When grinding and polishing my 8-inch diameter F6 parabolic mirror for my home-made telescope, the goal was to generate a paraboloid with an accuracy of ¼ wavelength of visible light (say around 580 nanometres for yellow/green light). This should produce an acceptable image at the eyepiece of the telescope.

So, the final polishing/figuring sessions using jeweller’s rouge involved getting rid of zonal and local defects which deviated from the desired paraboloid shape by some 0.000145 mm. This would only be an “average” quality mirror, an accuracy of 1/10 lambda would be highly desirable.

I wonder how much toothpaste rubbing it would take to produce a zonal or local defect of some 0.0001 mm at the lens surface?

Cheers

Dennis