IN A TIME OF TIMEBALLS
Introduction
The passing of time has been measured by humankind by ever increasingly sophisticated methods. These methods were closely aligned with the available technology of the day. Sun, Moon and major star movements, gave day, night, seasons, and major festivals. The sun dial and later burning candles, water draining from or filling a calibrated vessel and the hour glass showed the passing of time increasingly independent of the heavens and allowed the day to be divided into periods. As manufacturing technology developed, cogs, gears, weights and pendulums added the tick-tock sound to the passing of time, then periodically broadcast by bells, chimes, gunfire and timeballs. Miniaturisation and electronics meant vibrating crystals and atoms could measure, not only the passing of time to an unprecedented accuracy, but maintain this accuracy for a period in excess of Earth’s existence.
Timeball at Greenwich, Flamsteed House, London, UK.
Maintaining Timekeeping Devices
The 2019 article Shadwell’s Manual for the Accurate Determination of Meridian Distances by Chronometer indicated that the 1770s timekeeper (alternatively chronometer, clock or watch) of woodworker John Harrison was a turning point for marine navigation. Charles Frederick Alexander Shadwell (later Admiral Sir, 1814-1886), however, in his 1861 publication titled Notes on the Management of Chronometers and the Measurement of Meridian Distances, showed that the use of the new timekeepers for finding longitude was, for the most part, unreliable (Shadwell’s 1861 edition was a revised version of his earlier 1855 first edition). Thus, he concluded that information from voyages regarding longitude is kept in a state of perpetual fluctuation, from which it is impossible that universal precision can ever be obtained.
The aim of Shadwell’s book was to condense the dispersed information about the application of chronometers to the accurate interpretation of differences of longitude and endeavour to supply naval officers, and others entrusted with the care of chronometers, with a manual of instruction how best to use them, and how to furnish systematic results in recording the meridian distances of the several places visited during their voyages. It was slowly recognised, however, that shore based facilities at or near ports could greatly facilitate the maintenance of time as kept by ship’s chronometers.
The signalling of a specific, publicised instant of time is attributed to Robert Wauchope (1788-1862), a distinguished Royal Navy officer. His concept of dropping a timeball was trialled at Portsmouth, England in 1829, followed by the operation of a public timeball at Greenwich in 1833. The ball would be raised to the cross-trees (a nautical term for the horizontal crosspiece on a ship's mast to spread the sail; early flagstaffs on land which used signalling flags for messaging mimicked the ship’s mast design thus incorporating cross-trees; even on later timeball masts smaller versions of cross-trees can be seen) at a stated time before the signal, so that an observer would know that a signal was imminent. The time to be recorded was the moment a gap first appeared between the top of the ball and the cross-trees, as the ball was released to free fall.
In the earliest days of signalling time, noon was chosen as that was when the sun was observed to be at its highest point in the sky. It was a traditional observation understood by ship’s navigators. As instruments advanced and noon and therefore local time could be determined more accurately it was decided to separate observation and signalling. The noon observation could not be changed so the signalling was delayed to 1PM. Thus, at the noon observation a chronometer would be set and then that used to indicate 1PM for the dropping of the timeball. As is discussed below some ports signalled at other times of the day and some even more than once per day.
Reportedly Mauritius had a time signal operating in 1833, some six months before Greenwich. The apparatus consisted of a black ball on a white background that was hidden behind a white painted shutter. A flag was hoisted one hour before the proposed time signal and the shutter was raised to show the black ball against its white background. Five minutes before the proposed time the signal flag was lowered. At the published instant of time the shutter was dropped obscuring the ball. The moment the ball disappeared was thus observed as being the exact time signal. Other early timeballs were recorded as being at St Helena, off the west coast of Africa. Established by the British East India Company the service was available by 21 January 1834. At Calcutta (Kolkata) in 1835 and in 1839 at the Cape Coast Castle on the Gold Coast, now Ghana. A further report indicated that at the Cape of Good Hope around 1820, a light was eclipsed/obscured at 8PM indicating the instant of that hour to vessels.
A questionable time service, which may have seen the very first timeball operation being at the Cape of Good Hope, began following the 4 January 1833 purchase of a Flash pistol (a large brass barrelled pistol similar looking to a Flare gun today) and powder magazine for visually signalling time. Using a preset chronometer at the observatory, the Flash pistol was fired at the applicable time. Near the harbour, a signalman with the aid of a telescope, dropped the timeball when he saw the pistol's flash. While the Flash pistol exists today, among the concerns with such an operation was the lack of any warning of the events and to see the pistol flash the firing of the pistol at night would have made observing the timeball highly unlikely. More reliably was the recording of the first dropping of a timeball occurring at the Cape of Good Hope on 30 September 1836. This service followed the noting and publishing the time of the flash of a cannon at nominally 9PM.
In 1876, Edward Singleton Holden (1846-1914), then Professor of Mathematics at the United States Naval Observatory, was ordered to England. Specifically he was to attend the South Kensington (Museum, London) Loan Collection of Scientific Instruments, to examine and report on it. Holden’s report was tabled by the Secretary of the Navy in 1876 in documents of the Forty-Fourth Congress, 1876-77 (Holden, 1876). (It was reported that some 20 000 artefacts, relating to science, were exhibited during the Loan Collection of Scientific Apparatus exhibition in South Kensington from May to December 1876. It was a major cultural event attracting more than a quarter of a million visitors. Mr William Spottiswoode, Treasurer of the Royal Society, is quoted as saying : We have here brought together, not only a collection of remarkable instruments from all parts of the civilized world, and representing almost every school and period of research, but also a numerous gathering of the men who are at the present moment engaged in extending still further the range of discovery, and the practical application of its results.
Room 2 of the Loan Collection, South Kensington Museum, London, 1876
(courtesy Science Museum/Science & Society Picture Library)
While Holden’s formal instructions were vague his report clearly showed that the United States wanted to see how Time was disseminated for both marine navigation and public information.
During a stopover at Liverpool, enroute London from the United States, Holden spent time at the Observatory of Liverpool. Here he noted their systems of distributing public time, and of rating chronometers with respect to temperature.
The Liverpool observatory was responsible for dropping a timeball along with firing a timegun, thus regulating various clocks in the city and on the docks especially enabling any shipmaster to be independent of the chronometer-makers.
Over the years, the observatory had collected a large mass of statistics concerning the running of the chronometers of ships sailing out of Liverpool. Of particular interest was 1863 data on errors produced by 1 700 chronometers used by such vessels. The table below summarised these data noting that any voyages over 4 months were not included since a vessel could hardly be without means of correcting her chronometer for a much longer time than this.
Table showing error of longitude in geographical miles on the equator, deduced from 1 700 chronometers |
||||
|
Length of voyage (months) |
|||
Average error from the specified chronometers |
One |
Two |
Three |
Four |
Average error from 1 700 chronometers |
6 |
14 |
23 |
33 |
Average error from the best 10 in 100 |
0 |
0 |
1 |
1 |
Average error from the second best 10 in 100 |
1 |
2 |
3 |
4 |
Average error from the third best 10 in 100 |
1 |
4 |
6 |
8 |
Average error from the forth beet 10 in 100 |
2 |
5 |
9 |
13 |
Average error from the fifth best 10 in 100 |
3 |
7 |
12 |
17 |
Average error from the sixth best 10 in 100 |
4 |
9 |
15 |
22 |
Average error from the seventh best 10 in 100 |
5 |
11 |
18 |
28 |
Average error from the eighth best 10 in 100 |
7 |
15 |
25 |
36 |
Average error from the ninth best 10 in 100 |
9 |
24 |
41 |
61 |
Average error from the worst 10 in 100 |
25 |
62 |
101 |
143 |
These results starkly indicated the danger to which the merchant ships of Liverpool were actually subjected for many years, on account of the erroneous running of their chronometers. Not indicated in this table were vessels which were shipwrecked (on this and other accounts), thus the errors were considered to be under and not over stated.
Among the many vessels carrying chronometers assessed and whose results are included in this table, were a large number of vessels going on long voyages to India, Australia, and South America, and in many cases these vessels would necessarily be between three and four months or more on the voyage, often without sighting land. It appeared from this table that the average error to be expected on such a voyage and with such chronometers as they had (up to 1863) was no less than 33 miles!
The observatory was convinced that not only should ship’s chronometers be routinely tested but part of this testing be that their error rates be analysed with regards temperature change. To that end simple equipment had been procured which allowed chronometers to be rated at temperatures across the range that they might expect to experience at sea, being 50°F, 70°F and 85°F.
As an example, the report for a supposedly very good 1869 chronometer, contained the following results.
Rate (seconds) |
Temperature (°F) |
|
First |
+6.93 |
55 |
Second |
-3.40 |
70 |
Third |
-12.43 |
85 |
Fourth |
-3.03 |
70 |
Fifth |
+6.32 |
55 |
The chronometer was then returned to its maker and altered by him with the aim that its rate should vary as little as possible over the 30°F temperature range. On retesting the following far more acceptable results were recorded.
6 Day Period |
Rate (seconds) |
Temperature (°F) |
First |
+1.78 |
55 |
Second |
+1.05 |
70 |
Third |
+0.18 |
85 |
Fourth |
+1.12 |
70 |
Fifth |
+1.78 |
55 |
The chronometers of merchant ships were usually rated by the nautical instrument makers. In most cases the makers themselves could not determine the time with proper accuracy, and only a few were adequately informed as to how to make the best use of their imperfect instruments. Giving navigators an independent check upon the maker’s work was seen to be of great value, and the furnishing of time to the makers themselves seen as a very great and positive benefit.
The advantages of also providing a standard time, especially for inland cities, was seen to be that the hours of business and work, the starting of railway trains, etc. could be regulated. The time saved in the city of London alone, in the appointments of business men, was thought to be enormous.
The importance of providing correct time to vessels lying in sea ports, to chronometer makers and others, had led to the following information describing the dissemination of time at various places to be reported by Holden.
EUROPE
Helsingfors (Helsinki) - A timeball is dropped daily at noon at the observatory and a timegun fired, either on the guard ship or ashore.
Bergen - A timeball is dropped daily at noon by the observatory.
St Petersburg - Several public clocks in the city are controlled from the Pulkova observatory, and a timeball (or timegun) marks the instant of noon daily.
Danzig - A timeball is dropped at local mean noon, and one at Greenwich mean noon daily.
Kiel - Timeball (or timegun) for shipping.
Hamburg - Timeball (or timegun) for shipping.
Paris - The public clocks are controlled from the observatory.
Cherbourg - Paris mean noon is indicated by the dropping of a flat disc.
Lisbon - A timeball is dropped from the observatory at 1 hour Lisbon mean time, and one is dropped simultaneously on the south side of the river from the Prayal flagstaff.
Cadiz - A timeball is established at Cadiz.
Neuchatel - An extensive system of electrically controlled clocks and time signals radiates from this observatory giving the true time to the chronometer makers of the numerous factories, and for the general convenience of the public. The canton Vaud, adjoining the canton Neuchatel, has asked for an extension of this system to some of its important towns.
Berne - A system of time signals is established.
Toulon - Standard time is provided for the use of navigators.
AFRICA
The Cape of Good Hope - A timeball is dropped daily at 1 hour mean time.
Algoa Bay (Port Elizabeth) - A timeball is dropped daily at 1 hour Cape of Good Hope mean time (23h 46m 5s Greenwich Mean Time).
ASIA etc
Madras (Chennai) - A timeball is dropped daily from the custom house at 8h 20m 57.3s Madras civil time (15h Greenwich Mean Time). The time of the flash of the evening gun is noted at the observatory, and published at the master-intendant's office the next day.
Batavia (Jakarta) - A timeball is dropped daily at Batavia mean noon, and also one at 1h 7m 12.5s Batavia civil time (18h Greenwich Mean Time).
Vladivostok - There is a branch of the Russian hydrographic office here at which a naval officer obtains the time daily with a transit instrument, and keeps a clock regulated for the use of navigators.
Calcutta (Kolkata) - There is a timegun at this port.
AUSTRALIA
Melbourne - A timeball is dropped daily (Sundays excepted) at 1 hour mean time. The time is also given daily at 8 hours by the obscuration of a powerful light at two minutes before 8h; the instant of re-appearance of the light is the true time. The errors of these two signals on any day are published the next day in the newspapers.
Adelaide - A time signal is believed to be established at this port.
Sydney - The observatory drops a timeball daily.
Newcastle - A timeball is dropped by signals from Sydney.
SOUTH AMERICA
Rio De Janeiro - A timeball is dropped daily.
Buenos Aires - The observatory of Cordoba sends electric signals to Rosario and Buenos Aires for the regulation of clocks which are accessible to ship’s masters.
NORTH AMERICA
Saint Johns, New Brunswick - A timeball is dropped daily (except Sundays) at 1 hour. It is activated by a chronometer compared with a clock rated by a transit instrument.
Quebec - A timeball is dropped for the shipping at 1 hour, and a timegun fired in the city for public convenience at noon.
Kingston - Time is provided to the city and to the shipping daily from the observatory.
Ontario - For the past four years the observatory has struck its time signals using the fire alarm bells of the city.
Boston - For some years the Harvard College observatory has controlled clocks in Boston and furnished standard time to some of the railways.
Albany - The Dudley observatory furnishes standard time to the city and to the Hudson River Railway.
Pittsburgh - The city hall clock is controlled from the Alleghany observatory, and the fire alarm bells are struck at noon and every third hour. Time is furnished to the Pennsylvania Railway.
Cincinnati - The clock on the city hall is controlled by electric signals from the observatory every two seconds, and the fire alarm bells are struck at noon.
Chicago - Time is given to the city from the observatory.
Washington - A timeball is dropped daily (Sundays excepted) at noon. Clocks in the city are controlled (on Jones's system) at the Navy, State and Treasury Departments, at the Signal Office, and one is proposed for erection at the Western Union Telegraph Office. The city fire alarm bells are struck daily (except Sundays) at 7AM, noon, and 6PM.
Daily, at noon, (excepting Sundays,) a signal is sent by the observatory (by hand, not automatically) to the Washington office of the Western Union Telegraph Office, and by them distributed over such of their wires as are unemployed. Practically, this signal reaches New York, Philadelphia, etc., nearly every week day, not more than six or seven failures to transmit to New York occurring in a year. Chicago, Cincinnati, etc., receive the signals more irregularly. San Francisco receives them only at long intervals. The signals are better distributed along the seaboard (north and south) than toward the west. The sending of this signal is not obligatory upon the telegraph company, as it is sent over their wires without compensation. In the central office in New York City, various clocks about the building are practically kept right by the signals from Washington, and time is furnished to a few chronometer makers, etc.
Philadelphia - The Philadelphia Local Telegraph Company owns nearly all the lines in this city, and the manager of this company takes an intelligent interest in the question of public time. He has for some years supplied the noon signal from Washington to various establishments in the city, and to many of the private telegraph lines owned by merchants for the purposes of business.
GREAT BRITAIN
Liverpool - A timeball is dropped daily at 1 hour, which is visible to the shipping; a timegun is fired at the same hour from the Morpeth dock pier head (automatically); the public clock in the Victoria tower and various other public and private clocks are controlled by signals from the observatory.
Glasgow - The wire formerly owned by the observatory has been transferred to the government telegraphs, which distributes the time signals to the public at very low rates; for example, a clock within one third of a mile of the post office will be controlled for £1 per year, etc.
Dunecht (near Aberdeen) - Lord Lindsay's private observatory fires a timegun daily for the public convenience.
Edinburg - A timeball, visible to the shipping in the firth, is dropped from the top of Nelson's monument on the Calton Hill, a timegun is fired (automatically) at the castle, and several clocks are electrically controlled, some of them being visible for the convenience of the public. Time is sent daily to Dundee.
Dundee - Signals are received daily from Greenwich and Edinburg, and a timegun is fired (automatically) at 1 hour Greenwich Mean Time.
Dublin - The Greenwich time signal is received daily at 10AM and distributed.
Belfast - The Greenwich signal is received at 10AM and distributed.
Guernsey - The Greenwich time signal is received at 10AM.
Newcastle - Automatic timegun at 1 hour Greenwich time.
Sunderland - Automatic timegun at 1 hour Greenwich time.
Middlesbrough - Automatic timegun at 1 hour Greenwich time.
Kendal - Automatic timegun at 1 hour Greenwich time.
Deal - A timeball is dropped automatically at 1 hour Greenwich time.
Start Point and Portsmouth - Timeballs at 1 hour, are proposed.
Norwich - Greenwich time is distributed daily.
Stockton - Greenwich time is distributed daily.
Worcester - Greenwich time is distributed daily.
Nottingham - Greenwich time is distributed daily.
Greenwich - Timeballs : A timeball is dropped at the observatory at 1 hour daily, which is plainly visible to the shipping in the Thames. A timeball is dropped (automatically) at Deal; it has been recently proposed to drop a timeball at Portsmouth, and the Astronomer Royal suggests that one should also be established at Start Point.
Greenwich - Time signals : Sent hourly from Greenwich to the general post office. This hourly signal is transmitted to ten subscribers (mostly chronometer makers) in London. The Westminster clock on the Houses of Parliament, as well as the clock at the general post office, records its errors at Greenwich electrically. The 10AM signal is sent automatically to 21 provincial towns in England (where there are subscribers) and to Guernsey, Edinburg, Glasgow, Dublin, and Belfast. In addition to controlling the automatic sender, this signal influences a sounder, and when this is heard, the signal is sent by hand to over 600 offices in direct communication with the central one. This includes the principal railway termini. Many of these 600 offices redistribute the time signal at 10AM to the branch offices radiating from them. The 1PM signal is transmitted automatically to Newcastle, Sunderland, Middlesbrough, Kendal, Hull, Norwich, Stockton, Worcester, and Nottingham. At the first four, timeguns are fired; at the others, the current either drops a timeball, or registers itself through a galvanometer. It should be noted that the signals, as sent, are the best determinations of time possible under the circumstances, and that the final correction to the sending clock on any day is subsequently determined with great care. So much labour is expended upon the system that the Astronomer Royal considers that these hourly signals "may be used for accurate determinations of longitude." No pains are spared in the carrying out of this admirable system, and the observatory is rewarded by the very high appreciation which its labours in this direction have commanded from ship masters and owners, from the railways, and from the general public.
Holden then went on to remark upon the various systems suggested by the above data, and by the information he was able to acquire in England, insofar as they applied to the United States.
The distribution of time signals to railways, etc., is a most important matter, in which the United States is far behind England, for example, where about 500 railway stations receive a signal daily. This is partly due to the enormous extent of America in longitude, so that very different local times are used at different places of the same continuous railway line, and partly to the fact that the telegraphs are owned by the government in England, thus rendering the execution of a general system of time signals comparatively easy. In the opinion of many experienced and prominent railway officials, it is quite feasible and very desirable for all railways to be operated by one common time, and the first step toward this is plainly the certainty that the time signals which are now regularly sent from the Naval Observatory shall reach each railway station once daily, at least. This would require contracts to be entered into with the various telegraph companies, binding them to the delivery of the signals, and probably this change, which would involve considerable expense, is not at present called for. If the want is made known, ample facilities exist at the Naval Observatory, etc., for supplying it.
A timegun, if established in New York Harbor for example, would be used by a great number of vessels, which would be saved thereby the necessity of sending their chronometers to the city for rating, and thus exposing the instruments unnecessarily to the danger of a change of rate in the transportation back to the ships.
A timegun at Hampton Roads [in this context the name of the body of water that serves as a wide channel for the James, Nansemond and Elizabeth rivers between Old Point Comfort and Sewell's Point where the Chesapeake Bay flows into the Atlantic Ocean in Virginia and North Carolina], would be used by all vessels proceeding on long voyages from Baltimore, the Potomac, and Richmond, and by the large number of ships calling at Hampton Roads for orders where to carry their cargoes. It would be particularly valuable to ships using this roadstead as a harbor of refuge on their voyages, which ships at present seldom or never wait for fair weather to rate their chronometers, but on the first appearance of settled weather slip out to sea to continue their voyages.
Australian Timeballs
The above indicates that by the mid 1850s there were already some sophisticated systems in ports and cities disseminating a standard time at least once if not twice or more a day. The Australian ports of Melbourne, Sydney, Adelaide, Brisbane and Newcastle had timeball services which had replaced even earlier time signal services. The photographs below show some of Australia’s timeballs from that period.
Some Australian Timeballs
|
|
1940s photograph of Williamstown’s timeball tower, previously a lighthouse. The timeball was installed around 1861, but earlier in August 1853 a timeball, the first ever in Australia, had been dropped from a nearby flagstaff. (courtesy Museums Victoria).
|
|
The timeball at Observatory Hill, Sydney, was first dropped on 5 June 1858.
|
|
Adelaide timeball (left) and signal mast at Semaphore in 1877 (courtesy State Library of South Australia).
|
|
Brisbane Observatory and Windmill Tower (courtesy Henry Alcock, c1940s). From 1855 the tower was a signal station communicating shipping news between the entrance of the Brisbane River and the town. Renovations in 1861 saw the installation of a timeball.
|
|
In 1877 a timeball was mounted atop Newcastle's newly completed Customs House. It was activated on signals received from Sydney.
|
|
Signal station on the Round House at Perth’s Fremantle port, established 1900. |
Compiled by Paul Wise, December, 2021.
Sources
Anonymous (1876), The Press on the Loan Collection, Nature 14, Issue of 25 May 1876, pp.76-78, accessed at : https://www.nature.com/articles/014076a0.pdf
Bartky, Ian R and Dick, Stephen J (1981), The First Time Balls, Journal for the History of Astronomy, Vol.12, Issue.3, pp.155-164.
Holden, Edward Singleton (1876), Report upon the Astronomical Instruments of the Loan Collection of Scientific Instruments at the South Kensington Museum, as provided in the report of the Secretary of the Navy, November 29, 1876, in the Executive Documents of the House of Representatives, Second Session of the Forty-Fourth Congress, 1876-77, Vol.3, No.1, parts 3 and 4, Washington, 1877.
Kinns, Roger (2021), Time Signals for Mariners in South Africa, Journal of Astronomical History and Heritage, Vol.24, No.2, pp.285-314, accessed at : http://articles.adsabs.harvard.edu/pdf/2021JAHH...24..285K