Despite the fact that Newcomb in 1895 used the mean solar second between 1750 and 1892 as an independent unit of time, he had already in 1876 suspected that the Earth was a poor time-keeper.