In 1790 Mr William Nicholson devised a scheme for applying to ordinary type printing the already established process of printing calico by revolving cylinders. The impressions were to be taken from his press upon successive sheets of paper, as no means of producing continuous rolls had as yet been invented; but the machine worked far from satisfactorily, and practically came to nothing. A quarter of a century later Mr Edward Cowper applied Nicholson's idea to stereotype plates bent to a cylindrical surface. But till the advent of “Hill's machine” (described at the Patent Office as “A.D. 1835, No. 6762”) all plans for fixing movable types on a cylinder had failed. It is therefore incontestable that the first practical scheme of printing on a continuous roll of paper by revolving cylinders was invented and set to work by Rowland Hill in the year named. The machine was intended mainly for the rapid printing of newspapers, but the refusal of the Treasury to allow an arrangement by which the Government stamp could be affixed by an ingenious mechanical device as the scroll passed through the press—a refusal withdrawn later—deferred for many years the introduction of any rotatory printing machine. The apparatus was kept at my Uncle Matthew's chambers in Chancery Lane, and was often shown to members of the trade and others. Although driven by hand only, it threw off impressions at the rate of 7,000 or 8,000 an hour, a much higher speed than that hitherto attained by any other machine. But from 1836 onwards my father's attention was almost wholly taken up with his postal reform, and it was only after his retirement from the Post Office in 1864 that his mind reverted to the subject of the printing press. Several years before the latter date his brother had left London; but of the rotatory printing machine, bulky and ponderous as it was, a few small odds and ends—afterwards exhibited at the Caxton Exhibition in 1877—alone remained. In 1866 the once well-known “Walter Press” was first used in the _Times_ office. Of this machine my father has said that “except as regards the apparatus for cutting and distributing the printed sheets, and excepting further that the 'Walter Press' (entered at the Patent Office as “A.D. 1866, No. 3222”) is only adapted for printing from stereotype plates, while mine would not only print from stereotype plates, but, what is more difficult, from movable types also, the two machines are almost identical.” He added that “the enormous difficulty of bringing a complex machine into practical use—a difficulty familiar to every inventor—has been most successfully overcome by Messrs Calverley and Macdonald, the patentees.” By whom and through what agency the machine patented in 1835 was apparently transported from Chancery Lane to Printing House Square is a mystery which at this distant date is hardly likely to be made clear.