To make my first imperfect press I had bought a piece of wood from a wheelwright in order to have it turned into two cylinders. Hardly had the two been in the house a day before each one split so that a longitudinal crack of two inches width appeared. As the upper cylinder was thick enough to make an impression of a whole folio of sheet music without revolving so far as to let the crack reach the stone, I contented myself with it temporarily. Now, in order not to spoil the impressions, I had to begin each revolution of the cylinder at the crack, for otherwise the crack might have come at the middle or end of the impression and given no imprint of that part. Therefore, as the stone was pushed under the cylinder at the crack, it was already gripped before the impression began, and was drawn through at once. With the new press, however, the upper cylinder had to draw the stone between both cylinders in order to bring it under its pressure. But in doing this, the new press first pulled the linen stretched over the printing-frame till it would yield no more and forced the stone powerfully under the cylinder, during which of course the paper under the linen was pulled over the inked stone and smeared. Several attempts to rectify this trouble were unsuccessful. Probably I would have discovered the remedy finally,--either that the upper cylinder must not first be pressed on the stone, which must be under it before each impression began, or that I need only use tightly stretched leather instead of linen. But I decided, instead, in order to complete our work if possible, to have a press made in all haste by a carpenter, of a style like the book-printers' press, wherein the force is applied instantly from above. As everything was very rough, the new press was ready in eight days. The first experiment, with a small stone, seemed to succeed. But the larger stones would not give thorough impressions, probably because of the uneven surface of the press, which was merely of wood. I increased the power enormously. A stone of three hundredweight was elevated with pulleys and released suddenly to fall ten feet. It forced a lever down on the press with a pressure of more than ten thousand pounds. The plates gave fair impressions by this means, but generally they were cracked after the first, second, or third impression. To determine how much downward force was needed to print a sheet of music, I took a well-ground stone a square inch in area, laid moistened paper on an inked printing-stone, over this a sixfold layer of paper, then a double layer of fine cloth, finally the square inch of stone, and then weights ranging from one to three hundredweight.