25 comments

[ 2.7 ms ] story [ 70.9 ms ] thread
With very simple tools (and quite a bit of patience) it is possible to make spherical and parabolic mirrors that deviate from the true shape by (much) less than one wavelength of light, which is rather impressive.

https://en.wikipedia.org/wiki/Foucault_knife-edge_test

the candle and knife edge needed for the interferometry are indeed quite widely available in human history, but the necessary knowledge of the wavelike behavior of light is only about 200 years old
(If the navigation is confusing, this is a subsection of chapter 3 (in the ToC sidebar). The parameters for [Lorenzo] Lotto's hypothetical telescope are derived in chapter 5, from an analysis of one of his paintings (!)).
There is of course one thing here that isn’t 15th century technology! Aluminum, which was unavailable before the early 20th century. Brass or silver are better choices (or silver plate on brass)
late 19th

early 19th, as a powder

At extreme cost
yeah, brass or speculum metal (a high-tin bronze) were much more practical options until the late 19th century when aluminum became cheap
I forget the name, but some old paintings contain a visual equivalent of Rot-13, in that there are bits which look like odd blobs when viewed normally, but resolve to an image when viewed at the right angle or with a properly shaped mirror.

Anyone remember the proper term for this, or which century it was popular?

Hans Holbein’s “The Ambassadors” [1] is the one that immediately comes to mind. It has memento mori which is heavily skewed.

[1] https://en.wikipedia.org/wiki/The_Ambassadors_(Holbein)

Wow, I'm a fan of dutch golden age art and I've never seen that. Fantastic!
You can show with fluid mechanics is that all you need is a container, molten glass, and a steady source of rotational motion while the glass cools into a parabola climbing the walls of the container, give or take a factor of two on the final shape depending on how you feel about density.
How steady? Does a massive honking flywheel suffice, or does it require control?
Height varies quadratically with radial frequency and radius, mediated by gravitational acceleration, (rv)^2/?2g. I'll venture you want a small curvature with a great deal of uniformity, the model also discounts variable viscosity with temperature and bizarre patterns that would appear with inconsistent anything, I suspect you're gonna want a control system, and probably also to shave the tails to acquire a lens compatible with the paraxial model of magnification.
I recall seeing the technique used in the Amateur Scientist column in Scientific American, with epoxy or casting resin. The verdict was the result was "maybe for infrared"... found it: https://optica.machorro.net/Optica/SciAm/LiquidMirror/1994-0...

If you search around, you'll find people have kept trying and have got somewhat better results in the meantime. Fun stuff.

https://en.wikipedia.org/wiki/Rotating_furnace discusses that.

I think most mirrors cast this way still would need treatment. For example, the mirrors of the Giant Magellan Telescope (https://en.wikipedia.org/wiki/Giant_Magellan_Telescope#Mirro...) are being cast using this method to get a rough (for modern astronomy. For example https://www.techbriefs.com/component/content/article/tb/supp...: “after the casting, the surface “roughness” is about 2.5 millimeters, or a tenth of an inch, on average. “The polishing and constant measuring are what turn this amazing piece of glass into a mirror,” he said. “By the time we finish polishing, it will be accurate to better than 25 nanometers”) shape for them. That way, there’s a lot less glass to pour, the remaining glass cools down faster, and less glass has to be removed afterwards.

(For the GMT, an additional complication is that six of the seven mirrors will be off-axis. I don’t know whether they spin them off-axis, or accept that more grinding will be necessary to give them their final shape)

Instead of glass that solidifies, you can also use a container with mercury or another reflective liquid. https://en.wikipedia.org/wiki/Liquid-mirror_telescope:

“Liquid-mirror telescopes are telescopes with mirrors made with a reflective liquid. The most common liquid used is mercury, but other liquids will work as well (for example, low-melting alloys of gallium). The liquid and its container are rotated at a constant speed around a vertical axis, which causes the surface of the liquid to assume a paraboloidal shape. This parabolic reflector can serve as the primary mirror of a reflecting telescope. The rotating liquid assumes the same surface shape regardless of the container's shape; to reduce the amount of liquid metal needed, and thus weight, a rotating mercury mirror uses a container that is as close to the necessary parabolic shape as possible. Liquid mirrors can be a low-cost alternative to conventional large telescopes. Compared to a solid glass mirror that must be cast, ground, and polished, a rotating liquid-metal mirror is much less expensive to manufacture.”

Disadvantage is that you can only point the mirror straight up. Also, there’s evaporation of the mercury.

John Dobson used this technique to great effect to grind his own parabolic mirrors to build reflector telescopes. If you're not familiar, I highly recommend reading about his "Dobsonian" telescope design and its impact on amateur astronomy.
Yes, the grinding process is surprisingly easy. My son and I met John when he was showing one of his telescopes in San Francisco one evening in the '80s, and he gave us one of his instruction leaflets and an eight inch glass from a porthole, which we ground into a mirror for our own Dobsonian telescope. He was a generous and kind, humble man; at that time he was living in an ashram in SF.
My high school physics teacher, Mr Tolby, always had a cadre of students grinding telescope mirrors by hand from glass blanks. A candle was used to measure the accuracy of them to amazing precision.

When the mirrors were done, he'd send them off to get an aluminum coating, and then the students would get a tube and have their own telescopes.

It really was remarkable how low tech it was, and what fantastic results could be achieved.

my dad still has his from a different high school physics teacher. it's always been astounding to me that a parabolic mirror is something that can be made by hand (even though of course most things throughout history were indeed made by hand).