2. Provide three 3/4" upright supports topped with rubber cushions to hold the base plate (I Figure 2) about 5" above the pump intake.

3. Assemble electrodes and exhaust port in position. Seal outer joints with vacuum wax. Place bottom disk in position and connect exhaust port to pump with rubber pressure tubing (G), then seal with wax.

4. Insert the release valve through a 1/16" hole cut in the rubber connector. Also, insert small plug in end of valve and seal both with vacuum wax.

5. Clamp 40-mil tungsten wire in position in electrodes, with a 3/16" overlap at middle and with slight spring action of the wire holding them together. Wrap 3" of .04" aluminum wire on this overlapping portion. Be sure to wrap evenly, and in about two layers, to provide as much metallic contact as possible.

6. Clean the glass mirror, as previously described, and clamp it securely in the three-metal-strip holder. Tinned iron from a large tin can is heavy enough for a 6" mirror support.

7. Heat the feet of the three upright supports N, and cement to the base plate I with vacuum wax. Support the mirror, face downward, on these posts, as shown in Figure 4. If properly made, this assembly forms an extremely trustworthy mirror support.

8. Carefully dust the inverted mirror with a "camel's hair" brush until no particles can be sighted when glancing across its surface at a "flat" angle.

9. Lower the bell jar or cylinder over the whole assembly and seal to the base, as indicated in Figure 3.

10. Start the pump and record the time. After pumping for five minutes, impress about 2000 volts across the filament-exhaust-port space. The initial discharge should spread itself over the metallic surfaces and gradually soften into a patternless grey-blue glow. In about 15 minutes this should occupy the whole space inside the chamber. If by chance there be a leak, no discharge will be apparent or it will not progress as described. In either case the remedy is simple, i.e., heat all wax seals gently with the soldering iron until the discharge indicates a vacuum-tight chamber.

11. After about 20 minutes from the time the pump is started, or after a possible leak has been healed, the discharge, as viewed in the dark, will cease for a 2000- or 3000- volt potential. Progressively increase this voltage, always keeping it just high enough to maintain the discharge. When the latter ceases to exist at 4000 to 4500 volts, the pressure is low enough for successful aluminizing. During the discharge period the electron bombardment of the mirror cleans away all adsorbed gas molecules which might prevent the aluminum coat from adhering properly. Although successful evaporations have been carried out after 20 minutes or less of evacuation, it is desirable to wait two hours or more to insure best results.

12. When the vacuum has been judged satisfactory, eliminate the high voltage and impress 5 volts across the filament electrodes A, A. Be sure to use heavy copper wire to carry the large current. As soon as the circuit is closed, adjust the primary coil rheostat to a point where the heated tungsten quickly melts the aluminum. Then increase the current so that the aluminum can be seen to boil vigorously. When the filament can no longer be seen through the top edge of the mirror disk, break the circuit and pull out the release valve plug.

13. Scrape away the wax around the cylinder base with a wood chisel. Then use the tip of a penknife to cut the wax seal at the cylinder-base plate junction. Jar the cylinder with a sharp slap of the hand and remove. If any difficulty is experienced, the method of increasing the inside pressure, previously described, is useful. The writer much prefers the latter.

AFTER each run the aluminum coat must be removed from the base plate between the exhaust port and the filament electrodes, otherwise the high-voltage discharge of succeeding runs would pass through this metal film instead of the low pressure gas molecules remaining in the chamber. It is a good plan to shield these areas with strips of glass which are easily removed for cleaning. The filament electrodes may also be protected from excessive coating by glass shields, as shown in Figure 3.

Since successive layers of aluminum on the metal parts tend to become porous and troublesome from the standpoint of vacuum technic, it is desirable to sandpaper the brass surfaces after each set of about five runs. It is impossible to do this to the inside walls of the exhaust port, as dirt and grit would get into the pump. To eliminate this difficulty, place a short piece of glass tubing around the exhaust port, as is illustrated in Figures 2 and 3. No deposition of aluminum will occur on the above-mentioned surfaces, with this type of shielding.

WITH regard to the matter of volts in the high-voltage discharge and the amperes of the low-voltage discharge, I have purposely avoided the use of any ammeters or voltmeters in developing the technic and have proved by the results that they are not necessary or desirable.

I knew the approximate discharge voltage and filament current voltage, simply because the transformer ratings were stamped on the instruments themselves. I later measured them with meters to see whether they were about what I thought they were, which proved to be the case. This sounds like an old-fashioned housewife's method of cooking, but was right in line with my avowed purpose to simplify the method and avoid unnecessary complications. In order to improve the situation I will, however, give a more complete set of directions on the matter of knowing when the vacuum is good t enough for aluminum evaporation.

In the high-voltage electrical circuit (Figure 5) the letters a, b, c, d, and e, represent a number of possible settings on the high resistance rheostat. If, after the vacuum pump has been operating for 20 minutes, the full voltage (5000 v.) gives a soft grey-blue discharge (the purple or reddish color having faded out), increase the resistance in the rheostat Rp until the discharge is just able to persist. Each time it extinguishes, move the variable contactor C_V along from a to b to c, etc. At the end of two, three, or four hours the discharge should extinguish with the contactor well over toward e. This means that the primary coil is getting the full 110 volts, which in turn means that the filament-B-exhaust-port F gap (Figure 2) is getting from 4000 to 5000 volts. (The impressed high voltage is to the reading of the voltmeter across the primary of the transformer as 5000 is to 110.) If the discharge ceases, when viewed in the dark at about 4000 volts, the vacuum is O. K. (night time or a dark room is best for studying the character of the high voltage discharge during the clean-up period. I have often sat watching the final clean-up of a run, and the contemplation of the beauty and soft, soothing character the discharge always gives pleasure).

The beginner should make several dry runs before trying a mirror. Place in vacuum chamber, for the test runs, small vases or other non-porous objects. When cleaned by the method described these trinkets will he more beautiful than silvered ones and will hold their polish indefinitely if handled only with cotton cloth.

I would strongly advise taking written records of each run, just as the mirror maker does when figuring. A set of five or six of these will be a gold mine when future runs are made.

THIS ends Prof. Yeagley's description. From A. F. Hoeflich, 626 16th Ave. San Francisco, Calif., we received, some time ago, the photograph shown in Figure 6. He writes: "This is photo of W. M. Grant and his aluminizing apparatus which will handle mirrors up to 14". The mercury pump is backed up by a Cenco. All equipment, gages, etc., home-grown. It turns out fine work, as is attested by members of San Francisco Telescope Makers, among whom Grant is leading authority on mirror figuring. A 6" disk is seen suspended on the tripod under the jar. This jar was being pumped to a high vacuum, after which the aluminum around the tungsten coils in the base was flashed. The finish in this demonstration before the club members was a perfectly coated mirror."

IN ATMA, page 296, are photographs of a fine dividing engine made a year or so ago by Vard B. Wallace, one of the original TNs (his picture on page 65 of the 1928 edition of ATM, was taken before his Vard Mechanical Laboratory, Inc., at 2980 East Colorado St., Pasadena Calif., was organized). He now sends the photograph shown in Figure 7, and writes:

"Here is a picture of our new evaporating outfit. The diffusion pumps are our own design and built in our own shop. The bell jar has a 12" clear inside diameter. We have striven to make this a unit piece of equipment where everything is bolted to the table with nothing at loose ends. The mechanical pump is a Cenco Hyvac and is driven by a 1/6 H. P. motor. The box below is a low-tension transformer for heating the filament. The upper box is a high-tension transformer for testing the quality of the vacuum. The large steel plate on which the bell jar rests is provided with six electrodes sealed vacuum tight. Each of these electrodes is connected to one of the six holes in the panel at the left end of the table. This arrangement permits setting as many as six filaments at one time, with the possibility of selecting the particular filament that will be fired at will. Practice has proved this arrangement very satisfactory. The push button switch on the end of the long cord permits a close control on the filament current although the operator be, moving about the jar to find the best position for watching the progress of the evaporating.

"About 17 minutes are required to pump the jar down to a black vacuum. It is possible to clean a mirror, place it under the jar, pump down, fire the coil and admit the air again in about 45 minutes.

"In contrast to the hours I have spent trying to make a good coat with silver, this method is pure delight."

AT our request Wallace also sent in a photograph (Figure 8) and description of his knife-edge testing dingbat. Of such rigs there are now many (this old-fashioned scribe still rooting for the simple hoss-and-buggy kind Ellison recommends). Wallace writes:

"The knife-edge tester was designed after looking at one used by Mr. Dalton at the Mt. Wilson Laboratories in Pasadena. A ribbon filament bulb is housed in the vertical cylinder. Light from this source passes through a pair of condenser lenses and focuses on a pinhole in a piece of foil on the side of a 1/4" prism. Just clear of the prism is the knife-edge which takes the form of an adjustable slit. This arrangement permits cutting off from either side. The separation between the pinhole and the knife edge is about l/4". This whole optical system is supported by a rider on a vertical slide that can be adjusted for altitude. A tangent screw permits the system to be rotated a few degrees either side of normal. The vertical slide is supported in a horizontal plane on straight ball races and is controlled by a micrometer screw that reads to thousandths. This, in turn, is carried on a pair of lateral ball tracks and is moved by another micrometer screw.

"This may sound more complicated than it really is. Ball-bearing ways are very little more difficult to make than are the more conventional V ways, and their operation is a delight. The feel of the screws is smooth, yet there is no shake in the whole assembly. Leveling screws support the device at three points. There is some sentiment against equipment of this degree of elaboration, but our experience indicates that it has its place. There is classical precedent for the razor blade on a stick, but the above-described machine gives just as good results, and gives them quicker. And too, it was fun to build it."

Maybe that last sentence most nearly touches bottom. However, fun is fun-so what?-Ed.

PROFESSOR Yeagley sent us, a year or so before he sent the above aluminizing article, a description of his own pet knife-edge tester and this is a good time to fish it out and publish it. He described it thus:

"A, Figure 9, is a light house containing a 2.5-volt flashlight bulb. On one side is a 1/16" circular opening for use in lining up the mirror under test. Ninety degrees from this opening is a .001" slit, set on a ground glass, for use in the Ronchi and Foucault tests. The slit is made by aluminizing a small disk of glass and making a scratch across one of its diameters.

"B is an eyepiece containing a Ronchi grating (175 wires per inch, as indicated in ATM for optimum conditions), and a knife-edge parallel to but separated from the grating wires.

"C is a rack and pinion assembly for moving the eyepiece longitudinally.

"D is a millimeter scale and vernier which estimates accurately .01 cm.

FIGURE 10 brings us back to Vard Wallace again. (This thing seems to be becoming a duet, but that's only your scribe's whim of the day. These items have accumulated over quite a long period and are now brought together at one time.) Vard Wallace next speaking:

A pattern for the base of the circular dividing engine once described in these pages [and in ATMA, page 296.-Ed.] was dusted off, and an iron casting made that proved to be an excellent table for the polishing machine. Around this was built the rest of the engine. The other castings are of aluminum alloy. Central with the vertical column is a drive shaft on which is mounted a three-step cone pulley. From this runs a V belt to the shell of a clutch on the spindle. The cone of this clutch drives the spindle, and is so adjusted that when the lever control is pulled the spindle is stopped before the driving pin is lifted out of the depression in the polishing tool. This proved a big convenience, as it is not necessary to stop the whole machine each time new abrasive is added. The overarm and spindle can be swung to any position over the work and also can be inclined to the vertical for deep curve work.

"To date, two 6" parabolas, a 10" master sphere and a master flat have been made on this machine and the results have been excellent. The real way to appreciate equipment of this sort is to rub out several parabolas on the corner of a bench by hand. Then the delight of watching the machine work for a single hour will repay all the time and labor it represents." [But when it comes to character building, there is nothing like fighting a mirror 10, 20, 50 or 100 hours, by hand, and sticking to it till you lick it. Other tasks seem easy, after that.-Ed.]

NORBIDE is the name of an abrasive offered by the Norton Company. It is boron carbide, B₄C, and is the hardest material ever produced by man for commercial uses. It is made at very high temperatures in the electric furnace, from coke and boric acid. It is harder than SiC and the makers state that it will remove stock faster than Crystalon, their name for silicon carbide. And in one pound cans it costs about 18 times as much as silicon carbide!

TWELFTH annual convention of TNs will be held at Stellafane, atop Mt. Porter, near Springfield, Vermont, Saturday afternoon, August 14. Anyone interested is welcome. Many TNs and their telescopes, to "gam" with and look at, will be there.

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.