C19 sextant restoration

10 11 2009

Recently I returned home after a trip around the world during which I visited relatives in Britain, France and Texas. Shortly before setting out, I had secured a BU Ships Mark II sextant by Atlas Engineering of Chicago and while in Texas, I bought two more, a US Navy octant by Brandis and Sons, and a sextant from the nineteenth century. I am now busy restoring the instruments to good order, starting with the oldest.

 It has a bronze tulip pattern frame and, although it has no name, Spencer, Browning and Co (formerly Spencer Browning and Rust) made very similar or identical sextants in about 1840 to 50. As in the twentieth century, many of the component parts were standard and appear on sextants by various makers. The real heart of the sextant is its frame and divided scale and we know that these too were made by only a few makers, perhaps no more than ten in the whole of nineteenth century Britain. Thus, it is certain that many of the dozens of instrument makers whose names appears on sextants were in fact assemblers and finishers of parts made by others. We know that some makers actually did make most of the instrument and also that they were prepared to sell finished instruments for others to add their names.

As with so many old sextants, a previous owner had thought that polished bronze and brass looked better than whatever the maker had clothed it in, often black lacquer, but sometimes the bronze was chemically browned or blackened. In stripping it, it had probably been dipped in a bath of solvent. This treatment did not agree with the ivory main and vernier scales which had shrunk so that the main scale was loose and the vernier scale had cracked around the rivets which attached it to the index arm. Both had taken on a green tint.The ebony pear-shaped handle is intact apart from a fine crack, but at some stage the index arm clamp had been lost and replaced by a makeshift one fabricated from a 5/32 inch Whitworth screw and a disc of bronze. The scale magnifier had been similarly bodged together. The top part of the horizon mirror mounting was absent. The instrument was without a case.

I made a start with the angle base for the horizon mirror mount, which simply involved cutting a piece of heavy brass angle, filing it to size and shape, drilling holes in the right places and finishing the front of the angle to leave three tiny platforms opposite the tabs of the clip, yet to be made.

Copy of 100_3153

For the clip I first marked out what the finished object would look like unfolded and then cut it out of thin brass, using a jeweller’s piercing saw. The next picture shows the cut out piece in the rough state, before filing to size, bending into shape and soldering. A hole had first to be drilled for the threaded bush for the fixing screw, as it is easier to drill a small hole in thin brass when it is flat. Once the clip was bent into shape, the bush was rivetted on the inside.

100_3135

The magnifier called for some more work with the piercing saw, harder work this time, as the brass was thicker. To stay with the spirit of things, I used the front plate of a scrapped table clock. The next picture shows the cutting completed. As you can see, it is not the first time this plate has provided metal for a replacement part.

100_3147

As with any sawing, the closer you can keep to the line the less work there is to do to finish the part, but you also need to remember that putting-on tools are in short supply! With a little practice (quite a lot, really), it becomes easy to file to the lines. The secret with brass is to keep a set of files  that are not used on anything else. Use them on steel and they tend afterwards to skid uselessly over brass unless you use a lot a pressure, and then they tend to go where you don’t want them to go.

Once I had the outside filed to shape, I could put the part in the lathe to drill and bore the large hole to a size that fitted the outside diameter of a piece of thin walled tube from a scrapped Victorian something-or-other. If you haven’t got a lathe, this could at a pinch be done with piercing saw and files.

100_3148

I then glued the tubing in place. It would probably have been soft soldered in place in the nineteenth century, but I am not above using modern aids to fabrication. The lens was a scrapped field lens from an old microscope eyepiece. I had to make a piece of tubing for it and cut the 40 threads per inch internal thread using the lathe. The post also needed some attention, as the screw had been replaced by a soldered-in stud with a nut. I had to make a new washer, filing the square hole with a needle file. The washer fits over a squared section of the post and its purpose is to prevent rotational forces being transmitted to the screw and loosening it. The next picture shows the finished article with alongside it the monstrosity that it replaced.

100_3155

The new clamp screw was a fairly straightforward bit of turning and knurling, except for the thread which had to be 5/32 inch Whitworth. I went metric over twenty years ago and my odds and ends of Imperial screwing tackle do not include 5/32 x 32 tpi, so I had to screwcut it in the lathe. The clamp itself also needed attention, as it did not have a spring (and was the wrong size and shape anyway). I filed it to a somewhat better shape and made a new spring by hammering a sliver of sheet brass until it work-hardened and became springy, a trick that would have been well-known to C18 and C19 clockmakers. You can just see in the next picture traces of the solder that hold the spring to the clamp .

Copy of 100_3154

I was able successfully to glue one of the splits in the ivory of the vernier scale, but first I had to remove it. Traditionally, they seem to have always been rivetted into place, not a good practice, as the rivet inevitably expands a little in the hole and ivory tends to shrink as it dries out. Add a little corrosion and the stage is set for splits to develop. The rivets had rounded heads, so I made a little jig out of a stub of steel bar with a hole one end that just fitted over the head of the rivet with a through hole for a drill of the same (carefully measured) size as the shank of the rivet. Using the jig with a block of wood sawn to an angle of 20 degrees to support the index arm, I could drill through, confident that the drill would go through the centre of the rivet and cut off its head, without wandering. I elected to tap the resulting holes in the index arm 10 BA and to use screws to re-attach the repaired scale, partly because I have no tiny rivets, but mainly to avoid the very problem I was trying to cure. The next photo shows the finished result. You can see that the repair has been quite successful at the zero end.

Copy of 100_3160

I finished by stripping down the instrument to the last tiny screw, cleaning everything with an old toothbrush in a 50% solution of ammonia and washing up liquid, polishing the screw heads,  and painting the individual parts.

In the past, I have not been entirely happy with the appearance given by modern paints on antique instruments. Modern spray paints give a result that is almost too good and the paint film seems to be too thick. I mentioned this to an engineer friend when I called into his workshop to pick his brains about drilling out the rivets and he recommended a spray-on protective lacquer called CRC Black Zinc (also available in a variety of other colours). It sticks to bare metal without a primer, once cured it is tough and resists scratching and, best of all, gives an effect that pleases me. Take a look at the final appearance of the sextant and, as the TV shows say, you be the judge.

Copy of 100_3159

Once I have finished restoring the other sextants, I plan to try my hand at making  a wedge-shaped case out of my precious stock of African mahogany. If I make a mess of the bowed front, I will still have two sides to use in a square box.

Advertisements

Actions

Information

2 responses

10 11 2009
george

Bill, I have two questions.
1. Are you able to estimate the magnitude of the reading-error caused by those remaining cracks in the Vernier? Is there a case, if that’s serious (and I suspect it probably is) for replacing that Vernier with a new-machined and -divided strip, inauthentic though that may be? Ivory makes a lovely clear surface to divide and to read, but it does seem to be prone to such cracking.
2. Your magnifier is indeed a great improvement. There’s a difficulty with all such magnifiers, in that the radius of the arm they swing on is a lot less than the radius of the scale they are examining. So the bit you want to look at is always (except at two positions) off the axis of the microscope. In my own Vernier sextant, I find that this hampers precise reading, and it can be done better by detaching the microscope from the arm, and using it hand-held instead. Any comments?

10 11 2009
engineernz

C19 Sextant Restoration
Thank you, George.
1. There is a large error at 19′ 30″ and 20′. If I could complete the crack, I could probably glue it and correct the error. I may do so one day. Modern Computer Aided Design programs allow one to make quite accurate vernier scales on paper provided they are reasonably “open”. I have a headstock dividing device on my lathe and it would be possible to make a new vernier scale, at the cost of a great deal of set-up time. If I ever do so, I will be sure to write a post about it.

2. I wonder whether the problem is more about spherical aberration, quite severe off-axis with the 25 mm focal length plano convex lens used on cheaper sextants. I have vernier sextants by Heath and Hughes that use Ramsden compound magnifiers which gives nice flat fields.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s




%d bloggers like this: