The Micrometer

A two-monthly review of the engineering of time

Issue 11   May - June 2018

The Hartnup balance

     Made by William Shepherd of Liverpool in the 19th Century, his marine chronometer No. 206 shown on the Home page of this issue of The Micrometer is fitted with the unusual 'Hartnup balance'.  The Hartnup balance was designed by John Chapman Hartnup (1806 to 1885) who, with the support of Sir George Airy (the Astronomer Royal from 1835 to 1881), was appointed Director of the Liverpool Observatory from 1843 until his resignation due to failing health in 1885.

     Hartnup described his balance in the 1849 Monthly Notices of the Royal Astronomical Society, Volume 9 pages 206 - 215 which includes his drawing of the style of balance he envisaged and which is essentially identical to the one fitted in the movement of chronometer No. 206 photographed above.  This same Royal Astronomical Society Notice also describes the rating trials of three Shepherd chronometers fitted with Hartnup's balance, though Shepherd No. 206 is not amongst them*.   The bimetallic balance rim was made in a conical shape, the primary purpose of which, in Hartnup's words, was "contrived that the compensating rim and weights should move towards the centre with an accelerating velocity in an increasing temperature, while in a decreasing temperature they must recede from the centre with a gradually diminishing velocity."  Based on trials on Shepherd instrument Nos. 222, 228 and 230, he demonstrated some success at improving the middle temperature error**, though again in the words of Hartnup "... the disadvantage... is, so far as I know, merely in the additional trouble of its construction.  Mr Shepherd thinks this requires at least twice the labour of the ordinary balance."


     The full text of Mr Hartnup's comments can be found courtesy of the Royal Astronomical Society, SAO/NASA Astrophysics Data System and Harvard University at


* That Shepherd and Hartnup worked together on the development of this balance is in little doubt, Shepherd agreeing to make up the balances for Hartnup on condition that in the event of the results being successful they should be made available to the public.


** Middle Temperature Error (MTE): Cylindrical bimetallic balance rims reduced errors due to temperature variation but, because of the variation of the modulus of elasticity varies linearly with temperature, the moment of inertia of the balance varies in proportion to the square of the radius.  It was this middle temperature error that Hartnup sought to reduce.  There is still debate about the true source and magnitude of the MTE, which has been the subject of some recent discussion, and for those interested in a better understanding, the summary by David Boettcher at will provide greater enlightenment.

A 'Furnishing piece'

     Fitted with a Japy Frères movement, the four-glass clock illustrated dates from the latter part of the 19th Century.  Previous repairer's ministrations included 4BA threads and hexagon brass nuts which would not have been used at this time, while the relative lightweight support that the screwed rod provided for the spiral gong probably contributed to the rather feeble sound that it made on striking.   I assumed that the gong was original as, while there was a possible bell post securing screw hole in the backplate, no steady pin hole appeared to be associated with it and nor were there any tell-tale marks left by a bell post foot.  But whether the cast brass securing pad hung from the screwed rod is original I cannot say, but the bright zinc plated 6mm thread cheesehead screw seemed definitely non-original.  Whatever the maker fitted, it seems unlikely that he would have released for sale such a crude arrangement in a highly visible four-glass.

     The owner seemed oblivious to these shortcomings and was unwilling to pay for the reinstatement of something a little easier on the eye.  This may have been no bad thing as I do not know what might have been fitted; to use the rather pompous-sounding conservation terms, without some research I could not even start to ‘conjecture’ what might be appropriate for this 'artefact'.

Repairing a barometer pediment

     The ability to undertake simple wooden case repairs is always an advantage, and there now follows a brief description of just one such repair to a fine Ca. 1810 Dollond, London mercury stick barometer.  The barometer was in a distressed state from storage in a box-room, but otherwise it was worthy of restoration.

     Perhaps the most distressed component was the pediment, which is shown in the first photograph, the annotations indicating missing (lost) components.

     The first task was to remove the lacquer and old glue, after which some mahogany of a similar grain openness and colour was selected.  The new parts were sawn out by hand using a coping saw (fretsaw), copying wherever possible what remained on the other side.  The parts were then carved to shape using chisels and a craft knife, all of which need to be very sharp for such fragile cross-grain carving.  You will also notice that I deliberately snapped off the one remaining swan neck at the bad crack and re-secured it along with the other, new swan neck pediment.

     The missing mouldings were cut using a router cutter, itself made from a fly-cutter shaped to match the contour of the template.  The cutter itself was sawn and filed from a piece of 3 mm thick alloy steel gauge plate clamped by means of a loose cylindrical segment and 2BA cap screw into my purpose-made holder which fits into the ½ inch collet in my Router cutter holder (available as a download in this issue of The Micrometer).

     The cutter blade was hardened and tempered to straw, the blue colour in the photograph* being where it was heated to blue and the temper colours allowed to run towards the cutting edge before immediately arresting the tip at straw colour by quenching in oil.


* Note: this is not the actual cutter used for this job, but one made for routing a similar moulding.

     After glueing together using PVA 'Resin W' woodworking adhesive (which I find by far the best for cabinet work), the assembled pediment was finished and sanded before rubbing with a number of coats of Garnet polish (a dark French polish), staining, filling and re-sanding between coats as necessary.  The pediment was finally polished with wax polish to bring it to a shine commensurate with that on the rest of the instrument.

     A replacement finial was turned from CZ120 yellow brass to a style that research on the Internet (Google images) and the barometer publications by Edwin Banfield* suggested was appropriate for this Dollond stick barometer.  The finial was polished with Brasso and lacquered.

* Banfield published several most informative books on barometers, perhaps the most relevant for this barometer being Barometers: Stick or Cistern, ISBN 978-0948382000.  Generally these are kept in stock by horological bookseller Jeff Formby.

Deadbeat pallets

     Seemingly becoming a theme for the 2018 issues of The Micrometer, I saw no alternative to the wholesale replacement of some rather distressed Graham deadbeat pallets to a Vienna Regulator.  The pallets seemed to be made in soft, unhardenable mild steel and the wear was, to say the least, 'ferocious' as can be seen from the arrowed detail towards the centre of the photograph below.  I cannot believe these roughly-finished pallets were original, but - as is still happening today - enthusiastic clock repairers seem not to worry about steel pallet impulse and locking faces being finished in a soft condition*.

     My replacements were made in alloy steel ('gauge plate') and, as is my standard practise, the nibs heated to cherry red and quenched, leaving the upper end of the pallets in a reasonably soft (tough) condition.


* Using mild (unhardenable) steel, failing to have a heat source large enough to get the pallets to red heat, or gaily playing a heat-softening flame on pallet refacings made of blued mainspring steel in order to solder them to the worn pallet nibs, I never cease to be amazed that such practises are still published in periodicals that should know better.  Superficially seeming beautifully finished when photographed for (re-)installation in the movement their beauty is, to coin a variation of the old adage, not even skin deep.

     For interest, below are three views of the movement from which the pallets were taken which, apart from the pallets, was in remarkably good condition.  The segmented leather pad inserted into the gong hammer can be seen in the right hand photograph, as too can the four inverted key-hole slots used to hang the movement onto the backboard.  Clearly a movement made when machine-made screws were available, it is curious how the pillars are secured with screws at the backplate yet taper pins are still used at the frontplate

     I have been unable to identify the maker, but I believe it to be of German (not Austrian) manufacture, perhaps dating from the last quarter of the 19th Century.

A sad sight

     On a recent trip to Dunster on the edge of Exmoor, we spied a 'hole in the wall' which, on closer inspection, revealed a sad sight (inset, photograph below).  While this might seem just another watch and clock repair business ceasing trading, the name on the plate is something to be conjured with.  The proprietor - John Harwood - is the son of (also) John Harwood, the inventor and first manufacturer of the automatic wrist watch which he patented in the 1920s.  Those with an interest in this invention can read more about John Harwood senior and the watch at  and

     From studying the information at the links I also learned (or rather, if I am correct, deduced) that a 'bumper' watch is not one of enormous proportions as has been the recent fashion amongst gentlemen's wrist watches but one in which the winding weight oscillates between 'bumpers' rather than rotates a full 360 degrees.

Watch photo: Deutsches Uhrenmuseum/Wikimedia Commons

Need some 'Olbits'?

     I wanted some 8mm diameter HSS toolbits, so an Ebay search using these terms would surely quickly reveal whether such items were available.  But apparently they were not, while omitting 'toolbits' (or 'tool' and 'bits') resulted in all the twist drills, etc. under the sun popping up.  Hmmm; let me try Google instead - and, lo and behold, 8mm toolbits popped up.

     Google really does have a sophisticated search engine, and it managed to extract these from the supplier's not only serious mis-cataloguing of the items but also from his omission of the universally-used key key words 'High Speed Steel' (or abbreviation thereof).  And lest readers believe me not, a screen-grab appears below:

Beachy Head

Inspired by a post-war British Railways Downland Rambles poster by Adrian Allinson, we recently walked along the chalk cliffs to Beachy Head.   It was a lovely sunny day and near the top was the remains of a Lloyd's signal station (the mast in the upper right photograph).  The heyday of such signal stations was from around 1883 to the early 20th Century, from which (primarily) signalling flags were used as the means for transmitting weather forecasts to passing ships.

     Few Lloyd's Signal Stations now remain, perhaps the best preserved being on The Lizard and recently put up for sale by the National Trust for Ca. £700k.  Check out Wowhaus for details.

     Another pleasure during our visit to the South Downs was to view a few of the works of Eric Ravilious who, as a War Artist, died off Iceland during a search for a missing aeroplane in 1942.  What a pity that the Trustees' policy was to keep most locked away in order that we might enjoy the Gallery's vast acreages of emulsioned walls broken only by the occasional staff coffee machine.

    "Well, you can see them on the Web" was their cheery response, so clearly neither in need of foot-fall through their doors nor funds through their donation boxes, let me oblige them in their policy by illustrating two of his works below.

Source: Wikimedia Commons

A steam interlude

     Built in 1928 at Swindon Works, GWR 4575-class 2-6-2 Prairie Tank No. 5541 was shedded at Machynlleth (89C) from 1938 until the end of her working life in 1962.  She was saved from the breakers and currently works on the Dean Forest Railway over the former Severn and Wye Railway.

     The photograph of the locomotive waiting to depart for Barmouth and Pwllheli was taken by my Father at Dovey Junction in August 1959, the single lamp in front of the chimney indicating a stopping passenger train.

The British Watch & Clock Makers' Guild

     Mr Bahne Bonniksen (1859 to 1933) who owned a horological repair workshop employing 25 staff and course instructor at the Coventry Technical School was also the inventor of the Karrsuel watch.  Clearly a man of influence and position, Bonniksen addressed his audience at a lecture at the turn of the last Century: "It seems to me that watch makers are going downhill and becoming toy-makers... and that we should make our position one of some standing”.

     It was the spur to the founding of the BWCMG on 1st January 1907, and since it foundation membership has only been open by election to bona-fide members of the trade.  Even today its membership remains undiluted by enthusiasts and their very differing views on the purposes of an organisation.  Over the years, the BWCMG's principles have remained largely unchanged: to foster education, training and apprenticeship, maintain standards, and support fellow trade members and therefore the wider horological community.

     Now publishing an expanded bi-monthly newsletter, The TimePiece is edited by Director Jayne Hall.  Under her leadership, both The TimePiece and the BWCMG web-site have been significantly updated, and I can recommend readers check it out.  Said Jayne: “Our Directors are actively engaged across all aspects of the horological trade from clocks the size of the Great Clock at Westminster ('Big Ben') to the smallest of wrist watches.  Such professionalism ensures that our members' customers get the highest quality of service.

Next time

Full engineering drawings for a sensitive drilling attachment for the lathe tailstock.  Designed to provide rack feed pressure from finger and thumb, I hope readers will find it far neater than the standard rather gawky linkage arrangement.


Guy Gibbons

Copyright (c) 2018  G E Gibbons

The Micrometer

A two-monthly review of the engineering of time

Issue 11   May - June 2018

The Micrometer

A two-monthly review of the engineering of time

Issue 11   May - June 2018

Copyright (c) 2018  G E Gibbons