IN MAKING achromatic objectives for refractors, an approach from another direction than the more familiar one described below by Patrick A. Driscoll, 39 Bateau Terrace, Rochester N. Y. While this "testplate method" involves a little more work, it requires no optical flat and avoids other headaches.

Asked to tell about himself, Driscoll says: "In 1930 I stumbled across the telescope making hobby when I was a ship model fanatic, and suffered a direct hit in the superstructure. I have had Carbo in my teeth ever since, having been an 'A.T.M.' (amateur telescope maniac) and made a 5" two 6" achromats, also eyepieces, prisms, and so on. I am now an experimental lens maker employed by the Eastman Kodak Co., Hawkeye Works. Thus you might call me a sort of amateur professional amateur." His article:

SPECIFICATIONS-The data in Figure 1-the radii, r, each r being numbered with a subscript, and thickness t, similarly numbered-have been very carefully calculated to produce an ideal telescope objective. All tolerances have been worked out to give the greatest permissible leeway, for error in measurements.

Crown component: Bausch and Lomb Optical Co. (Rochester, N. Y.) Light Barium Crown LBC-2 N_d 1.5711, V 57.3, thickness at center 17 mm, diameter 4-7/8" (thickness of blank 19 mm, diameter 5-1/4'').

Flint component: Bausch and Lomb Extra Dense Flint EDF-2, N_d 1.6890 V 31.0, diameter 4-7/8", thickness at center 14.5 mm (thickness of blank 18.5 mm, diameter 5-1/4'')

Finished diameter, 125 mm; e.f.1., 1480.82 mm (58.30"); focal ratio 11.84.

Hold thickness of each component to plus or minus 1 mm; radii to 1 in the third significant figure (for example, plus 666.46 would be passable at 665.46).

To convert millimeters to inches divide by 25.4.

METHOD-In place of the method involving regrinding or repolishing the fourth surface to correct variations of curves on the other three, we shall first make concave test plates, or masters, to match the desired curve of each lens surface, controlling them by our familiar old friend, the Foucault knife-edge test, and then test our lens surfaces against these test plates by means of Newton's rings and the simple setup described in "A.T.M.," page 244. We shall then be sure that, since the original computations and specifications by which we make the masters are correct; and since we shall make the masters correctly to them; and since we shall make the lens surfaces faithfully to these masters, our achromatic lens itself will be-must be-correct.

Making our test plates will be exactly like making spherical mirrors, but four spheres look like a lot of work. Let's investigate.

Surface 1, the r₁ surface of Figure 1, calls for one test plate.

Surface 2 will be worked against surface 3, tool-and-mirror fashion; surface 3 becomes the test plate for surface 2.

Surface 4 calls for a test plate.

Net extra labor, therefore, two temporary test plate curves-minus the job of making a flat.

For the first of the surfaces-surface 1 (+935.51 mm)-we shall use the crown blank as the tool and work a 5-1/4" disk of common one-inch plate glass to the corresponding concave radius; roughing, grinding, and fine grinding precisely as in producing a spherical mirror, as described in "A.T.M." To obtain the desired curvature, an arc of radius 935.51 mm is scribed on a thin piece of window glass at least 14" long, using a glass cutter, and this glass is broken apart to make a template ("A.T.M.," pages 310, 344); the best 6" section that shows evenness of break being used. Our tool when this is done, will be our crown component and will have been completely fine ground to curve on one side.

For the other test plate, the one for surface 3 (- 666.46 mm), we use the other side of the crown as the tool and the flint as the "mirror," proceeding as before. The result is the flint fine ground to -666.46 mm and the second surface of the crown similarly fine ground to the corresponding plus curve. This finishes the grinding of the crown; care, of course, having been taken to bring it to the desired thickness-the computations called for 17 mm and we started with 19 mm, a 2-mm leeway for grinding. We also have, as has already been stated, a 1-mm plus or minus tolerance on thickness of either component. This will eliminate elaborate miking devices and automatically chloroform that bugaboo of the amateur optician, the fear of deviating from stated formulas. We would not ruin our reputation in optics nor would we spoil the performance of our objective lens if, on completion, the latter were to come out with a crown curve, for example, of plus 666.00. Any such I small percentage of error still will produce a good achromat and a pleased maker.

In grinding a lens having a curve on either side it is important that the surfaces be centered with respect to each other; that is, that the optical axes of the respective surfaces be made to coincide longitudinally; otherwise we shall have what is termed a prismatic lens and this "prism" then could be removed only by centering the lens far below the diameter we want. The 5-1/4" blanks have enough excess of glass for centering out slight errors in edge thickness but, when grinding the second sides, we should occasionally check the thickness at several points around the circumference. Micrometering is best but measuring with a steel scale will be satisfactory at this stage. The mike shown in Figure 2, with some kind of stopblock attached, will give very accurate readings, but such close readings-an alternative for the most exacting amateur-are unnecessary.

If one side shows thicker than the other, apply more grinding to the thick side. Stop occasionally moving around the lens and make extra strokes on that side.

For the last step in the grinding operations we now take our flint blank which is already ground to the -666.46 curve (surface 3) and, turning it over and using it as a tool, place the plate glass test plate (which already has the -935.51 curve fine ground on it) flat side down and once again make a concave surface, this time -2884.64 When this concave test plate is finished in its fine grinding, the last side of the flint lens is also finished to +2884.64 so far as grinding is concerned. Care is taken again to bring it to tolerated thickness and to check for prism. Our test plate also is finished, so far as grinding is concerned, since it now has concave curves on both sides. This eliminates a second plate glass blank and gives us something to feel really bad about if we drop it.

The flint lens, having a concave surface on one side exactly the same curve as the convex surface of the crown-automatically becomes the third test plate surface; the only drawback being that, in placing the flint and crown together for testing purposes, care must be taken not to slide them and scratch the flint lens.

In all stages of the fine grinding on all the surfaces of the test plate and components we have made every effort to fit the lenses tightly to the template.

Because emery, which is graded in mesh sizes, is more consistent in grain size than Carbo, I prefer to do grinding with emeries. Suitable emeries are made by Bausch and Lomb and by the American Optical Co. Since the time lost in washing up and changing during use of very elaborately graduated series of emeries, such as eight sizes, tends to annul the time saved by their use, three sizes only-No. 180, No. 500, and No. 1200-will fully suffice and will produce perfect work.

This will be especially true if the micrometer is used for measuring glass removal and for controlling the old bugaboo, pitted edge. Before grinding is started, a permanent mark-nick with a file or a glass cutter-is made on the edge of the blanks. After the lenses have been brought to fit the template with No. 180, they are miked near this mark and the reading is written down. At least 0.008 of glass thickness must then be removed with No. 500 in order to wipe out all pits left by No. 180. At least 0.004" additional must similarly be removed by No. 1200.

During work on all sizes the familiar trick of inverting and working the tool on top should be made available for helping to arrive at gage fit.

In roughing out with No. 180, a long stroke-almost edge to center-may be used.

If the mark is overshot in the 500 stage, the curve may be brought closely to gage through resort to several alterations of inversion and return. No departure thicker than an average sheet of writing paper should be permitted to creep in at the second and third stages.

In all stages of emery it is most important to spread it evenly. A blob of emery in the center will dig its own hole, and a smear on the edge will turn the edge.

The emery should be watery, not thick. In the 1200 stage it should not be allowed to dry out too much. A wet mixture cuts faster and reduces the likelihood of scratching.

After the 500 stage, and again at the 1200 stage, the edges should be beveled slightly with some 500 and a small piece of glass. To eliminate any chipping of the edge the bevel should be retained throughout the 1200 stage throughout polishing.

POLISHING -Now, after many tedious hours, we are ready to polish. Let the good worker's rule be: "Never polish one side until the other side is ground.'' Thereby hangs the tale of no scratching.

The test plate must be polished first in order to give us a master reference to apply to the crown and flint curve.

In polishing we can once more refer to "A.TM.," page 85, for procedure; except that, having polished innumerable lenses both by machine and by hand, I wish to make a suggestion that will immensely benefit the amateur both in polishing time saved and in the figure produced. That is, to make our polisher 6/5 the diameter of the lens when working with lens on top of polisher, and 5/6 the diameter of the lens when working with the polisher on top.

In polishing lenses of such long radii it is not necessary to have a curve polisher base. For a base we can us flat plywood disks, heavily shellacked. Let the thickness of the pitch be approximately 3/8" and, when the pitch is shaped to curve, scratch lines into it to form about 1/2" squares about 1/16" deep. Lines this close and this shallow will flow in more quickly than a more familiar kind but will give more polishing facets. They will also facilitate the rouge flow, and keep better contact with the lens. Do not try to cut them in an absolutely even pattern, merely scratch them freehand, as they flow in.

The-935.51 curve on the plate glass test plate will be our first polishing surface, and here we can take heart in the fact that neither side of the test plate has to be completely polished out. It is necessary only that we have fairly good polish on it so that under Foucault's test it will give a reasonably bright reflection. Don't worry about pits but, if you feel that a test plate surface should be perfectly polished, then polish it perfect and accept my commendation on your perseverance.

We set up the Foucault test to find our radius of curvature and in so doing we once more digress from the familiar method, substituting for the opaque knife-edge a piece of thin glass (microscope specimen slide?) frosted on one side, and for the pinhole a 1/4" hole covered with coarse wire mesh (window screen), arranging these so that knife-edge and pinhole will move always equidistant from the lens. These are shifted to the point where the image of the pinhole is sharpest on the ground glass. With a steel tape, one end held by our partner (friend wife?), this distance is measured from the center of the test plate. My experience has been that any good steel tape will measure accurately within the tolerance to which we are working. One half millimeter is close enough. If over or under the specification, the test plate curve must be brought within the tolerance by polishing.

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.