324.  It is partly owing to the imperfection of the original trials, and partly to the gradual improvements in the art of making machinery, that many inventions which have been tried, and given up in one state of art, have at another period been eminently successful.  The idea of printing by means of moveable types had probably suggested itself to the imagination of many persons conversant with impressions taken either from blocks or seals.  We find amongst the instruments discovered in the remains of Pompeii and Herculaneum, stamps for words formed out of one piece of metal, and including several letters.  The idea of separating these letters, and of recombining them into other words, for the purpose of stamping a book, could scarcely have failed to occur to many:  but it would almost certainly have been rejected by those best acquainted with the mechanical arts of that time; for the workmen of those days must have instantly perceived the impossibility of producing many thousand pieces of wood or metal, fitting so perfectly and ranging so uniformly, as the types or blocks of wood now used in the art of printing.

The principle of the press which bears the name of Bramah, was known about a century and a half before the machine, to which it gave rise, existed; but the imperfect state of mechanical art in the time of the discoverer, would have effectually deterred him, if the application of it had occurred to his mind, from attempting to employ it in practice as an instrument for exerting force.

These considerations prove the propriety of repeating, at the termination of intervals during which the art of making machinery has received any great improvement, the trails of methods which, although founded upon just principles, had previously failed.

325.  When the drawings of a machine have been properly made, and the parts have been well executed, and even when the work it produces possesses all the qualities which were anticipated, still the invention may fail; that is, it may fail of being brought into general practice.  This will most frequently arise from the circumstance of its producing its work at a greater expense than that at which it can be made by other methods.

326.  Whenever the new, or improved machine, is intended to become the basis of a manufacture, it is essentially requisite that the whole expense attending its operations should be fully considered before its construction is undertaken.  It is almost always very difficult to make this estimate of the expense:  the more complicated the mechanism, the less easy is the task; and in cases of great complexity and extent of machinery it is almost impossible.  It has been estimated roughly, that the first individual of any newly invented machine, will cost about five times as much as the construction of the second, an estimate which is, perhaps, sufficiently near the truth.  If the second machine is to be precisely like the first, the same drawings, and the same patterns will answer for it; but if, as usually happens, some improvements have been suggested by the experience of the first, these must be more or less altered.  When, however, two or three machines have been completed, and many more are wanted, they can usually be produced at much less than one-fifth of the expense of the original invention.