If the piece of cast-iron should have been turned in the lathe, or planed in the planing machine, instead of having been wrought entirely with the chipping chisel, plane and file, the former instructions would be uncalled for, as the remaining steps alone would remain to be followed. Unless, indeed, the work had been so imperfectly fixed as to have been strained, and thence become distorted on being released from the machine: on the latter supposition the grosser errors would probably require correction with the bastard file, before the smooth file could be judiciously used.

The necessity of the convex form of the file will now be rendered most striking, as were the file absolutely flat on its face, it would be scarcely possible to reduce with it any small and isolated spot that might become coated with the red chalk from off the planometer; but as the file is a little rounded, any precise spot on the work may be acted upon, as the end of the file may be pressed with the fingers of the left hand on the exact spot to be reduced, whilst the remainder of the file is held just out of contact with the rest of the surface.

When, however, the points of bearing become numerous, the file cannot even thus be managed with sufficient discrimination, and notwithstanding the best efforts it will act on too large a part, and thereby lengthen, or it might be said altogether to prevent the complete correction of the work, because the file is not sufficiently under control. Before the file has assumed this questionable tendency, it is politic and usual as a measure of economy, to discontinue the use of the file, and to prosecute the work with a scraper, which having a sharp edge, instead of a broad and abrading surface, may be made to act with far more decision on any, even the most minute spot or point. A worn-out triangular file, ground at the end on all the faces, so as to make thin keen edges, is generally used as the scraper; this should be keenly sharpened on an oil-stone, so as to act without requiring much pressure, which would only fill the work with striae or utters.

The continual reduction of all the points, which are sufficiently prominent to pick up the colouring matter from the planometer, is now persevered in with the scraper instead of the file, and precisely in the same manner, except as regards the change of the tool; and if the process have been carefully performed throughout, it will be found at the conclusion that if the work and plano-meter are both wiped clean, and rubbed hard together, that the high points of the work will ho somewhat burnished, giving to the work a finely mottled character.

In producing metallic surfaces, the constant effort should bo to reduce all the high places with as much expedition as circumstances will admit, but avoiding, on the other hand, that energetic use of the tool, which may too hastily alter the condition of the surface, and in expunging the known errors, induce others equal in degree but differently situated. Throughout the work, attempt should be made to keep the points of bearing, whether few or many, as nearly equidistant as may be, instead of allowing them to become grouped together in large patches.

In respect to the tools, there should be a gradual diminution in their cutting powers, and also of the vigour with which they are used, as although energy is wise at the commencement of the work, it should gradually subside into watchfulness and caution towards the conclusion. The periods of alternation between the hand-plane and the file, and also the times when these are successively rejected, in favour of the scraper as the finishing tool, must be in great measure left to the judgment of the operator.

There should be a frequent examination of the work by means of the straight-edge and planometer, which latter should at all times be evenly tinted with the colour. At the commencement it is necessary the coating of red stuff on the planometer should be moderately abundant, so as to mark even those places which are minutely distant, but with the continued application of the work, the colouring matter will be gradually removed from the planometer, and which is desirable, as towards the conclusion the quantity of red should be small, so as but faintly to mark the summits of each little eminence, the number and equality of which are dependent on the perfection of the planometer, and the steady persevering watchfulness of the operator.

It is not to be supposed that it is in every case needful to proceed in the careful and progressive mode just described, as the parts of different works require widely different degrees of perfection as to flatness. For instance, in many it is only necessary they should be clean and bright, and have the semblance of flatness, with such even the straight-edge is little if at all used as a test. Those surfaces by which the stationary parts of framings are attached, require a moderate degree of accuracy, such as may be comparable with the perfection in the hewn stones of a bridge or other massive edifice, which require to be flat, in order that they may bear fairly against each other, as without a certain degree of truth the stone might break from the unequal strain to which it would be exposed.

The flat parts of metallic works, if similarly imperfect, would bend, and perhaps distort the remainder; but although it is of great importance that bearing surfaces should be out of winding, or not twisted, it is by no means important that such bearing surfaces should be continuous, as a few equally scattered bearing points frequently suffice. Thus it was the common practice before the general introduction of the engineer's planing-machine, to make fillets or chipping places around the margins of the bearing surfaces of castings, which fillets alone were corrected with the chisel and coarse file, for the juxtaposition of the larger pieces or frame work of machines, the intermediate spaces being left depressed and out of contact. This mode sufficed, provided the pressure of the screw-bolts could not, by collapsing the hollow places, distort the castings, with which view chipping places were also generally left around the bolt holes of the work, this method greatly reduced the labour of getting up such works by hand; but fillets and chipping places are now in a great measure abandoned. Smaller and more delicate works, requiring somewhat greater accuracy than those just described, are left from smoother files, but in most cases without the necessity of scraping; but the rectilinear slides and moving parts of accurate machinery, and the trial or surface-plates of the mechanician, require beyond all other works, the most dexterous use of the file and other means, from which it is again repeated, grinding should be entirely excluded.

Until very recently, when the points of bearing had been so multiplied by the file and scraper, as not to exceed about half an inch in average distance, and that a still higher degree of accuracy was desired; it was the ordinary practice to attempt the obliteration of these minute errors by the method of grinding.

Supposing only one piece or surface to have been required, it then became necessary to grind the work upon the planometer itself, but to avoid the necessity of so injurious a practice, it was usual, when practicable, to make three similar pieces at one time, in order that, when all three had been separately filed and scraped to agree pretty nearly with the straight-edge and piano-meter, the three pieces might then be mutually employed in the correction of one another, by grinding the faces successively together with emery-powder and water.

The one piece was laid down horizontally, wetted all over with water, and then strewed with emery powder, after which, one of the other surfaces was inverted upon it, and rubbed about in various ways with longitudinal, lateral, and curling or circular strokes,on the supposition that as the two pieces came into contact respectively at their highest points, these highest points became mutually abraded, with a tendency to reduce them to the general level. After a short period, the top surface was removed, fresh emery and water were applied, and the third surface was rubbed upon the first; after which, all three were variously interchanged, by placingeveryone in succession as the lower surface,and robbing the two others upon the lower until it was considered from the uniform but deceptive grey tint thus produced, that the errors in all were expunged, and that the three surfaces were all true.

It is, however, considered quite unnecessary to enter more into detail on a process that may be considered to be nearly obsolete, as regards the production of plane metallic surfaces, especially as at a future part of this Volume, the practice of grinding will be noticed, in reference to surfaces requiring inferior exactness, and consisting of materials that do not admit of the employment of the file and scraper.

That two surfaces which are very nearly accurate, if ground together for a very short time, do in some degree correct each other, is true, but it has been long and well known, that a continuance of the grinding is very dangerous, and apt to lead the one surface to become convex, and the other concave in a nearly equal degree, and on this account, three pieces were usually operated upon that the third might act as an umpire, as although two pieces possessing exactly opposite errors may appear quite to agree, the third cannot agree with each of these two, until they have all been made alike, and quite plane surfaces.

But the entire process of grinding, although apparently good, is so fraught with uncertainty, that accurate mechanicians have long agreed that the less grinding that is employed on rectilinear works the better, and Mr. Whit worth has recently shown in the most satisfactory manner,* that in such works grinding is entirely unnecessary, and may with the greatest advantage be dispensed with, as the further prosecution of the scraping process is quite sufficient to lead to the limit of attainable accuracy; the only condition being, that the mode of continually referring the work to the planometer, and scraping down the points sufficiently high to be coloured, should be steadily persevered in, until the completion of the process, and works thus treated assume a much higher degree of excellence than is attainable by grinding.

Mr. Whitworth stated a further and equally important advantage to result from the discontinuance of grinding, as regards the slides and moving parts of machinery. Some of the grinding powder is always absorbed in the pores of the metal, by which the metallic surfaces are converted into species of laps, so that the slides and works carry with them the sources of their depreciation and even destruction. The author's previous experience had so fully prepared him for admission of the soundness of these views, that in his own workshop he immediately adopted the suggestion of accomplishing all accurate rectilinear works by the continuance of scraping, to the entire exclusion of grinding.