be found some details respecting clock and compensated pendulums, together with figures of the gridiron and mercurial pendulums. We here insert a few miscel kincous details on the same subject The way in which motion is communicated to a pendulum is there described. It may be added that the clasp of the pendulum by the crutch should neither be very close nor very loose, and the axis of motion of the pendulum should be exactly in the continuation of the arbor of the crutch. The clock-frame itself must be firmly fixed, for any shake or looseness of the support of the pendulum will alter its time of oscillation : the bob should be heavy, the arc of vibration small ; and when good performance is required, the rod should be of deal, well varnished or gilt. Iron-wire makes the next best rod, and brass the worst. Whenever a brass rod is used for an ordinary clock pendulum, the maker is not master of his craft, or he thinks little of his customer's knowledge. The errors arising from changes of tempe rature when deal, steel-wire, and brass are used, are as I, 3, and 5. It must however be remembered, that unless the deal rod pendulum be coated over, so as to be impervious to moisture, it will be liable to small errors arising from that cause. A range of 50° will not alter the rate of a clock with a white deal pendulum more than 5' a day, if Capt. Kater's value of the expansion of deal be correct.
In a well-made clock, the error arising from expansion from tempe rature is the most considerable, and is that which must be guarded against. Mr. Baily's contrivance with a leaden weight is given under Hosotoov. We may add that a lenticular form may be given to the weight, provided the proper length be preserved, and this will be an improvement, as it diminishes the resistance of the air, and the error which arises from the inertia of the air carried by the pendulum.
It is stated in the article just referred to, that in Harrison's pen dulum foreign artists use zinc rods instead of brass. Zinc is objection able as being a weak metal, and it is said by some persons to expand by jumps, which seems not very probable unless much exposed. The
great objection to the gridiron pendulum is that the astronomer, if he be no workman, must rely upon the artist for perfect compensation (and this is perhaps never achieved); and again, if from rust or bad fitting the slipping parts should bind, the action will necessarily be by bounds and irregular. After the clock-maker has done his best, a year's experience will point out the error in the compensation, which can easily be remedied, if the fault be over-compensation, by cutting tbe zinc rods shorter. All the fixed parts should then be secured as firmly as possible by steady pins, as any attempt to reserve a power of further adjustment would be too dearly purchased by a ricketty frame. The bands attached to rods 1 and 5 for the purpose of keeping the zinc rods in their places should not squeeze the zinc, and but just press against them with the spring of the metal. Harrison is said, by Short (` Phil. Trans.,' vol. 47, p. 517), to have invented his pendulum about 1725.
The construction of Graham's pendulum allows the astronomer to alter and finally perfect the compensation. This quality, and the absence of any nice fitting or slipping parts, give Graham's construc tion a very decided superiority over Harrison's, especially for a clock which is not to be moved from place to place. These mercurial pen dulums are generally constructed in a more expensive manner than is necessary. The vertical rods may all be made of stout steel wire, and the base and top of the stirrup of brass. Mr. Bally, in his valuable memoir (` Astron. Soc.,' vol. i.), justly recommends that the thread of the regulating screw should be deeper and coarser than it is usually made, and be tapped on a stouter cylinder. Graham's pendulum is described by himself. (` Phil. Trans.,' vol. xxxiv. p. 40.) It was in vented in 1722. There are many ways of compensating a pendulum for the effects of temperature, some of which 'are very ingenious, and others very simple. Those described under FIonoLoor will suffice for an explanation of the principle ; they are the most usual, the most easy of execution, and most exact in their operation.