In some cases one side is left quite flat and the other side is ground to two or sometimes three angles, as in the carpenter’s chisel, a section of which is shown in Fig. 8, where _a a_ shows the angle made by the body of the blade; _b b_ the angle of the bevel formed by the grindstone, and _c c_ the angle of the cutting edge which is formed by the oilstone. In this case the flat side A B forms a straight guide and enables the workman, in cutting, to make straight, clean work. In sharpening such a tool on the oilstone the flat side is always laid flat on the oilstone and merely smoothed off,[19] and the skill of the workman is chiefly shown in the accuracy with which he forms the angle _c_ B _c_. The dexterity with which skilled workmen can effect this by the hand alone is astonishing, but in the hands of those who have had but little practice the edge infallibly becomes rounded as shown in Fig. 6. Therefore wherever a knife with a flat side is used, no attempt should ever be made to grind this flat side on the hone. Such an attempt will almost always result in an edge so irregular that it will be almost impossible to cut a good section with it. But if the under side of the knife be left truly flat, we will have the very best tool that can be had for use with the microtome. Unfortunately, however, it is very difficult to get a knife with a truly flat surface, that is, one that is perfectly “out of wind” as mechanics say. As a general rule the flat sides of knives are not true planes but irregularly curved surfaces; consequently when laid on a plane surface, such as the table of a microtome, they rock, and the edge is in contact with the table at some points and separated from it at others, and the points which are in contact change as the knife is moved diagonally across the table. This makes the section irregular and worthless. But a really good knife with a perfectly true flat side is such a valuable tool that a good deal of effort may be profitably expended in getting it, and when once obtained never let the flat side be touched by a grindstone, or more than touched by a hone. It is not absolutely necessary, however, that the sides of a delicate knife should be curved, provided the knife is used for cutting very soft materials. An angle of 10° or even 5° gives a pretty stout blade when carried back three-quarters of an inch, as may be seen by examining Fig. 1, and the guiding action of the very thick back may be obtained by means of an artificial guide, applied like the brass backs of tenon saws. Such an arrangement is shown in Fig. 9, and has been found very effectual. Footnote 19: Holtzapffel tell us that chisels that are required for paring across the end grain of moderately soft wood are considered to hang better to the work when they have a very slight keen burr or wire edge thrown up on the face or flat side of the tool. But this does not apply to section knives.