NavList:

You wrote:

"Do you know
- if exist a book describing precisely the use of this instrument,
- if exist always a manufacturer of this instrument
- any other information or comment is welcome."

 

I really doubt that anyone still manufactures instruments like this, but I would bet there are a few in storage at older observatories around the world.

 

The links you passed along were very interesting. The Hezzanith patent was very detailed, and it's clear from the description how it works. For those who didn't read it, I'll give my summary...

 

The purpose of the instrument(s) is to measure "arc error". This is a combination of eccentricity (centering error), division error, and any other errors which are fixed over time and depend on the angle observed with the sextant. It's the (presumably) fixed, correctable error tabulated on the certificates pasted to the inside of the sextant case.

 

The idea behind the described system is to insert a prism with a known angle of deflection into the light path in front of the horizon glass of the sextant or in front of the telescope such that it occupies the part of the field of view where the horizon side is normally visible. The simplest case is a 90 degree prism. You place it in front of the horizon side of the sextant's telescope, and it lets that side of the telescope see straight up instead of straight ahead. You then set the sextant arm to 90 degrees and aim it at some object. For example you could look at the sea horizon by pointing the sextant straight down. The mirrors let you see the horizon on the right side. The prism simultaneously shows the horizon on the left side. Then the procedure to measure the error is exactly the same as an ordinary index error sight. You line up the two images and read off the error. Assuming the prism is an exact 90 degree prism, the "index correction" that has just been measured using the prism is actually the "arc error" at 90 degrees (arc error plus normal IC?). In the "patented" system as described, the sextant user would be provided with a set of prisms and holders for angles of 30, 60, 90, and (optionally) 120 degrees. The prisms do not need to have exactly those angles, but they need to have their actual angles accurately measured before they may be used. By making observations with each prism, a rough correction table for the sextant's arc error could be constructed. It's a very clever idea, but if you had such a system today, would it be any more useful than ordinary sextant observations (or maybe lunars) in assessing arc error?