The cover of the box must now be put on, the lathe set in motion, and the outside, and also the top of the cover, carefully turned and finished. If the box is to be cylindrical, care should be taken that it is truly so, and that the angle formed by the junction of the top and sides is sharp. If sand-paper is used to finish the work, the edges will be rounded and the workmanlike appearance spoiled. If, therefore, the article is made of box or other hard close-grained wood, this finishing-off may be done with a carpenter's chisel held so as to act as a scraper. The turning chisel will answer the same purpose, but it is a pity to spoil the edge, which should be always preserved keen and fit for use. If the box is made of soft wood, scraping will not answer; the turning chisel must then be made use of, held as previously, described. If the cylindrical form is not proposed, the sides of the box must be left thicker, and after the cover is fitted on the outside may be moulded by the gouge and chisel, and tools like 60 and 68 to 70, to any desired pattern. The only thing remaining to be done is to cut off the box with the parting tool, the same precautions being observed which we spoke of in separating the cover. If there should be any defect in the bottom after the work is detached, the box must be placed in a cup chuck turned to receive it, and the above defects removed.

Figs. 69, 70, 71, 72, 73.

In hollowing out a piece of solid ivory or similar costly material, it would be exceedingly wasteful if the central part were removed as in a common box, by being reduced to small chips. It is possible to remove the whole interior in a solid block, and with exceedingly trifling loss of material. This is effected by means of side parting tools, 71, 72, 73. A common parting tool is first used and a groove cut therewith in the face of the block to be turned. [Fig. 74] represents this face 75; the section after the groove is cut the depth of the box required. The shaded part in the centre represents the part to be removed. The smallest parting tool, [Fig. 73], is now introduced, the back of the tool being laid across the rest, so that the crook takes a perpendicular position, A, B, [Fig. 74]. When at the bottom of the groove the hook is turned to the left, so that it may cut a groove underneath the block, until stopped by its shank. It is then withdrawn and [Fig. 72], and subsequently [Fig. 71] introduced, and used in a similar way. In [Fig. 76] the black line shows the tool in position, with the under cutting done by it. The sizes are thus increased until the last tool removes the block entire.[3]

[3] The side parting tools are sometimes inserted in the centre of the work, a hole being made for their introduction, they then cut from within outwards. In this case, however, instead of a solid piece a thick ring of the material is detached.

Figs. 74, 75, 76, 77.

We now propose to describe the method of turning a round ball or globe, and, to make the work more interesting, it shall contain a small box. The first thing necessary is to decide upon the diameter. In the present case let it be an inch and a half. Turn a cylinder of boxwood a little exceeding this, and cut off from it rather more than an inch and a half in length. The excess is merely to allow for waste. You will thus have a cylinder whose diameter equals its length. Before removing it from the lathe, mark its centre by a groove with a point tool—subdivide the outer spaces with five lines, and from the latter remove the corners of the piece, thus reducing it to the form 77. Test the length and breadth by the callipers and take care that the ends of the cylinder are at right angles to the sides. Now place the piece in the chuck in the position shown in the figure, that is, at right angles to its original position in the lathe. It must be tested as to truth by holding a point tool on the central line E, F. If correctly placed this will only make a dot when the lathe is put in motion. If the piece does not lie evenly the point of the tool will make a small circle—it must then be corrected with a light tap or two, until it runs evenly.

If the inside of the chuck is rubbed with chalk the work will be less liable to slip. The following operation, however, must be conducted very gently and with exceeding care, or a satisfactory result will not be produced. It will be observed that the central line having been marked or cut upon the side of the cylinder is necessarily a circle, and its revolution on its axis forms a sphere.

We have therefore only to cut away the piece truly down to this line to finish what ought to be a perfect globe. Bergeron, however, justly remarks that although the theory is correct it is next to impossible to manage the tools with sufficient skill to complete in this way a true sphere. One great cause of this difficulty, is that as the work revolves in its new position the central line is not visible as a line, but simply becomes the boundary of the sphere. This may be in part done away by making a red line or black one (red is the best) instead of a mark with the tool. The work will then appear red as it revolves, and the gouge and chisel must be used to cut away this red part, great care being taken only just to remove what appears coloured. Thus you will in the end have cut the work away so as barely to remove the line. Work from the central part outwards, and always with exceeding care, and you will eventually succeed to your satisfaction. It is, nevertheless, a very difficult bit of work to finish even fairly well—mainly on account of the great obscurity of your landmark, the red line. For the more perfect finishing of the above a template of steel may be made like [Fig. 78], with which to test the work—its diameter is equal to that of the sphere, and it will serve as a gauge or scraper. It should be made of saw plate if intended for the latter purpose, otherwise sheet brass will answer as well. After the semi-globe has been turned in the first chuck, it will be necessary to turn another to receive the finished part, and for the more perfect formation of the same a semicircular template of the same gauge as the concave one first made may be provided, as the more nicely the ball fits the chuck, the less chance there will be of the work shifting during the turning of the latter half of the sphere.