In his Popular Lectures and Addresses, vol. ii, 1894, p. 360, Lord Kelvin published a paper which he had read before the Institution of Engineers in Scotland in 1867, on “The Rate of a Clock or Chronometer as influenced by the Mode of Suspension.” The effect has long been known. Almost two centuries ago, Ellicott observed the marked beats exchanged by two similar regulators when their cases were connected, which is due to the same effect (Phil. Trans., vol. xli, 1744, p. 126). The fact has crystallised into the rule that the cock that holds the pendulum ought to be as firmly fixed as possible; but as the question is not familiar, and as the observations of Lord Kelvin are given in general terms, and his description—though remarkably correct—contains some oversights, I thought it worth while to repeat the theory and the experiment.