The file, if concave or hollow in respect to its length, in the manner coarsely exaggerated in fig. 832, might be used for works of corresponding convexity; but it would be impossible to flat a flat surface therewith, as the concave file would only touch the surface at its edges, but the convex side of the same file might, as in fig. 833, be made to touch any and every part of the surface if moved in a right line. On this account most files are made thicker and wider in the middle, or with both faces convex, and the error of hardening will then rarely make either side concave, but will leave both faces convex, although differently so; and consequently, both sides, notwithstanding some irregularity, are useable upon flat works, provided the operator can move them in a right line across the work.
In reference to the manipulation of the instrument, it is to be observed that the most natural movements of the hand and arm are in circular lines, the several joints of the limbs being the centers of motion; but, as in filing a flat surface, it is needful the hands should move very nearly in right lines, a kind of training becomes necessary.
If, however, the file were carried quite straight across a wide surface, the central part of the file would be alone used; but as the continual effort of the individual is to feel that the file lies in exact contact with the surface being filed, the hands imperceptibly depart so much from the exact rectilinear path as to bring all parts of the file from point to heel into use.
Again, it might be urged that the file, from being itself in the form of the arc of a large circle, would reduce the work to the counterpart form, or make it hollow in the opposite degree; it is true this is the tendency, and may by dexterity become the result, even on narrow pieces; but the contrary error is more common, so that the surface of the work becomes rounded instead of concave or plane.
If the surface to be filed is four or five inches or more in width, the risk of departing from the true figure becomes reduced, as the file has then a wide base to rest upon, and the pressure of the hands readily prevents any material departure from the right position of the file; but the difficulty becomes greatly increased when the surface to be filed is narrow.
The file held in the two hands upon the narrow work, may be then viewed as a double-ended lever, or as a scale beam supported on a prop; and the variation in distance of the hands from the work or prop gives a disposition to rotate the file upon the work, and which is only counteracted by habit or experience.
Assuming, for the moment, that in the three diagrams the vertical pressure of the right hand at r, and the left at l, to be in all cases alike, in fig. 834, or the beginning of the stroke, the right hand would, from acting at the longer end of the lever, become depressed; in fig. 835, or the central position, the hands would be in equilibrium and the file horizontal; and in fig. 836, or the end of the stroke, the left hand would preponderate; the three positions would inevitably make the work round, in place of leaving it plane or fiat.
It is true the diagrams are extravagant, but this rolling action of the file upon the work is in most cases to be observed in the beginner; and those practised in the use of the file have, perhaps unconsciously, acquired the habit of pressing down only with the left hand at the commencement, and only with the right hand at the conclusion of every stroke; or negatively, that they have learned to avoid swaying down the file at either extreme, and which bad practice will necessarily result, if the operator have not at first a constant watch upon himself, to feel that the file and work are always in true contact, throughout the variable action of the hands upon the instrument.
When the work is fixed in the bench or table-vice, the file is almost always managed with both hands, as above described; but 1 lien the file is held in the one hand only, all the circumstances are altered, except the continued attempt to keep the work and file in accurate juxta-position; and to assist in this, the work when so small as to be filed with the one hand only, is almost invariably held on the filling-block with the left hand, occasionally through the intervention of a hand-vice, as in fig. 858, page 862.
In this case the two hands act in concert, the right in moving the file. the left in adjusting the position of the work, until the indi-vidual is conscious of the agreement in position of the two parts. Sometimes indeed the partial rotation of the work, in order to adapt the work to the file, is especially provided for, so as to compensate for the accidental swaying of the file; such is the case in the various kinds of swing tools, used by watchmakers in tiling and polishing small flat works. A similar end is more rarely obtained, on a larger scale, when the file is required to be held in both hands. For example, filing-boards resembling fig. 837, and upon which the work is placed, have been made to move on two pivots, somewhat as a gun moves on its trunnions; consequently the works, when laid upon the swinging board. assume the same angle as that at which the file may at the moment be held.
A more common case is to be seen in filing a rectangular mortise, or key-way, through a cylindrical spindle, as in fig. 888; the hole is commenced by drilling three or four holes, which are thrown into one by a cross-cut chisel, or small round file; and the work, when nearly completed, is suspended between the centers of the lathe, so that it may freely assume the inclination of the file. At other times, the cylinder is laid in the interval between the edges of the jaws of the vice, that arc opened as much as two-thirds the diameter of the object, which then similarly rotates on the supporting edges; this mode is shown in fig. 889.