Instead of a great wheel revolving horizon tally, like Hebert's, he employed a little wheel (about 30 inches diameter), which revolved vertically like a grindstone; on the outside edge of which were fixed, in suitable carnages, the glazers and grainers, provided with proper adjustments. The table on which the skins were laid was a hollow segment of a circle, of the same radius as that described by the rubbing surfaces of the glazers and grainers; thus the skins were impressed with right lines by a curvilinear motion, owing to their lying in a hollow curve. The specification of this invention, very ably and fully described, is given in the thirteenth volume of the Repertory of Arts, Second Series, to which the reader is referred for the details; an inspection of the drawing* in which inclines us to think the machine may have failed from two causes; First, the finisher incurs great risks from blows by the revolving tools as they successively descend into the hollow curve wherein the skin is operated upon. He has to look down this curve to see his work; and as the view is a very unfavourable one for examination, he might inadvertently put his head too near and get a fatal blow.

If a guard were put up to prevent such an accident, it would be in the way, and obstruct light in a situation where more 13 wanted. Second, The necessity of extraordinarily accurate workmanship to make a perfect adjustment of the concave surface of the table, with the curve described by the revolving tools. A third objection will probably lie against the direction in which the table is brought into or goes out of action; it is made to slide in a horizontal and tangential line with the lower side of the circle described by the tools, consequently the lower extremity of the table comes into action first, which must subject this important part of the machine to shocks very unfavourable to the preservation of a perfectly true bearing. We submit that it would be better to make the table move in a radial line to the centre of the wheel, or at the least, in a tangent to a very small inner circle.