RAF Air Ministry Bubble Sextant, Mk IXB, 1943

	Contents Introduction The basics Interesting features Repairs Re-filling the bubble chamber Unsticking the 5° increase control Results The Hughes company Downloads Hit statistics
Steve Grobler Perth, Australia. 2003-2006.		View My Guestbook Sign My Guestbook

Introduction

This my World War 2 aircraft sextant. It was built around 1943. I bought it for about AUD$40 in January 2003. Before the days of GPS technology, aircraft navigators would use a bubble sextant to get a "fix" on their position by making measurements on the stars, the moon or the sun. This process is called "celestial navigation" and it is a dying skill, thanks to the advent of GPS. But it's a lot more fun, and more satisfying than using a GPS, so many people still enjoy using sextants.

It is something I have been vaguely interested in for about 10 years, but never seriously enough to warrant buying a sextant. When this one came up at the right price and with the added bonus of historical interest, I bought it without hesitation. At the time I had no idea how to use it, whether or not it worked or even if it was complete :). So it's been a lot of fun repairing it, learning how to use it, eventually taking some sights and finally plotting the my longitude and latitude coordinates.

The basics

Marine sextants use the visible horizon as the reference for the horizontal when making measurements of a star's angular "altitude" above the horizon. Light from the star S* is reflected off mirrors at I and H, to the observers eye at E. Light from the horizon also reaches the observer by passing through the un-silvered half of the mirror at H. By moving the mirror at I, the image of the star is brought into line with the horizon and the angle is read off the scale W.

Because the aircraft is moving at high speed, manoeuvring and generally bumping around, the bubble is not steady. This introduces errors that can be reduced by taking an average of several readings over a period of 1 or 2 minutes. The Mk IX bubble sextant has a built in automatic averaging device to do this. The navigator keeps the star cantered in the bubble as best he can for about 2 minutes whilst the instrument takes a reading every few seconds and works out the average.

The Mk IX sextant has the angle scale graduated in minutes of arc, with a resolution of 1 minute. Since 1 minute of arc corresponds to about 1 nautical mile (NM) on the earth's surface, it is possible to measure position to approximately 1 NM or 1.85 km. In reality, the accuracy decreases to about 2.5' if the bubble is not cantered exactly. Furthermore there is a certain amount of skill and practice involved, so that 1'-2' is probably the limit of what can be achieved. Marine sextants typically have scales with finer divisions (0.1 minute) and the real horizon (but not the boat) is much more steady than a bubble, so that more accurate measurements can be made, possibly allowing position accuracy down to about 200 m.

The processing of the angle measurement and time is quite an involved process (of which I know very little) and these days lends itself very well to computers. Ten years ago I copied a DOS BASIC program from a boating magazine to do all the calculations and plotting, but I never had an instrument to play with then. These days there are much better programs around. I found one called Navigator and that's what I have been using. A plot from the program appears later on.

Interesting features

The design of the sextant is really quite clever and it has all sorts of features to make it possible for the navigator to get sights in a bouncing, speeding aircraft at night. Two things I found particularly interesting were the averaging mechanism and the lighting system. These functions would be so easy to do today with electronics, but they had to be achieved mechanically.

The averaging mechanism is a wind-up clockwork mechanism. It has two settings to allow the averaging time, or the number of averages to be changed - I still have to work out which. When it is fully wound, the averager readout is automatically set to zero and a screen is mechanically raised so that the user can see through the sextant, thereby indicating that the instrument is ready. When the star has been positioned, pulling a trigger initiates the averaging and the sextant takes a reading every few seconds whilst the navigator continually adjusts the mirror to keep the star cantered. The averaging is all done mechanically by the clockwork mechanism. Today it would be simple do the averaging in a few lines of computer program, but the mind boggles at the intricacies, complexity and cost of achieving the result mechanically - clockwork mechanisms are full of springs, gears, ratchets and levers.

The bubble has to be illuminated so that it can be seen both during the day and at night. During the day, ambient light does the trick, but at night a tiny lamp is used. The lamp has filters to regulate light levels so that faint stars are not drowned out. The sextant has a mechanical selector device with holes of varying sizes and equipped with coloured shades to let varying amounts of light in. When the averaging process is complete, the clockwork mechanism automatically switches off the lamp and drops the screen down to cut off vision of the object.

When the sight is complete, the navigator has to immediately note the time and then read the angle off the sextant scale. The sextant has a single lamp that illuminates all the relevant items: the coarse degree scale, the fine degree scale, the minute scale, the averager scale, a small note board on the side of the instrument, and the navigators watch, worn on the inside of his left wrist. This is all done very precisely and elegantly with a single lamp, directing light through carefully positioned apertures and prisms.

Repairs

Before I could use the sextant, I had to do a few repairs:

• Refill the bubble chamber - most of the fluid had evaporated or leaked
• Replace the main mirror due to damaged silvering
• Unstick the 5° increase mechanism on the "fixed" mirror
• Unstick the clockwork lamp cut-off contacts

Re-filling the bubble chamber

I had to find out what the bubble fluid was and how to re-fill the chamber. I resorted to the internet for this information. Initially I was not sure who had made the sextant - the only clues were the letters A.M., Mark IXBM, BRIT. PATS. and 1943/44 (I discovered later that the sextant was made by Henry Hughes & Son of London). A web search brought me to a celestial navigation mailing list NAVIGATION-L@LISTSERV.WEBKAHUNA.COM which in turn pointed me to the University of Queensland Physics Museum, where I found some images and a manual for the MK IXA variant. I now had some instructions on how to use the instrument, but I still had to work out how to fill the bubble chamber.

I subscribed to the mailing list and asked the other members if anyone could help. Robert Eno, in Canada, kindly responded by emailing me a set of scanned, hand written notes and sketches made by Frank Bartlett of NSW, Australia :). The fluid turned out to be HEXANE and refilling involved de-soldering a seal on the bubble chamber, refilling with a syringe and re-soldering.

I managed to obtain 50ml of HEXANE from a small chemical laboratory. Thanks to restrictions on sale of chemicals these days, it is hard for individuals to obtain odd things like HEXANE. HEXANE is highly flammable, and I was a bit dubious about soldering a sealed chamber of flammable liquid. (HEXANE is used because it does not harm the SHELLAC seals in the chamber). So I looked up the properties on the net and found that the auto-ignition temperature is about 260°C - above the soldering temperature of about 200°C. I also tested it first by dipping the soldering iron into a spoonful, and there was no ignition. I then made sure I didn't put much too heat in when I was re-sealing. The first time I was too cautious and I didn't get a good seal. After a few days, I could not remove the bubble and it grew and grew as the fluid leaked out, so I had to re-do the whole operation. That lasted several months, but I still have a leak somewhere because I am unable to remove the bubble again. I think I might change the filling arrangement so that I can fill through a removable cap (screw) to make re-filling easier, and just live with the slow leak.

There are a few photos of the bubble assembly below, and a schematic of the bubble assembly, which I found in a 1938 Air Ministry (A.M.) publication at the Perth Aviation Museum. The bubble is produced by winding a control that moves a diaphragm, sucking fluid out of the chamber until some air is sucked in. This allows you to control the size of the bubble, and remove it when not taking sights.

Left: Bubble chamber and reflector lens assembly, with lamp and diffuser in the foreground.
Right: Bubble in the chamber!!! :) (thanks to Robert & Frank)
Bot left: Diaphragm exposed. Thankfully it was not cracked.
Bot right: Schematic of bubble system on the earlier MKVIII version of the sextant

Unsticking the 5° increase control

There are two mirrors in the sextant. One is adjusted to bring the star down to the bubble. The other remains stationary. However, the fine adjustment of the movable mirror only allows about 5° of adjustment, and the coarse adjustment is in steps of 10°. So, to ensure that the star can always be brought to the bubble, the fixed mirror inclination can be changed by 5° by moving a lever. This mechanism was jammed, so I had to strip it down to free it.

The internal workings of the sextant are full of adjustment screws and stops that are sealed off at the factory to maintain instrument calibration. I was worried about having to break these seals and destroy the calibration. As it turned out, I could strip out the mechanism without affecting any of the adjustments (at least none that I was conscious of !!). The problem was that the grease had dried and hardened on the tapered shaft that carries the mirror. The leaf spring in the mechanism was not strong enough to push the mirror firmly against the stops when the lever was operated. After cleaning and re-greasing, it worked fine.

Results

So after a getting the instrument functional, I was able to take some sun, star and planet sights and plot my position on a suitable map. I got the map from the Geoscience web site, where you can download (for free) scanned 1:250,000 topographical maps of Australia.

I also needed to know accurately what my actual position was. Without knowing that, I could not determine how good or bad my sextant fix was. I got this data from the Geographical Information System (GIS) boys at my office. They were able to provide me with the coordinates of the centre of my property here in Perth, (to 6 decimal places!).

Since time has to be known accurately, I also had to figure out exactly how fast or slow my wrist watch is. Every 4 seconds of error in the time at which each sight is made, results in about 1 NM of error, so one really needs to know GMT to an error of no more than 1 s. To do this I "synchronized" my watch to UTC time using the NIST on the internet.

So, here is the plot of the end result. My actual position is the lower arrowed yellow dot, and my calculated position is the arrowed circle above. The distance between the two is about 0.7 nautical miles or 1.25 km. To get this accuracy I used 9 Sun sights. Most of these sights had a delta (distance from the actual position) of less than 3 NM. This is about 3 or 4 times better than I have been able to achieve with star sights. I think the reason is that the stars appear so much smaller than the bubble that it is tricky to centre the star accurately. In contrast, the Sun is quite large – same sort of size as the bubble so it is easier to line up.

I can't see how navigators in the war could get a fix any better than about 10 or 15 miles in an aircraft bouncing around and flying at 300 mph! Perhaps that level of accuracy was all that was required?

The Hughes company

surviving with original first hand knowledge of the Mk IX bubble sextant and has become a bit of a legend in certain quarters. He lives in Chatswood, North Sydney. Frank's son Ian provided the following commentary on the Henry Hughes & Sons company that manufactured the Mk IX sextant.  (The chart adjacent, from the Kelvin Hughes company traces the origins back to 1750).

Ian writes:

"I  thought you might wish to know that Frank, along with Francis Chichester (the round the World sailor) were two of the handful of instrument makers who actually built these sextants at Henry Hughes' factory in Ilford, Essex, England, during WW2. Incidentally, at the age of 83 he is probably the only person in the world who is still providing a professional repair service for these sextants. BTW, he told me that all of the instrument makers at Hughes used to inscribe a special mark inside the body of the sextant to denote who actually built it. Thus it is possible that there is a one in four or five chance that your sextant was actually made by Sir Francis Chichester and it may be possible to determine its provenance by examination.

Frank told me when I was with him a month back (Dec 2004), that he is in touch with an English fellow who acts as curator for the broader aspects of Henry Hughes history. He looks after a collection of information and artefacts relating to the company throughout its history.

Regarding the company, I thought you'd like to know that whilst the original manufacturing facilities have long gone, it has evolved to become Kelvin Hughes which currently supplies marine navigation aids and instruments to professional and amateur sailors. As a recreational sailor, I can speak very highly of a number of them.

From the notes I have, the history of the company goes something like this. Henry Hughes & Sons was founded in 1838 in London as a manufacturer of chronographic and scientific instruments. Later, in 1903, the firm was incorporated as Henry Hughes & Sons Ltd. and in 1923, the company produced its first recording echo sounder, putting them into a much bigger league. In 1935, a controlling interest in the company was acquired by S Smith & Son Ltd resulting in the development of marine and aircraft instruments.

Following the London office's destruction in the Blitz of 1941, a collaboration was entered into with Kelvin, Bottomley & Baird Ltd. (of Glasgow, Scotland), resulting in the establishing of Marine Instruments Ltd. Following the formal amalgamation of Kelvin, Bottomley & Baird Ltd. and Henry Hughes & Sons Ltd. in 1947 to form Kelvin & Hughes Ltd., Marine Instruments Ltd. acted as regional agents in the UK for Kelvin & Hughes Ltd. who were essentially now a part of Smith's Industries Ltd. founded in 1944 as the successors of S Smith & Son Ltd.

Kelvin & Hughes Ltd. developed various marine radar and echo sounders, supplying the Ministry of Transport, and later Defence (MoD Navy). The firm was liquidated in 1966 but the name was continued as Kelvin Hughes, a division of Smiths Group plc. At the present time, Kelvin Hughes continues to produce and develop marine instruments for commercial and military use.

From what I remember my father telling me, bubble sextants were made in several locations during the war. One of these was the main Hughes factory near Ilford, Essex (just East of London) and in the Smith's factory at Cricklewood (in North London). The third one I cannot recall.

Downloads

• Frank Bartlett's instructions for re-filling the Mk IX bubble (168kb)

• Manual for the Mk IX sextant (2Mb)

Hit statistics