If grinding be not adapted to form a true general outline, neither is it to produce accuracy in the minuter detail. There can be little chance of a multitude of points being brought to bear, and distributed equally, under a process from which all particular management is excluded. To obtain any such result, it is necessary to possess the means of operating independently on each point, as occasion may require, whereas grinding affects all simultaneously. It is subject neither to observation nor control. There is no opportunity of regulating the distribution of the powder, or of modifying its application, with reference to the particular condition of different parts of the surface. The variation in the quantity of the powder and the quality of the metal, will, of necessity, produce inequalities, even supposing they did not previously exist. Hence, if a ground surface be examined, the bearing points will be found lying together in irregular masses, with extensive cavities intervening. An appearance, indeed, of beautiful regularity is produced, and hence, no doubt, the universal prejudice so long established in favour of the process. But this appearance, so far from being any evidence of truth, serves only to conceal error. Under this disguise surfaces pass without examination, which, if unground, would be at once rejected.
