[37] If that is too expensive, some of those given in the preceding pages will probably answer your purpose.
[38] "Soils which are naturally porous, from which rain rapidly disappears, are known to be the healthiest for the sites of houses. In this the action of the soil oxidizes all organic impurities, the resulting product is washed away by the rain, and the soil remains sweet and wholesome."—Latham.
[39] To find the number of steps for a given situation, find the height, as just shown, from floor to floor, [102]" for example. Assume, for trial, a satisfactory height for each step, as 7". Divide 102 by 7, which gives 144/7 for the number of steps. To make the number even, call it 14, and you have only to divide 102 by 14 to get the exact height of each step.
[40] Unless too heavily loaded, a canvas-covered canoe will float in case of a capsize, but some form of air-chambers is desirable and a safe precaution in any small boat. It is hardly safe to rely upon your ability to build water-tight compartments in the ends of canvas (or wooden) boats, as is sometimes recommended—that is, as a part of the regular construction of the boat. It is not easy for an amateur to do this. It is better to have the air-tight compartments made separately and independent of the boat itself. Copper boxes or air-tanks fitted to the space at the ends are the best and the only really reliable expedient, but they are expensive. Light wooden boxes covered with canvas and thoroughly painted can be used, as well as galvanised boxes or even varnish cans sealed and painted. Any such contrivance can be made tight at first, but is always liable to become leaky (except by the use of copper tanks), particularly as it is usually concealed from examination.
[41] Mortise-chisels with great thickness of blade (Fig. 486) are not likely to break, and the width of the sides bearing against the sides of the mortise tends to make the cutting more accurate.
[42] You may be told that perfect joints do not require much clamping, but a perfect joint is impossible, and as a practical matter, only the skilled workman or the most accurate machinery can make even a good joint of much length, so great is the difficulty of avoiding little inaccuracies. Besides this, there is always the liability to more or less springing or change of shape on the part of the pieces. The joint which was good when you stopped planing may not be as good by the time the glue has set, particularly if the gluing does not immediately follow the jointing. In addition to this, the pressure from clamping at only one or two points, or at points too far apart, may force the joint to open elsewhere. Do not infer from this that even the beginner should be content with a poor joint, with the idea that it can be squeezed and jammed to a sufficiently good fit by applying muscle to the clamps. Of course this jamming or mashing of the fibres to fit occurs, to a microscopic degree, in even the best joint, and it can sometimes be done to a perceptible extent with soft wood, but to do this intentionally is very unworkmanlike, and the greatest care should be taken to make as good a joint as possible before gluing and applying the clamps. Do not, however, flatter yourself that you can make so accurate a joint that you can afford to neglect proper clamping, unless, in such cases as that shown in Fig. 488, you adopt the old-fashioned way of rubbing the two edges together and then leaving the rest to the glue, but this is not so good a process for the beginner, except with small pieces, such as corner-blocks (see Corner-blocks). See Jointing.
[43] Shellac is, strictly speaking, a kind of varnish, but it is so different from many kinds of varnish in common use that it is quite commonly spoken of as shellac, in contradistinction from what is popularly known as varnish, and the term is so used here.
[44] In shell acing doors or panel work, first shellac the panels, then the rails, and finally the styles (see Fig. 505), because daubs or runs can be wiped off and covered better when you thus follow the construction of the work.
[45] As an extreme illustration, it may be interesting to note the way the best lacquer work (which is so durable) is made by the Japanese, an article being given, as Professor Morse tells us, one coat a year, the finest work having twenty-one coats and the artist rowing out to sea for miles each time to make sure that all dust is avoided.
[46] Two pieces properly glued are often stronger than one solid piece—that is, the glued joint is stronger than the wood itself, as you will probably discover some day when you have occasion to break apart a piece of good gluing; but after a long time the glue is apt to deteriorate in adhesive or cohesive force, particularly if the joint has not been protected by paint or varnish, so do the best work you can if you wish it to last. Nevertheless, in important work it is usually safest to take a whole piece when you can, rather than glue up two or more pieces, except in cases, perhaps, where the matter of warping, etc., is concerned, when it may be better to build up the desired shape of pieces selected for the purpose.