THE FOLLOWING IS the concluding part of the article on mirror making on the hand-lever spindle machine, which was begun last month by Patrick A. Driscoll of the Eastman Kodak Company Hawk Eye Works.

CLEAN up all traces of 180 emery around the machine, make up a jar (one pint) of 500 emery, and use a new brush. Mike the glass and write this figure down. We must grind with 500 emery until we have removed at least 0.008" and until the surface once more exactly fits the gage.

Before proceeding with the 1200 emery, the mirror must be re-miked on its location mark and the reading noted town. To wipe out all No. 500 pits we must remove at least 0.004".

Assuming that the surface is once more a gage fit we now can look forward to polishing, with the assurance that we have a perfect fine-ground surface, free from pits or scratches. A spindle speed of 200 r.p.m. was recommended for grinding. For polishing, we must reduce this to about 36 r.p.m. Any speed within those respective ranges will, of course, prove quite satisfactory. In Figure 1 (in last month's installment) the pulley or gear trains have been left to the builder's discretion, in the belief that many will have their own gear boxes and pulleys and individual initiative for devising ways to change spindle speed.

For polishing, Hindle's machine ("A.T.M.," page 234) will give a close approximation to the action of the machine in Figure 1 and will be quite satisfactory if the worker prefers it. In either case, the spindle should have a speed of about 36 r.p.m. All belting Hindle or Driscoll, should be double V belt. The speed reduction from the motor can be accomplished by belting or, in the probable event of financial affluence, by a gear reduction unit. Try the gear shift unit from the family flivver.

Elaboration or addition of automatic stroke is left to the individual's discretion.

To shape the cold polisher to curve we place it and the mirror in a large pan of water and slowly heat them until the pitch is fairly soft to the touch and the glass is fairly warm to handle. Next, place the hot, wet polisher on the machine, apply thick rouge, start the spindle rotating and form the polisher to curve, using the mirror as a forming instrument.

If the polisher and glass cool down too soon, dunk them back into the pan of water for a while and proceed as before. The pan of water will stay good and warm if you make it big enough.

Do not use soap and water as a lubricant.

The rouge will color the polisher as it conforms to the glass, and any black areas may be assumed not to be in contact. But when the lap is a dull red all over it is formed to curve and contact.

To cut the polisher we place it, cold, on the machine and, with the spindle rotating, hold the corner of a single edged razor blade against it and scrape, not cut, grooves into it, spaced about 3/8" apart and 1/16" deep. They need not be wide nor deep. I have found nothing better than the single-edged razor blade for cutting the polisher, not because of its cutting edge but for its sharp corner when used as a scraping device.

During the forming of the polisher and in all polishing it is most important that the grooves in the polisher or lap be kept open. They do not have to be deep or wide.

On a concave surface a polisher having the same diameter as the work will have a tendency continually to shorten the radius of curvature, that is, it will "drive the center low," polishing the center more than the edge. Therefore a hyperbola develops, even with a rational stroke. In trying, then, to lengthen the radius and get out of his hyperbola, the worker resorts to a very short stroke, and thus, by polishing the edge much more than the center, he lengthens its radius and winds up with two foci, a short center and a longer (turned) edge. The turned edge is caused by trying to correct the tendency of the shortening effect on the radius.

Thus, three fourths of the polishing time is lost in trying to keep the focal length up to standard and, in so doing, three fourths of the abrasive effect of the polisher is put on the edge of the mirror. This is one of the causes of that great bugaboo, turned edge.

By inverting polisher and mirror the worker can reverse the action of his equal diameter polisher but, while the polisher when on the bottom unduly hits the center because it is too small, so does the polisher when on top unduly hit the edge because it is too big-and in both cases much too hard. Thus we have the dilemma: either the turned edge or hyperbola. Of course, by juggling the polisher and glass, and by using all sorts of strokes, a spherical surface is finally attained-if our judgment, timing, and patience are working overtime.

A convex surface with equi-diameter polisher on top will be hardest hit on the edge (and once again a shortened radius). The equi-diameter method holds the worker to a short stroke, less abrasive or polishing action, and a greatly increased polishing time.

Proceeding once more on the assumption of a concave surface to be polished, we place the six-fifths diameter polisher on the spindle, lay the mirror face down on it, adjust the stroke neither long nor short but medium, and give the work a 15-minute polishing spell.

The rouge should be of a creamy consistency. Keep it in a jar and apply it with a soft 1/2" brush.

The center of the mirror should cross center of polisher at every stroke.

For a 6" mirror a pressure of about eight pounds will suffice. A fair rule for weight on the drive-pin of the stroke arm (which has a slip fit so that the weight of the worker's arm, itself, is not transmitted to the work) is 1-1/2 pounds per inch of diameter.

Remove the work and measure the radius by the Foucault test. Use a 1/4" pinhole covered with window screening, and substitute a piece of fine ground glass for the knife-edge, the two mounted so as to move as a unit together. When the image of the squares of screening is sharpest on the ground glass (scan with a 1" f. 1. magnifier) we have ascertained the radius of curvature, and half this distance is the f. 1. of the surface within 1 mm.

If the radius is too long the center of the mirror must be polished more than the edge. Therefore, increase the length of the stroke. If the radius is too short the stroke must be shortened to polish the edge more than the center.

By testing often, the action and effects of the machine will become more and more familiar as we proceed.

Thus far the radius of curvature has been referred to as too long or too short. In optics a concave surface is often referred to as "high" (long radius) or "low" (short radius), and hereafter we shall use these terms.

If the low surface is fitted to the gage it will show space under the center, because the curve is too deep. The high surface will show space under the edges because the curve is too fiat.

Let the polishing spells be about 15 minutes in length. When approaching the desired radius of curvature, shorten these spells and test often.

Ring grooving of the polisher affords an additional correction to obstinate zones. An example: The curve is too low (deep) and short strokes do not seem to change the radius during a few polishing spells. Therefore, from the center to half way out on the polisher we scrape additional grooves between each present groove. Conversely, if the surface is high we scrape more grooves on the polisher's outer area.

A warning: Do not make these corrective grooves too deep. Be easy with pressure on the razor blade scraper. The grooves are to correct a temporary fault only, and when this has been accomplished we do not want them to continue their effect. Regroove them lightly and often but only as long as they are needed. If, however, they should remain after the surface is corrected, scrape balancing grooves on the remainder of the polisher, to nullify their effect and from now on let these intermediate grooves close up but do not fail to keep the regular grooves open at all times. Never let them close.

A corrective method, alternative to local grooving, is to scrape the center or edge with the razor blade held square on. Scrape ever so lightly, my friend, ever so lightly. Even with this warning some probably will scrape too much and get into trouble.

If a slight stain appears on the mirror, a 15-minute spell using plain water and no rouge will usually remove it. If not, add about ~ teaspoonful of vinegar to ~ pint of rouge.

It is not necessary to remove the work every time the tool needs recharging. In applying emery or rouge use a brush and apply while the machine is running.

Since all the foregoing has treated the concave surface for the benefit of the mirror maker, we must now make amends to the refractor builder and not leave him high and dry. Let us bring him down to our lowly level and carry this discussion into the field of convex surfaces.

Our rule calls for the convex surface to be ground and polished face up, tool on top. The grinder will be 4/5, and the polisher 5/6, the diameter of the lens.

In reversing the position of the lens from the top to the bottom we also reverse the action of the stroke for correcting high or low test findings. The high concave would require a long stroke. The high convex would require a short stroke. The low concave would need a short stroke. The low convex would need a long stroke. In sum, on a concave surface, when the curvature is not deep enough it is high. When the curvature is too great, it is low. On a convex surface the conditions are directly opposite. The polishing technique for a convex surface is different from that of the concave only in the inversion of the lap and action of the stroke.

A last and oft-repeated warning: Do not attempt to polish until the surface exactly fits the gage after fine grinding. We do not wish to polish ourselves into an early grave.

Finally, I extend my condolences, in the form of a more simplified polishing machine, to the amateurs who lack the material, space, and so on, for a better one. This machine will do everything the one already described will do, except that the time factor will be extended considerably. The spindle should be vertical, running at about 36 r.p.m., and the stroke arm should be double, consisting of your own two strong arms with sleeves rolled up. Merely place the work in a chuck (Figure 3, insert, last month), and apply your strokes exactly as if you were a part of the machine. The chuck for hand polishing may be made of metal or wood, and lined with glued-in felt. The drive pin notch shown on its top is not for hand work, but with it this same chuck may, if preferred, be substituted for the metal button of Figure 1, in polishing.

After each 15 or so strokes, let go of the chuck. The work will revolve with the spindle. After about three seconds drop the hands back on the chuck and proceed. Do not be afraid of the Work flying off the spindle when you let go. it won't, since 36 r.p.m. is not fast enough to cause alarm on this score

With this machine no single condition or relation of work to polisher will be duplicated, and neither monad, astigmatism, nor one-sided test (one area of different radius) will occur.

It is my humble opinion that amateur optics can well afford to get away from some of the do's and don'ts prescribed by workers in the past, not omitting the wearing of round-the-barrel ruts in the rug. So roll out the barrels.

End of Driscoll's contribution.

READERS are warned that the 200 r.p.m recommended in the above articles should not be carelessly exceeded. "Faster, disaster," is Driscoll's warning, in a private communication, while Leo J. Scanlon of Pittsburgh, who has used a hand lever machine, is of the same mind and urges not falling asleep even at 200 r.p.m. Especially is this pertinent in the final stages of grinding since at this stage the grinding surfaces become dry or warm and seizure may take place so quickly as to end possibly in the disaster against which Driscoll warns. Henry Paul, Norwich, N. Y., also has used a hand lever machine and protects his mid-section by embodying a heavy steel splash pan in his equipment; an added piece of plank would serve the same purpose. In the next number an illustrated description of Paul's hand-lever machine will be published, with numerous sidelights on its operation.

GLEANED while reading a back number of the Journal of Scientific Instruments (London, February, 1938, p. 3) is the following fragment abstracted from a paper on "The Mechanical Amplification of Small Displacements," by Professor A. F. C. Pollard, and of probable interest to Carborundum-conscious readers.

Place a well-made agate knife-edge from a micro-chemical balance in contact with a microscope cover-glass under a high power of the microscope, in monochromatic light. Note fringes from the two facets, also black areas surrounded with very irregular contours from the edge.

"One wonders," Prof. Pollard remarks, "how it is that agate knife-edges function as well as they do."

Now select a perfect Carborundum crystal about 1/8" long and similarly test it. Though the crystal edge included by the facets is at an angle of about 117 degrees, this edge is almost perfect and is seen as a black fringe along its whole length. Such perfect crystal edges should form ideal knife-edges for highly sensitive and delicate lever systems, Professor Pollard states, "but I know of no case in which the crystals have been used for this purpose." But he adds that Professors McBain and Tanner have used two sharp points of Carborundum crystals to function as the knife-edge of their highly sensitive microbalance, and with great success.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.