Figure 2 is a combination photograph drawing of the parts and assembly of an ideal evaporation chamber. The top and bottom consist of 1" glass plates I, strong enough for vacuum chambers up to 14" in diameter. The 1/4 ", cylindrical wall K may be of ordinary glass, Pyrex, or metal, such as brass or aluminum. In case metal is used, a concentric cylinder J is desirable, to receive the stray aluminum (easily removed with a copper sulfate hydrochloric acid solution). A large bottle with bottom removed, as suggested by John Strong in ATMA, serves adequately for either an inner or outer cylinder. If the outer cylinder is of this type it should be completely covered with a wire screen to prevent injury in case of an implosion.

Holes in the glass base are cut with Carbo and a rotating pipe. The use of these holes for sealing in the electrodes and exhaust port obviates the need for tricky glass blown seals.

The operations entering into the making of the electrode parts A, B, C, and D, are ordinary metal turning, drilling and threading. The lead in and low voltage connection A is turned and threaded to fit into the lower drilled and tapped portion B. The L-shaped members C, C, are bent from 1/4" brass rods and held in place in the drilled holes in the upper part of B by set screws. The horizontal portions have 1/16" holes and are slit lengthwise by a hacksaw. Rings D, D, with set screws, serve to clamp the filament wires in these jaws.

The exhaust port F, preferably of brass, is also made on a lathe. The opening should be at least as large as that entering the Hyvac rotary pump, and the sides of the hole as smooth as possible, to facilitate the passage of air molecules. The metal wall should be at least 1/16" thick at every point and it is a good idea to cover the outside parts entirely with vacuum wax.

All joints, including metal-to-glass, metal-to-rubber, and glass-to-glass, are joined with vacuum wax, as indicated by cross section at L. The method of applying the wax is shown in Figure 3: an electric soldering iron is held so that the vacuum wax may be melted and guided into place by it. Care should be taken not to flow the wax between surfaces, as in metallic soldering, since this prevents easy separation o the parts later. It is desirable to build u a good backing of wax around the electrodes and exhaust port joints, since they become slightly warm during the high voltage cleanup. This prevents an undesirable flow of the wax into the joint. If leaks are present after evacuation is started, the remedy is always the same; i. e., simply warm the wax joints with the soldering iron until the leak ceases to exist. The writer has always succeeded in quickly healing leaks (which seldom occur) in making some 200 evaporations with this technic.

To open any of the seals the wax is easily scraped away with a wood chisel and penknife. To free the main cylinder entirely, after atmospheric pressure has been restored inside, force more air in through the release valve opening, either with the lungs or a small air pump, and then force a sharp wooden wedge under the cylinder-base plate joint. A slow, stead push is more desirable than strong-arm methods, for obvious reasons. The remove wax may be used over and over again.

The filament is of new, overlapping, a adjustable type. The older type helical and sine wave filaments are costly and require hot winding and bending of the tungsten wire. The single type will largely eliminate these features but often permits the molten aluminum bead to travel up on side and freeze. This effect is due to occasional greater heating on one side of the molten ball, which causes a relatively low surface tension on that side, permitting to be pulled up the opposite side in spit of weight.

The filament shown at E is somewhat similar to the electrode arrangement in carbon arc light. Two straight pieces tungsten wire are clamped in the holder C, D, so that they overlap 3/16" at the mid point between the electrodes. Three inches of .04" aluminum wire are carefully wound on this overlapping portion. When the current is established in the filament, the overlapping portion becomes welded immediately. Because of the greater cross-sectional area, less heat is formed here, resul ing in a "cold spot" which causes the molten aluminum to hug this position tenaciousl If the aluminum is all boiled off of the filament after a run, the overlapping tips tungsten should be sanded on the contacting sides and again sprung together. This is to insure low resistance and keep the local temperature low enough to maintain the aluminum ball in position.

Successive evaporations gradually amalgamate and weaken the tungsten wherever the molten aluminum is in direct contact with it. After about 15 runs the filament will break during one of the cooling periods. The break always occurs on one side of the doubled portion, which is removed by breaking off the opposite side with tweezers. The remaining portions of wire are then released and moved toward the middle, to form a new filament. By this means a hundred or more runs may be made with two 3" pieces of tungsten. With the older types of filament a considerably greater portion of wire must be discarded when they become weakened by amalgamation. (O and P, Figure 2, may be ignored, being another type once used.)

The release valve H began life in the form of a hypodermic needle which was inserted through the rubber connecting tube G after each run. The same hole could be used and resealed with wax an indefinite number of times. Although this was an entirely satisfactory method, the valve H was evolved to take its place. The small plug, indicated in the illustration, is easily removed, reinserted and sealed as often as necessary. Either the hypodermic needle or brass valve will eliminate the need for costly and troublesome glass stopcocks.

The inverted position is best for aluminum coating small mirrors.

When mounted in the inverted position the mirror may be freed from lint and dust particles with very little danger of further contamination of this type. This insures freedom from "pin holes," provided also that the pre-cleaning by washing has been thorough. This position also eliminates the chance of molten aluminum falling on the mirror and melting a hole in its surface- an extremely remote but possible occurrence. The depth of coat is easily and accurately controlled by observing through the top edge of the disk, when the film becomes opaque.

The surfaces to be coated may be cleaned satisfactorily by scrubbing with Ivory soap and hot water for ten minutes. A soft cotton cloth makes a good wad for this purpose. Subsequent rinsing with distilled water, and firm rubbing with cotton or cotton cloth, assure a surface conditioned to take an extremely adherent coat of aluminum.

A Cenco Hyvac rotary oil pump, connected to the vacuum chamber by a short piece of large diameter rubber pressure tubing, will produce the required low pressure if leaks are absent. About 3" of 1/4" walled, 3/8" inside diameter pressure tubing is desirable. The oil in the pump should be renewed after about 15 runs. No. 30 Quaker State Oil is excellent f this work.

The necessary low pressure is indicate when 5000 volts across the exhaust port filament space ceases to produce a visible discharge when viewed in the dark. The latter is true when the distance involve is about 2" to 5" or more. The necessary voltage for the above purpose is obtainable with a neon sign transformer.

The heating current for the filament best obtained by means of a 110-10 volt heavy current transformer. Both of the transformers mentioned require the use of suitable rheostats in the primary side to vary the impressed voltages.

Hyvac Pump. Central Scientific Co., Chicago (price includes motor and base. Pump alone is $50.00 and requires 1/6 H.P. motor), $90.00.

110-5000 volt transformer, No. 721,161-2, Jefferson Electric Co., Chicago ( a neon sign transformer is suitable), $6.50.

110-10 volt transformer with secondary current rating of 40 amperes (though low ratings could be used, since it is all over in so short a time), $6.50.

Rheostat for primary of high volt. trans Chicago Apparatus Co., 1735 N. Ashland Ave., Chicago (No. 67680-2, 0- 1570 ohm .3 amps.), $6.00.

Carbon rheostat for primary of 10 transformer, Chicago Apparatus Co. (No. 67700 carbon compression rheostat), $12.00.

10" x 10" x 1" glass base plate, Pittsburgh Plate Glass Co., Pittsburgh, Pa (order with unfinished edges), $2.00.

Two pieces tungsten wire 3" long an .04" diameter, approx. (.03" to .06"), General Electric Co., Schenectady, N. Y.,

Brass stock: 2' of 3/4"; 1' of 1/4"; 4' of 1/2" and 1/2' of 1" brass rod (estimated) $1.50.

Piece rubber pressure tubing, Chicago Apparatus Co. (1 foot of No. 78880 D -5/16" inside diameter, 1/4" walls), 30 cents

Bell jar, Chicago Apparatus Co. (No 75640D 8-1/2" inside diameter, 15" high) $5.50.

Vacuum wax (vacuum sealing compound 94216), Central Scientific Co., 79 Amherst St., Cambridge A Station, Boston, Mass., per can, .70.

Total: $131.50.

Note: This list is only to suggest possible sources of the materials. Most amateur groups will be able to gather all the necessary equipment for a fraction of the amount listed.

Next month, Prof. Yeagley will give the actual, step-by-step instructions for aluminizing.

HERE are the data on designing Barlow lenses, by Jack Haviland, author of the chapter on O. G. design, in ATMA.

"In designing a Barlow the formula of page 227, ATMA, becomes:

"Since the V figure for crown is larger than that for flint, the numerator of this fraction will be negative which, when multiplied by -F will give a positive result Thus the flint turns out to be a positive lens.

"Still keeping track of signs and substituting in the formula at the bottom of

" is positive for reasons stated above, so that when this is multiplied by -F the result is negative.

"This is the reverse of what happens in designing a positive objective. The Barlow lens is also used 'backward' with the flint facing the incident light. Of course it is probably possible to design the lens the other way with the flint negative and the crown positive. The Germans make telescope objectives with the flint leading and positive. Either would have to be reconsidered from the standpoint of spherical aberration, however. But, since we already have empirical data that give suitable corrections for spherical aberration, let's not go into that, just for the fun of turning a lens around.

"Regarding spherical aberration: for incident light, parallel or converging, a convex surface facing the light is best. For diverging light a concave surface is indicated. The general idea back of this is that the above prescriptions lead toward making the deviation a minimum; that is, the bending of the ray at incidence and emergence is more nearly the same at both surfaces.

"A Barlow designed as above will have a slightly convex surface on one side and a concave on the other. Since the flint is the one with the convex surface it is faced to the light to get the best correction for spherical aberration. This is the only reason the flint leads. It is also the reason why the crown leads in a telescope objective.

"Barlows add color, especially with short locus-f/8 or less-reflectors. With a long focus reflector or refractor they are not needed to get power with reasonable (3/8" or longer) eyepieces. So what?"

Thinking moods of two veterans are shown in Figures 4 and 5. Philip E. Myers of San Diego caught Porter, while R. W. Munn of Pittsfield took the picture of Everest thinking how much work it will be to grind the 20" Pyrex disk before him.

IN a letter, J. D. Beardsley, of the Washington Instrument Co., present address 1718 I ("eye") St., Washington D. C., who is the author of the notes on metal mirrors, ATMA, page 67, says:

"One little point about lead laps. Lead is rather gummy, as you know, and small abraded particles of it will roll up into small balls and lift the mirror off the tool. An ordinarv 50-50 solder lap is much better and it would seem that a 90-10 mixture would be better. This incidentally is the composition of Chase 'sweat' solder, rather expensive item for a large lap, but thin film can be attached to cast iron if it is first copper plated and then tinned.

"You can suggest to the boys to try chrome green (CrO), which has bee washed, for polishing metal mirrors. Th is the stuff the commercial boys use for buffing; they say that rouge dirties stain less steel. I have used the stuff on crystal quartz and it cuts much faster than rouge. It costs more than rouge but if time is worth anything it is a money saver.

"Dr. Canfield tells me that substituting nickel for copper in speculum metal give a very elegant substance free from the reddish tinge of the copper alloys."

OLD settlings from local eyeglass opticians' shops contain rouge that have already been worked down and, after proper washing, will not scratch. Often the owners are glad to be rid of it. This tip recently reached us from Dr. S. H. Sheib of Richmond, a contributor to ATMA, just before his unfortunate death from pneumonia, late in April. We should not have said, "will no scratch," but "should not scratch."

IN ATMA, Clark tells how to make an optician's spindle, and now a dealer's descriptive literature on a factory-built vertical spindle reaches us. Price only $245. Most TN's will make Clark's kind, instead.

IT seems that we have erroneously "promoted" Dr. Gaviola, author of the chapter in ATMA beginning on page 76, to be director of the great observatory at La Plat, Argentina, giving him that title in t by-line at the head of that chapter. Mostly he now "resigns," saying there is director already, and asks readers to cross off the word "Director" after his name.

TWELFTH annual powwow of amateur will be held at Stellafane, Saturday August 14. Porter is expected. So are you.

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.