
I have an excellent Pentax Super-Takumar 50mm f/1.4 normal lens from the late 1960's. I got it on e-bay for a relatively low price.
Just this morning, after reading
an extremely informative thread on the Manual Focus Lenses forum, I was finally able to verify that, based on my lens' serial number of
37801, it
is one of the earlier Super-Takumars with radioactive thorium glass in one of the high-index elements.
Turns out, the yellowing of these radioactive lenses is actually the reaction of the Canada balsam (a tree sap, used to cement lens elements together) when it gets bombarded by particles over a period of time.
Prolonged exposure to ultraviolet light actually
bleaches the Canada balsam (UV light is well-known to fade dyes and other polymers), so the yellow cast disappears.
Here's a comparison of the Super-Takumar with an SMC Pentax-A, which shows the slight color cast of the Super-Tak:
Super-Tak:
SMC Pentax-A:
The Super-Tak has an obviously warmer cast (which lots of people like, incidentally..) but a histogram shows that it's only about half-a-stop faster than the SMC-Pentax (which is an f/2 lens), probably due to the yellow cast absorbing some light. I don't like my f/1.4 lens suddenly becoming an f/1.8 lens, so I decided to try to bleach the yellow cast out of it.
People supposedly expose their lenses to sunlight (which is rich in UV) for weeks or months at a time, but it's rainy (typhoon season) and besides I don't have a window which sees constant daylight. Also the lens would get wet as it rains in the afternoons (and I can't keep the lens behind the window, as window glass blocks UV).
Anyway, I went to Ace Hardware to look for a black light and found this:

It's an "Omni" brand bug zapper bulb, 25 watts and 220 V, quite cheap at 70 pesos (about $1.50). It's not quite a black light, as it's an incandescent (not surprising given its rock-bottom cost!), so it has really pitiful UV output. However, I think it's still better than sunlight:
The mean solar constant is 13.67 lumens/square centimeter. A 50mm f/1.4 lens has a front element that's 36mm in diameter, which gives an area of 1,000 square millimeters or 10 square centimeters. So if the Sun's rays are fully perpendicular to the front element, at most 137 lumens of solar energy goes into the glass, not all of which is UV of course.

My bug zapper bulb is 25 W, since it's an incandescent it's quite inefficient and would only put out about 8 lumens/W or 200 lumens. Only a small fraction of that is UV light. But I can make certain that this light is
always going into the front element.
I put aluminum foil (shiny side down) over the front element of the lens, as I'll send the UV light in through the back. The discolored element is supposedly somewhere in the back.

I took a cardboard tube (actually an old piggybank that I got for exchange gift in Mozcom about five years ago) and lined the insides with aluminum foil, shiny side in:


By doing this, I'm hoping that a good fraction of the bug zapper bulb's luminous flux will end up inside the Super-Takumar, rather than being uselessly dissipated into free space.
And the
piece de resistance:
It's
The Blue Light!!!
(fortunately Hitler won't see this)
Update: 13 September 2008
What I didn't mention after I posted this is that the UV light from the incandescent bug zapper did not provide any more improvements to the color cast. That cheap bulb had just too little UV output to bleach the yellow-colored Canada balsam.
Last week though, while on one of the usual lunchtime Power Plant expeditions with Dennis, I found a fluorescent UV lamp at True Value. The shelf guy said it was used for money detectors and such. It was 6W and T-5 form factor. I figured that it would have about the same light output as a 30W incandescent bulb, but due to the more efficient UV emission of charged gas versus a hot tungsten filament, it would put out a lot more UV.
And I was not far wrong. After three nights of bathing under the relatively cool light of the UV fluorescent light (no electric fan cooling needed) the lens was obviously almost yellow-free.
Super-Tak:
SMC Pentax-A:
Super-Tak after UV treatment:
Actually the above photo is too blue (must be a white balance issue). But if it's blue, it's not yellow! (they're opposite)
Here's a photo taken with the Super-Tak after the UV treatment:
and compare with the output of the SMC Pentax-A 50/2 (a bit dimmer because I used the same shutter speed for both, and the Super Tak is an f/1.4 lens):
As Borat might say, Great Success!