The first thing is, I and am sure many others, would like to emulate to a certain extent those accessories normally reserved for the very few and fortunate folk in laboratories who can use extremely expensive Berek and other variable compensators. Unlike fixed compensators such as the 1/4 or full wave plate the variable compensator can provide a range of interference colours which can be useful in familiarizing oneself with the Michel-Levy chart and see for real the colours that are normally reserved for books and which are nothing short of amusing in the variations of print quality. I have several books with the Michel-Levy [birefringence] charts and they differ enormously; so much so it is difficult to know which to trust. You can also do simple experiments determining the birefringence of a mineral in thin section by canceling out the colour and measuring the figure on the compensator, this is detailed in my last article.

To allow the idea to come to fruition I used a working prototype which could be modified to include improvements as I used it in practice and found any shortcomings. The main problem for the prototype is the likelihood of the unit falling off the eyepiece tube and with the unit only weighing 10 grams but with the measuring disk to one side, the centre of gravity tends to pull it off the tube, so I included a small brass counterweight to compensate which worked well. As with my previous design I used a thin cleaved piece of mica which rotates about its axis to provide different amounts of compensation introducing the various interference colours from lower first to mid order second in distinct bands, the colour bands are much more clearly defined than my first design probably because the mica is so close to the eye. The various parts are held together by glue stick and rubber solution allowing them to be dismantled to a certain extent and re-assembled to include new ideas. Some of the parts are from an old video recorder including the steel spindle running from the mica plate to the adjustment disk and the grey plastic bearings are from transistor insulators. A possible source of mechanical parts for projects like this are discarded 3 1/2" and 5 1/4" floppy drives and CD drives from desktop computers where you will find a number of small parts from the mechanism which may work.

In  Fig 4. you can see how the prototype is built up from layers of thin card glued together to increase strength and stability. The lower disk fitting around the eyepiece tube is a precision fit to accurately centre the unit on the microscope. The cheap polarizing disk forming the microscope analyzer [available from Knight Optical UK] is temporarily glued into place on the upper ring with spots of rubber solution easily removed when necessary. The picture gives an idea of the close tolerances between the various parts and eyepiece tube. The length of plastic tube to the rotating measuring disk is calculated to allow sufficient clearance not to interfere with the face when in close proximity to the unit. A delicate touch is required when adjusting the unit so as not to displace it off the microscope. Incidentally the inverted 'L' shape black plastic holding the grey bearings is a small section cut from the cable retainer found in a very old household 13 amp plug [these normally have the two screws passing through it holding the cable into the body of the plug but sadly the more modern ones are flimsy affairs]. However in the last year or two plugs in the UK have become molded and no useful parts, it worked so well it was used in the final version.

For stability the final design main components are made from perspex which gives the unit much longer life and more stable operation than card unfortunately I have no easy way of cutting accurate larger circles in perspex so I have to do it by drilling a number of small holes neatly around the required hole size snapping it out and finishing by hand as shown at the end of the article in Fig 18. The compensator shares a number of concepts from my previous design including the measurement scale built around a small disc shown on the left hand side of Fig 5. A very rigid plastic tube holds this fixed disk in position with its calibrations allowing the steel spindle to rotate within when moving the free disk far left. The analyzer is fixed uppermost on the unit whilst a sufficiently wide window accepts the additional compensator below it without interfering with the field of view looking at the subject. End stops on the calibration disk prevent the mica rotating more than 90 degrees.

Once the perspex has been cut and filed completely round I applied a thin coat of black gloss spray paint to give a nice finish and allows the various calibrating markers shown in white Letraset decals to stand out. A thin slice of Knight Optical full wave compensator cut to the right dimensions and orientation allows an addition to the variable mica plate plate to allow measurements up to third order increasing its versatility over my previous unit.

Fig 6. shows the 2 mm thick high density foam which grips the eyepiece tube flange firmly on my stand but allows slight rotation to allow the unit to be aligned accurately in conjunction with the microscope polarizer to give optimum extinction, this could also be achieved by rotating the eyepiece tube with the unit in position if free movement is available on the stand. The mica plate is held by a small rubber cup with a sharp knife cut to accept the plate and it is not held in place by any adhesive. The additional full wave plate which is a slide fit is held by card rings cut in the correct size and orientation as shown above.

Fig 7. shows the sturdy brass pillar holding the two perspex rings apart this has a long brass screw running through it fixing the two together however a nylon washer allows free movement of the upper ring holding the analyzer and additional compensator to be swung out of position and together with the mica plate in its vertical position allows normal viewing of subjects. This picture also shows clearly the additional compensator slot in the top ring in this case end on.

Fig 8. shows the matching Knight Optical polarizer swung out of its normal position on a mounting ring with a long arm allowing rotation in the closed position beneath the condenser to provide optimum extinction with a known polarizing source such as the Zeiss Jena eyepiece unit set up N-S shown in Fig 9. Once we know that we have the polarizer accurately set for E-W and analyzer N-S measurements made with the compensator set for 45 degrees to the polarizer following standard petrological techniques. I go through this procedure every time the compensator has been off the microscope to provide optimum results only taking a few seconds to check. The home made mount shown above allows quick removal of the centre polarizer disk to take dark-field stops, rotatable colour Rheinberg rings stackable with the polarizer if desired to give different hues and plain green and blue acetates.

Fig 10. shows the home made compensator fitted to the microscope in its normal operating position, the eyepiece tube has been purposely pulled out by about 10mm to show the unit is stable without resting on the flange of the main brass gear tube assembly below but the design was based on the tube being normally closed. One important point which is not clear in Fig 10. is that the hole drilled in the base ring of the compensator must allow the eyepiece to pass through, the compensator is gripping a small knurled flange about 2mm below the eyepiece as seen in Fig 9. which is part of the design of my particular microscope. This is why it is important to look at your own microscope for some time make a prototype and adapt the compensator to fit your own needs. It was common for Victorian and Edwardian brass microscopes to have these larger flanges just below the eyepiece which is a convenient place for the unit to rest and grip.

A nice shot of the compensator is shown in Fig 11. here the analyzer can be clearly seen on the uppermost perspex disk held in position by a friction fit card ring. The rotating and measuring disk looks unnaturally large here being a macro shot and distorted perspective. The perspex rotating disk is accurately drilled to be a friction fit on the steel spindle but due to being only 1.5mm thick requires a stabilizing dense foam disk which also doubles as a comfortable and convenient point to rotate the disk.

Fig 12. zooms in on the mica showing the very close working distances between the rotating mica and the top of the eyepiece together with the lower face of the card holding the additional compensator. The 'L' shaped black plastic bearing holder carefully sawn from the cable retainer mentioned earlier can be seen together with the two grey bearings made from transistor insulators the two together giving a smooth damped feel to the rotation of the mica.