English: A tubular temperature compensated pendulum used in precision pendulum clocks, from a 1917 clock book. A major cause of inaccuracy in 18th century procision clocks was contraction and expansion of the metal pendulum rod with temperature changes. In 1726 John Harrison invented the gridiron pendulum, in which rods of two metals with different temperature coefficients, steel and zinc, compensated each other, so that the pendulum stayed the same length with temperature changes. However when expanding and contracting, the rods of the gridiron had to slide through holes in a metal frame in a series of tiny jerks, which caused the rate of the clock to change suddenly. Therefore clockmakers worked to invent a compensating pendulum using Harrison's principle which didn't have sliding friction.

This tubular design was invented several times by different clockmakers, and used in a number of precision clocks during the 19th century. It consisted of a steel rod (A), inside a zinc tube (B), inside a steel tube (C), with the tubes large enough that there was space between them.. The steel rod was suspended at top from the pendulum's suspension spring. It supported the concentric zinc tube at its bottom, and the zinc tube supported the steel tube at its top. The pendulum's weight (D) was attached to the bottom of the steel tube. An increase in temperature would cause all three parts to expand in length. However because of their interconnection the expanding steel rod and tube would make the pendulum longer, while the expanding zinc tube would push it up, making the pendulum shorter. Since zinc has a greater thermal expansion than steel, by making the parts of the proper length these two effects could be made to cancel, and the greater expansion of the zinc would exactly compensate for the expansion of the steel parts which had a greater combined length, so the pendulum would stay the same length with temperature changes. The advantage of this design over Harrison's gridiron is that the rod and tubes are self-supporting, so there is no friction between parts and thus the expansion and contractions could occur smoothly, without little jumps. The pendulum's bob (weight) is attached at its center to the steel tube, so expansion of the metal bob itself with temperature will not cause the center of gravity of the pendulum to change, changing the rate.