[ 223 ]

ness, so that the pressure varied in different points of itsevolution. On trying different kinds both of Europeanand Indian woods, all the hard ones split, the soft oneswarped without splitting, and, of more than twenty rolls,there was not one which continued round for twenty-four hours even without being worked in the machine.

These failures put him upon contriving another methodof pressing the rolls together, so that the force shouldalways accommodate itself to whatever inequalities mighthappen. The axis of the copper roll being made to turnbetween friction wheels as before, that of the wooden oneis pressed upwards by a lever at each end furnished witha half collar for receiving the end of the axis. Each leverhas the end of its short arm supported on the frame ofthe machine, and the long arm is drawn upwards by aniron rod communicating with the end of the short arm ofanother lever placed horizontally : to the long arm os thislast lever is hung a weight, and the levers are so propor-tioned, that a weight of thirty pounds presses the rollstogether with a force equivalent to 17536 pounds, whichwas found to be the proper force for the sufficient exten-sion of the gilding. By this contrivance four men canturn the rolls with more ease than ten can turn thosewhich are kept together by screws ; and the fame weightacting uniformly in every part, the pressure continuesalways equal, though the wooden roll should even becomeoval, and though the stuff be of unequal thickness.

A piece of cloth, of about two ells, is sowed to thebeginning and end of the stuff, to keep it out to its widthwhen it enters and parts from the rolls, which could notbe done by the hands for fear of burning or bruisingthem : as it would take too much time to sow these clothsto every small piece of an ell or two, a number of these issowed together. The stuff is rolled upon a cylinder, which

is