I don’t normally quote myself, but my earlier comments need correction or clarification.
The sample plate is blue, the cover plate is clear. The wort is placed between the two.
Since all light passing through the refractometer is "filtered" by the blue lens, shouldn't the blue light indicate the prism and the wort [rather than the white as you've described above]?
If so, the white is simply the color of the reticule with no light falling on it as observed by us looking through the objective lens of the refractometer.
I am mistaken. I took a good look at my refractometer with a bright light and realized that the prism itself is clear glass and that the lumen (the inside) of the refractometer is painted blue! I now believe that Kai is correct in saying that the white light is the light that gets refracted by the sample and the prism. See further explanation below.
When calibrated with 1.000 SG distilled water and using a hypothetical 1.000 SG prism it will be calibrated to 0 since no deviation (refraction of light) occurs.
When a 1.040 SG sample (10 Brix) is placed on the sample plate and the cover plate applied, the light is refracted towards normal by the sample and then refracted away from normal when it passes from the sample into the prism. This results in upward deviation of the blue line (which, I agree, is a little blurred rather than crisp).
Again, I am incorrect. It turns out that the prism must always have a higher index of refraction than the sample. As light passes from the sample into the prism, it is refracted (“bent”) towards "Normal" (defined as a perpendicular line that bisects the interface between two optically different media where the incident ray hits the interface). The angles of incidence and refraction are drawn in reference to Normal, not the interface.
As the refractive index of a sample increases (as would be the case with higher SG wort), there is less of a difference between it and the refractive index of the prism. Therefore, the light would be refracted less. This would cause the white “illuminated” portion of the reticule to increase (the ‘blue/white” interface line moves up). As the SG of the wort declines, the refractive index of the wort sample declines (but the difference in refractive index between the sample and the prism increases results in a greater amount of refraction of the light.) This results in the “blue/white” interface line moving down as the illuminated portion of the reticule decreases.
Taking this concept further, as one approaches the refractive index of distilled water, the blue/white interface line reaches zero. Given that air has a much lower index of refraction than water, the difference in refractive index is even greater, and we see no illumination of the reticule whatsoever because one of two things must be happening: either the critical angle has been exceeded and white light cannot exit the back side of the prism, or the amount of refraction is so great that white light cannot exit the back of the prism (look carefully at the prism, the exit side is pretty small and situated quite high up on the prism. White light will be refracted too far downward.)
There are some refractometer designs that use prisms with complex shapes (e.g., trapezoids and rhomboids) that utilize the principle of critical and total internal reflection, but I don't think that principle is what causes the blue line we see on the refractometer to move.
I should clarify this. The concept of Critical Angles and Total Internal Reflection (TIR) most likely do impact how the refractometer works, particularly in explaining why we see the color blue on the reticule. For angles that are less than “critical” some light is reflected and some is refracted out of the prism (all 3 sides apply). For angles that are greater than “critical,” TIR occurs (and the light is reflected).
I presume that the blue color comes from refracted white light that had previously entered the prism, then exited the base of the prism, struck the blue paint [that lines the lumen of the refractometer], which absorbed all the colors of the white light while reflecting the blue color back into the prism), and that TIR occurs between the front and bottom sides of the prism until it exits the rear side of the prism and casts its blue light upon the reticule.
Unfortunately, I cannot find a good, detailed schematic online for a good hand held refractometer, nor do I possess CAD skills. As I roughly estimate the angles of my prism, the nose end seems to be around 15 degrees, the top back end seems to be around 30 degrees, and the bottom back end seems to be around 135 degrees. Does anyone have a good explanation of how a refractometer really works?