Baselines for Geodetic Triangulation Surveys in Australia
Compiled by Paul Wise, August 2021
The successful early settlement of Australia depended on the release of land to new settlers. This posed a problem for successive colonial administrations as land released without a prior survey was quick but ended up later in litigation when many landholders found they owned overlapping properties. A prior survey meant that land release was delayed in the short term but ultimately meant peace of mind for landholders. It took some forty years and the arrival in 1827 of then Major (later Sir) Thomas Livingstone Mitchell (1792-1855) as Surveyor General that any systematic surveying and mapping of the new land of Australia was commenced. In fact, towards the end of Mitchell’s management he was able to say that due to the extensiveness of his surveys, any land sold since 1831 had been measured and mapped in its true place. Please refer to Figure 1 below which shows the areas discovered in Australia as at 1832.
Figure 1 : 1832 Map of the discoveries in Australia : copied from the latest M.S. Surveys in the Colonial Office / by permission dedicated to the Right Honourable. Viscount Goderich, H.M., principal Secretary of State for the Colonies, and President of the Royal Geographical Society, by his Lordship's obliged servant J. Arrowsmith [Arrowsmith, John, 1790-1873]
(courtesy National Library of Australia, nla.obj-231944045).
By this time, survey by triangulation had become the accepted method of establishing points of known coordinates in the terrain which could be used for mapping and the determination of individual property boundaries. Generally high points in the terrain were chosen to form a series of triangles, hence Triangulation. The internal angles of these triangles would then be observed with optical instruments of the day. Knowing that the angles of each triangle had to equal 180 degrees allowed a check to be kept on the quality of the angle observations (over large areas the curvature of the Earth meant that the angles of a triangle actually added up to more than 180 degrees but this spherical excess could be calculated so as to still maintain a check on the angular observations). To provide scale for a triangulation survey at least the side of a triangle near the start of the survey was measured and another near the end. Using the measured length of the first side of a triangle, all the sides of the triangles were then calculated and the calculated length then compared with the measured length near the end of the survey. The misclose between the measured and calculated length then provided an overall check on the triangulation’s accuracy. As the extent of a triangulation survey grew, further sides of triangles were required to be measured so any errors could be quickly isolated.
In practice the difficulty arose in that the angular observations had to be taken on high points to have visibility over the intervening ground. Measuring a triangle’s side with the equipment of the day, however, up and down significant slopes did not achieve an accurate length for a triangle’s side. Thus, rarely was the side of an actual triangle measured. Instead the solution was to find a relatively flat piece of ground from which high points in the terrain could be seen and from which the triangulation proper could commence. On this flat terrain a baseline was prepared and very accurately measured and from the terminals of the baseline angles were observed to points in the triangulation proper. This configuration was known as a basenet. Please refer to Figure 2 to see an indicative basenet, where points Trig 1 to 4 are on high points and part of the required triangulation, whereas the points Base north and Base south are on a plain with visibility to some of the surrounding high points of the triangulation as shown. With the necessary angular observations taken to all visible points and the distance Base north-Base south known by accurate measurement, the distances Base south-Trig 3 and Base north-Trig 3 could be calculated. In turn triangle sides Base north-Trig 4 and Trig 3 to Trig 4 and sides Base south-Trig 1 and Trig 3-Trig 1 could then be calculated. The remaining sides of the triangles of the triangulation proper could now be calculated progressively.
Figure 2 : Indicative basenet; angles to visible points in the triangulation from the baseline terminals allowed the sides of the main triangles to be calculated.
The use of triangulation in surveying came from the work by Dutch mathematician Willebrord Snell. Willebrord Snellius, born Willebrord Snel van Royen (1580-1626) was a Dutch astronomer and mathematician, known in the English speaking world as Snell. Snell in 1615 surveyed the distance from Alkmaar to Bergen-op-Zoom with a chain of triangles controlled by a baseline at the start and two check baselines at the end. Snell, however, appears to have been influenced by the earlier work of Gemma Frisius (1508-1555) and Tycho Brahe (1546-1601) dating back to the late 1500s. Between 1733 and 1740 Jacques Cassini (1677-1756) and his son Cesar Francois Cassini de Thury (1714-1784), also called Cassini III or Cassini de Thury undertook the first triangulation of the whole of France. Their work subsequently led to the publication in 1745 of the first map of France constructed on rigorous mathematical principles.
In 1785 Cassini de Thury proposed the French and English cooperate in a survey to determine an arc of parallel between the longitudes of Greenwich and Paris. The English survey was placed under the control of then General William Roy (1726-1790) founder of the Ordnance Survey in 1791, and the French survey under Dominique comte de Cassini (1748-1845), son of Cesar Francois Cassini de Thury, Pierre Francois Andre Mechain (1744-1804) and Adrien Marie Legendre (1752-1833). The triangulation network ran from London to Dover across to Calais joining with the existing French triangulation near Dunkirk to Paris. This achievement led to a geodetic survey of the United Kingdom by triangulation. The Great Trigonometrical Survey of India, as it became known, started on 10 April 1802 with the measurement of a baseline near Madras, now Chennai. Then Major William Lambton (c1753-1823) selected the flat plains with Saint Thomas Mount at the north end and Perumbauk Hill at the southern end for a baseline of 7.5 miles (about 12 kilometres). The Indian triangulation was not completed for another seventy years.
The triangulation baselines for these, and all later triangulation surveys, were thus measured with the highest possible accuracy of the day. Special wooden bars, steel rods, and steel bands were finally replaced with Invar bands (Invar is a nickel-steel alloy invented by Swiss physicist and metrologist Charles Édouard Guillaume in 1896. It was named Invar for its Invariability under extremes of heat or cold; its coefficient of expansion is 15 times less than that of steel. Guillaume was awarded the Nobel Prize for Physics in 1920 in recognition of the service he has rendered to precision measurements in physics by his discovery of anomalies in nickel steel alloys). In the early 1900s an Invar band’s calibration was given to be the equivalent of 1 millimetre in 1 kilometre (one part in 1 000 000) at temperatures from 5° to 40°C.
For the baseline for the triangulation of Ireland, Thomas Frederick Colby (1784-1852), devised a dual arrangement of brass and iron, which he called a compensation bar. The Colby compensation bar or just Colby bar, as his apparatus became known, was used for baseline measurements in all parts of the world, but were relatively expensive. Colby clamped together a bar of brass and a bar of iron both about 10 feet long. The central clamping meant any expansion/contraction of the materials due to temperature would occur at the extremities of the materials. As the two materials reacted differently to any expansion/contraction he was able to devise a method where two positions on the combined bars always remained 10 feet apart despite nominal temperature changes.
The National Mapping Office in 1953, produced on behalf of the National Mapping Council, Standard Specifications for Horizontal and Vertical Control. At this time, as triangulation was still the dominant survey method in Australia, these specifications provided an insight into the accepted Australian standards for such surveys. Of note was the following :
chains of triangulation shall not normally exceed 200 miles (300 kilometres) in length between baselines;
baselines shall be measured with an accuracy represented by a probable error not exceeding 1 part in 1 000 000 (1 millimetre in one kilometre) and an estimated actual error not exceeding 1 part in 300 000 (1 millimetre in 300 metres); and
chains of triangulation shall be so designed and executed, that after the side and angle equations have been satisfied, the probable error of the length obtained for the closing base, when computed from the starting base as an ordinary side of the triangulation shall not exceed 1 part in 100 000 (1 millimetre in 100 metres) and the actual discrepancy between the computed value and the measured value of the closing base shall not exceed 1 part in 25 000 (1 millimetre in 25 metres).
In practice, however, finding a suitable site for a baseline along the route of the triangulation every 200 miles or so, was not always possible. It would seem then that this clause was read as being more of a desirable feature of a triangulation survey rather than a mandatory requirement.
Triangulation Baseline Measuring Apparatus
- early Australian linear standards
A Surveyor’s or Gunter’s chain was probably the first linear standard in Australia. The chain made of thick steel wire, consisting of 100 main links connected together by about 300 smaller rings, was invented by Edmund Gunter (1581-1626). From about 1620 the Gunter’s chain became the instrument for distance measuring in Britain, with a length equivalent to 22 yards or 66 feet; ten chains were a furlong and 80 chains a mile. Gunter's chain then integrated two seemingly incompatible systems; the traditional English land measurements, based on the number four, and decimals based on the number 10. Since an acre measured 10 square chains (or 100 000 square links) in Gunter's system, the entire process of land measurement could be computed using measurements in links, and then converted to acres by dividing the results by 100 000. The Gunter’s or Surveyor’s chain was different from an Engineer’s chain which had a length of 100 feet.
A Gunter’s chain would have been among the equipment bought to Australia by the First Fleet in 1788 for use by Australia’s first Surveyor General, Augustus Theodore Henry Alt (1731-1815). (As the Gunter’s chain construction made it prone to distortion over time it was probable that more than one such chain was landed; at least one working chain and one other chain kept theoretically pristine to be used to periodically standardise the working chain). In May 1787, Alt had been appointed surveyor of lands for the new colony of New South Wales. In late 1791 because of ill health, Alt asked to be replaced as Surveyor General. Then Governor Arthur Phillip supported his request, and while waiting for a replacement employed Lieutenant William Dawes and David Burton to survey settlers’ farms. Lieutenant Governor Francis Grose subsequently ordered Charles Grimes to Sydney to act as deputy surveyor general in 1794. Grimes virtually performed the work of Alt, until 1803 when he finally replaced Alt as Surveyor General and in turn in 1827, Thomas Livingstone Mitchell became Surveyor General.
It was not until 24 August 1832, however, that An Act (designated 3 William IV, No.4) for establishing Standard Weights and Measures, and for preventing the use of such as are false and deficient, determined that : Whereas it is necessary to provide against the use of fraudulent weights and measures in New South Wales and its dependencies, and for that purpose to establish certain standards by which all other weights and measures may be regulated, and to prohibit the use of any other weights and measures than such as shall agree with such standards; and whereas certain weights and measures of the standard now in force and in use in the United Kingdom of Great Britain and Ireland, denominated Imperial Weights and Measures, (a schedule whereof, marked with the letter A, is hereunto annexed,) have been deposited in the Colonial Treasury, in the town of Sydney. Be it therefore enacted by His Excellency the Governor of New South Wales, with the advice of the Legislative Council thereof, that the said several Weights and Measures now deposited in the Colonial Treasury, in Sydney, as aforesaid, shall be there safely kept, and shall be, and they are hereby declared to be, the Standard Weights and Measures of New South Wales [the schedule marked with the letter A mentioned above, listed as the Standard Measures of Length; One Yard, One Foot, One Inch]. A standard set of imperial weights and measures included a standard yard (approximately 1 metre).
By 1848, in New South Wales, their standard yard had been used to establish a baseline for the calibration of Gunter’s chains then in use in that colony. More specifically a standard of 100 links...established by cutting marks in the stone flagging in front of the old Lands and Survey Office in Bridge Street, Sydney, which according to one authority, was probably correct to within half an inch or so, was prescribed in 1864 to be used for the official calibration of Gunter’s chains.
Steel bands replaced the chains in the 1880s. While more rugged than the chain the growth of cities and towns required a consistency of measurement necessitating steel band calibration facilities be more accurate and more widely available. With the adoption of steel measuring bands, a more rigorous approach to the definition of distance was now possible. While the Gunter’s chain had performed well, its vagaries in producing consistent results could no longer be tolerated. Land, especially in the cities, was now becoming too expensive to be defined by approximation. Even though the stability of the steel band far exceeded that of the linked chain a repair or bad kink could still jeopardise its legitimate length.
Figure 3 : Sydney Observatory ‘Standard Measure’ or baseline
(after 1881 publication by the Government Astronomer H.C. Russell entitled ‘Results of Astronomical Observations made at Sydney Observatory – 1877 and 1878’ and held at the Mitchell Library).
Standards of 66 feet and 100 feet with stone terminals carrying the marks showing the distances had been laid down at Sydney Observatory around 1880 following the destruction of the one chain length in the pavement at the front of the Old Lands Office, Sydney. Figure 3 above shows the location of the Standard Measure in relation to the observatory. Although, of great service to surveyors generally…the observatory standards were apparently damaged by rough usage and were later found to be 0.04 inches [about 1 millimetre] short.
In June 1863, the Melbourne Observatory began operations following its move from its original site at Williamstown. A baseline of 100 feet was established nearby in 1882-83. The Victorian 10 foot standard OI6, discussed below, was used to set out the baseline to enable the Gunter’s/Surveyor’s 66 feet/100 link and Engineer’s 100 feet chains, to be accurately calibrated. Figure 28 below shows its location in relation to the Melbourne Observatory.
The first Intercolonial Conference of Surveyors in the southern hemisphere was convened in Melbourne on 31 October 1892. Among the outcomes was that the conference resolved to recommend that the measure of length used in all Australasian surveys being the English measure of length, as provided by English Statute law, standards 66 feet and 100 feet in length, in terms of such legal standard, should be established in the principal Australasian cities and adopted as the standard of surveys in all the colonies.
Nevertheless, other wooden or metallic linear standards were generally separately obtained for baseline measurement purposes.
- brass and iron, steel and Invar to light and microwaves
In addition to his compensation bars, already described above, Thomas Colby also had two ten feet iron standard bars manufactured. These two bars were to serve as a permanent record of the length of the compensation bar and of the base measurements when undertaken at a temperature 62°F. Further, two three feet (1 yard) bars, were also manufactured.
When the Houses of Parliament in London, burnt down in 1834 the primary Imperial standard yard for England and its colonies was destroyed. Due to Colby's foresight in having copies of the standard made the primary Imperial standard yard was able to be re-established some ten years later. Thus in 1845, forty bars were cast and the one that best matched the length of the existing yard became the new Imperial Standard Yard. The next best bars were approved as Parliamentary Copies and sent to England's cities and colonies. Parliamentary Copy No.18 was received by the Colony of New South Wales, and likewise copy No.34 by the Colony of Victoria, in 1855. There is also reference to South Australia having a brass yard standard in 1840. The source of this standard is, however, unclear.
These yard standards, were used to standardise chains for local land titling and triangulation surveys but, as will be seen, it was the 10 feet standards that were the reference during the major early baseline measuring activities.
Sir Thomas Makdougall Brisbane (1773–1860), during his term as Governor of New South Wales from 1821 to 1825, established a private observatory at Parramatta. Later, in 1828, the observatory took delivery of the requisite rods and cylinders for the trigonometrical survey. This survey was for the determination of an arc of the meridian which never eventuated.
In April 1846, Sir William Thomas Denison (1804-1871) was appointed the Lieutenant Governor of Van Diemen's Land, today’s Tasmania, and resided in Hobart from 1847 to 1855. It was while in Tasmania he saw the use of wooden bars for baseline measurement for triangulation surveys, as in a public lecture in 1857, Dennison not only spoke of the Tasmanian work but also voiced the opinion that : in a country [now referring more broadly to Australia] like this, where the atmosphere though it has a great range of temperature, is generally dry, to use the simplest instrument - namely, wooden bars. This thinking was driven by the fact that although Colby bars were available, they were beyond the financial resources of any Colony to be acquired.
George Robarts Smalley (1822-1870), then Government Astronomer for New South Wales, refers to these same wooden bars in his published 1869 letter at Annexure A. In March, 1867, I proceeded to Melbourne…[and] was fortunate enough in recovering some well-seasoned pine bars belonging to the Government of New South Wales, and which appear to have been sent out some thirty years ago to the Government of Tasmania, and afterwards, through some mistake, deposited at the Lands Office, in Melbourne. As will be seen later, when used at Lake George during 1873-74, the thought to be now well seasoned wooden bars reportedly acted in a most incomprehensible manner. By now however, the iron standard was now available.
In December 1858 and March 1862, New South Wales and Victoria respectively received their 10 feet wrought iron standards. The standards were designated OI4 and OI6 (the numerical identifier 4 and 6 is often shown in literature as a suffix making it very hard to see in old documents; here the full size numeral will be used for clarity) respectively. The original 10 feet wrought iron standard, designated O1, was constructed for the United Kingdom Ordnance Survey in 1826-7. The manufacturer was the British instrument making firm, Troughton & Simms Company, formed when Edward Troughton in his old age took on William Simms as a partner in 1826. So as to keep their standard O1 pristine, and not only to be able to obtain the length of O1 in terms of the existing standard yard but also to provide a standard which had other markings for comparing with other international standards of the day, Ordnance Survey had another intermediate standard made which was designated OI1. The Australian wrought iron standard bars were copies of OI1, hence their designation as OI4 and OI6 (an OI2 was also made at that time and may have been eventually sent to Australia in 1881). The position of the markings on iron standards OI4 and OI6 is shown in Figure 4 below, with Figure 5 below showing a ten feet iron measuring rod from the period, belonging to the now State of Victoria.
Figure 4 : The markings on iron standards OI4 and OI6 (after Clarke, 1866).
Other standard bars constructed for Ordnance Survey for comparison with other international units of measure, were the Ordnance Toise, designated OT, Ordnance Metre, designated OM, and Ordnance Foot, designated OF. Butterfield (1906), compiled the following table relating most of the common European measures (note that these values will vary with modern adopted national and international values of today) :
1 toise (t.)
1.949 03 metres (m.)
0.914 399 2 metres (m.)
1.093 614 3 yards
39.370 111 3 inches
202.25 English yards
1 Rhinland perch (R. per.)
12 Rhinland feet
1033 Rhinland feet
1000 English feet
0.52724 Vienna fathom (V. fa.)
Figure 5 : Ten feet iron measuring rod (courtesy Museums Victoria).
Attention to detail, please refer to Figure 6 below, when measuring baselines with rods, meant that geodetic triangulation baselines were time consuming operations sometimes taking many months to complete. As engineering methodologies were refined whereby a thin strip (ribbon) of steel and later Invar could be manufactured, the heavier and cumbersome rods where replaced by these more portable and stable measuring apparatus for baseline measurement. With the improvement in the measuring technology, however, baselines were remeasured to ensure the most accurate results were derived for the survey. Errors arising from uncertainty in temperature measurement were always a concern – was the temperature given by the thermometer(s) that of the metallic band? After 1938 in Australia, such errors were greatly reduced by standardising the metallic measuring bands in terms of their electrical resistance instead of at a stated temperature. More on this resistance thermometry technique is available at Annexure B. Around that time, then Major Thomas Alexander Vance (1882-1959), officer commanding the Australian Survey Corps published a paper The Present Methods of the Australian Survey Corps. Among other topics, Vance’s paper described RA Survey’s approach to baselines to that time :
RA Survey aim at placing baselines about 200 to 250 miles apart and adjusting the triangulation in sections between baselines.
It is not easy to find a good base site. In addition to facilities for measurement, the base terminals must be intervisible and considerably elevated above the surrounding terrain, and an extension through well conditioned figures to the main triangles must be possible. It has been found by experience that it is more important to have good conditions for the angular work than it is to have a perfectly or nearly flat surface for the lineal measurement. Slopes up to 5° or 6° can be measured without a falling off in accuracy if proper precautions are taken. A convenient length for a baseline is from four to six miles.
Invar bands 330 feet [5 chain] by one-eighth inch wide are used. In the past these have been standardised by comparison with the 66 feet standard in the basement of the Lands Department, Sydney. It is well known that Invar bands have a tendency to alter in length as they age and during use. They have to be handled with great care, as they are easily kinked. They should be standardised immediately before and after the measurement of each base, and should be compared frequently with the field standard during the measurement. Probably the greatest error in the measurement of a baseline is due to the uncertainty of the length of the band. Owing to the difficulty and delay caused by frequently having to send the bands to Sydney for tests and to the value of the Sydney standard itself being open to criticism, the Survey Corps are importing four steel bands which are being standardised at the National Physics Laboratory, England. Steel bands are not suitable for measurements of precision, owing to their high coefficient of expansion, if the only control of their length is obtained by taking the temperature with thermometers. To overcome this disadvantage, we are using a new method whereby the bands are standardised in terms of their electrical resistance instead of in terms of temperature, the resistance of the bands being measured simultaneously with the comparison of their lengths. It is estimated that by this means they can be standardised to one part in one million, against one in 400,000 by the temperature method; and that their change in length corresponding to a difference of temperature of less than ½°F, can be measured under field conditions. It is not intended to use the steel bands in the actual measurement of the base, but to have a ready means of making frequent comparisons with the Invar field bands. Experience has also shown that the coefficient of expansion of Invar as supplied by the makers is not reliable. A band drawn from an ingot may have a different coefficient from another drawn from the same source. We found the actual value of two bands to be more than three times that given by the makers. It is necessary to determine, experimentally, the coefficient of each band.
The measurement proper is with the Invar band suspended in catenary. At least two measurements of the base must be made and if these do not agree within the specified limits, a third measurement would be necessary. The measurement must be suspended during windy or wet weather.
The working Invar band is compared [pre RA Survey’s adoption of the resistance thermometry technique] with the reference bands before and after each day's measurement. This can most conveniently be done on a temporary ground standard, established in a convenient position near the base. The bands to be compared are suspended side by side, in catenary and several comparisons are made.
Corrections that have to be applied to the base measurements are for grade, temperature, difference of scale readings, sag, deformation of catenary on slopes, variation from standard, and height above mean sea level.
Figure 6 : Baseline measurement by rods in America (courtesy Penry, 2019).
In the mid twentieth century, the ability to measure travel times of frequencies in the electromagnetic spectrum meant that the distance between two points could not only be measured accurately but the intervening terrain no longer had to be physically traversed with the measuring apparatus. Thus, not only could triangulation baselines be measured with such electronic equipment but eventually the portability of such equipment resulted in complex Triangulation being replaced by simple Traversing. The era of triangulation ended.
In May 1954, National Mapping received a Geodimeter (a name derived from GEOdetic DIstance METER) Model NASM-1, manufactured by AGA (Aktiebolaget Gasaccumulator) of Sweden. Today the term Electronic Distance Measuring (EDM) equipment covers any device that uses an electronic means to measure distance and the NASM-1 Geodimeter was not only an EDM instrument but in fact the first line-of-sight EDM instrument. The Geodimeter generated frequencies in the visible light region of the spectrum. While at the Nobel Institute of Physics in Stockholm, Sweden in the latter 1940s, Dr Erik Osten Bergstrand developed an instrument to measure the speed of light. In 1947, Bergstrand took his instrument to a 7 734 metre triangulation baseline, from the island of Lovö to Vårby near Stockholm, and obtained a measurement for the speed of light. With a value for that constant, Bergstrand now adapted his device to obtain the distance between any two intervisible points and later in 1953, the first commercial model of the Geodimeter, the NASM-1 emerged. After successful testing, in 1955 Nat Map now believed that with the Geodimeter available to measure the side of a triangle every 400 kilometres or so in the main triangulation network, there would no longer be a need for tedious baseline measurements, or the observing of the associated and time consuming basenet schemes. Almost immediately however, the arrival of the Tellurometer EDM in 1957 made triangulation obsolete. The Tellurometer generated frequencies in the microwave region of the spectrum. There is further discussion of EDM and its effects on triangulation later in this article.
Australian Triangulation Baselines 1827-1970
The earliest triangulation baselines were established under Surveyor General Mitchell in the late 1820s and through the 1930s. In this era the only Australian colony was that of New South Wales. As is discussed later in the paper the colony of New South Wales recommenced its triangulation again in 1867. By this time however, other colonies had come into being. It was only Victoria that for a while adopted a different survey methodology but eventually, they too reverted to a survey by triangulation.
At the 1912, Conference of the Director of Commonwealth Lands and Surveys, the Surveyor-General and the Government Astronomer of New Zealand, and the Surveyors-General of the States of the Commonwealth of Australia, held in Melbourne, the Surveyors General each presented a review of their Colonial and now State survey activities to date. Several States declared that their triangulation activity was complete although it did not cover the whole of their State. The major reason was that the specific States had run out of hilltops and the plains/desert country was completely unsuitable for continuing with triangulation. In addition, much of such country was also unsuitable for settlement.
With a World War coming soon after the 1912 Conference and then a Depression there was an understandable lack of triangulation activity. Activity accelerated once again when World War Two demanded accurate surveys for strategic requirements. Post war development priorities then kept triangulation activity from slowing again. Thus, the discussion of the baselines for triangulation surveys in Australia has three parts; Triangulation Baselines of the 1830s being essentially the Mitchell era up to 1867 in New South Wales; Triangulation Baselines 1840-1912 being the triangulation work reported by the respective Surveyor’s General at the 1912 conference; Triangulation Baselines 1920s-1950s as documented in reports of the National Mapping Council and described in various sources.
Following this discussion, a summary of the baselines is tabled (even with best efforts this table may still be incomplete). All geographical coordinates quoted are on the Australian Geodetic Datum 1966 (AGD66) and are the best available based on the information obtainable about the location of the baseline and/or its terminal stations. The names of the terminal stations are spelt according to modern nomenclature. From the terminal coordinates the length of the baseline was calculated in metres (rounded to the nearest metre) using Vincenty’s formula and the AGD66 spheroidal parameters. The baseline distance in metres was then converted to miles (rounded to a tenth of a mile). Where only the length of a baseline was accurately reported this Imperial length is noted and converted to metres and miles.
In the pre 1912 era some baselines were measured with special apparatus others were not. Thus, unless the measuring apparatus used on a baseline was specifically mentioned the table lists the apparatus as Unspecified although in the text the probable measuring apparatus might be given, based on the era. In the case of the South Australian baselines it was known that the 66 feet/100 link Gunther’s chain was standard equipment. This knowledge meant that if the measuring apparatus used on a South Australian baseline was not mentioned it could rightly be assumed that a 66 feet/100 link chain was used. In the post 1912 era Invar bands were the norm, although information on the length and number of bands used was not always available. In the case of RA Survey, 5 chain (330 feet) Invar bands were used in the 1920s and 1930s, for their baseline measurements. Later during the war however, when second order standard work sufficed, it appears that more durable, 300 feet Steel bands were used. These two facts are thus reflected in the table. (Note that the word band is preferred instead of tape for describing the measuring apparatus as today the word tape describes a more flexible, general purpose, measuring apparatus, material.) The Reference numbers shown thus (NN) in the text refers to the numbers in the Reference column of the table.
Individuals who are mentioned as being involved with historical and mainly pre 1912 baselines are listed alphabetically at Annexure C. The list provides a link to where more detail on the individual, if available, may be found.
In all the now major capital cities of Australia, and indeed many of the then isolated larger centres of population, the first triangulation surveys were only for town planning purposes so a single baseline was sufficient. It was only as the Colonies started to expand inland were the wider implications of a geodetic triangulation with adequate, highly accurate baselines considered. Please refer to Figure 7 below. It is these geodetic triangulation baselines that is the focus of the remainder of this article. Nevertheless, baselines used to control significant mapping surveys are discussed as appropriate.
It should also not be overlooked that the initial surveying and mapping of some of the larger islands of Australia also required the establishment of baselines. Of note were the baselines of Flinders Island, Tasmania; Montebello Island, Western Australia; Melville Island, Northern Territory and Long Island and Fraser Island, Queensland. Before being connected to the mainland two baselines were established to control the triangulation of Tasmania; at Longford, 25 kilometres south of Launceston and at Cambridge, 15 kilometres east of Hobart. It was also found that in 1890-91 a survey of Moreton Bay, where the whole coast line was traversed with theodolite and band, and the triangulation computed from a baseline 26 227 feet [some 5 miles or 8 kilometres] in length on Moreton Island. In undertaking their regional mapping responsibilities RA Survey also established baselines. Of note was the survey and mapping of New Britain where 19 third order baselines were measured with astrofixes for latitude, longitude and azimuth. (While not strictly correct RA Survey is used throughout to refer the Royal Australian Survey Corps or their predecessors.)
Figure 7 : (Centre) October 1967 photograph, courtesy of Andy Rodgers, of Klaus Leppert, Supervising Surveyor, Geodetic Branch, inspecting the Oxley Theodolite Memorial, commonly known as the Big or Giant Theodolite, erected at Booligal, New South Wales, as a memorial to Explorer and Surveyor General John Oxley, unveiled by Sir Eric Winslow Woodward KCMG, KCVO, CB, CBE, DSO (NSW Governor 1957-1965) on 8 August 1967, the simple plaque reads: Lieut. John Oxley, Surveyor General of NSW reached this point on 5th July 1817;
(Left) John Joseph William Molesworth Oxley (1784-1828, Mitchell’s predecessor as Surveyor General, 1812-1828);
(Right) More recent photograph of the Oxley Theodolite Memorial.
New South Wales to 1867
As stated above it was then Major Thomas Livingstone Mitchell as Surveyor General that became the driving force behind the triangulation surveys in New South Wales. He and his colleagues left their legacy in not only today’s New South Wales but also today’s Queensland and Victoria. Unfortunately, some of those later involved with the results of these surveys were not impressed with Mitchell’s implementation. One was George Robarts Smalley (1822-1870), then Government Astronomer for New South Wales. In an 1869 letter, Base Line for the Triangulation of New South Wales, published in The Sydney Morning Herald newspaper, Smalley wrote…In measuring a baseline, it is of the greatest importance that in future generations someone should be able to verify it; yet this is rarely possible. I may instance…Sir Thomas Mitchell’s small baseline on the Botany Sands; also, his measured check line somewhere near Lake George. Please refer to Smalley’s published letter at Annexure A.
Later, Augustus Charles Gregory (1819-1905), the first Surveyor General in Queensland, recorded in 1885 that about 1850 Mitchell began a triangulation survey of Moreton Bay and the Darling Downs (it was actually Mitchell’s assistant surveyor Robert Dixon who did the actual work as is discussed below) : but the imperfections in the measurement of the baseline on the Normanby plains by the use of hardwood bars of ten-foot lengths, the small size of the theodolites, and somewhat unfavourable selection of the trigonometrical stations,…left the work in such an imperfect state as not to be of much value (Cumpston, 1954). In 1828 and 1829, Dixon had carried out triangulation in the Camden area and near the Murrumbidgee and Molonglo Rivers.
Nevertheless…The advent of Major Thomas Livingstone Mitchell…was to transform the department and the profession of surveying in Australia. Mitchell took in the department’s disarray at a glance. There was, he noted, ‘not a single theodolite in the colony that was fit for use’ and he at once set about repairing this accurate survey instrument with its clear line of sight for measuring horizontal and vertical angles. He also gained permission for making a trigonometrical survey of the colony (Moyal, 2017).
It would appear that Mitchell, while wanting to achieve a high quality survey, felt he did not have the time to meet all the requirements of such a survey. Time consuming and also expensive tasks like the monumenting of observation stations and baselines was thus avoided. On the most prominent of the visible hill/mountain tops he had all the vegetation cleared save for one conspicuous tree which was used as the target in his theodolite. This decision in no way impacted his survey’s quality but in time physical traces of those points disappeared diminishing his surveys future value; this was the point Smalley made in 1869 mentioned above.
Mitchell’s approach has been described as a reconnaissance type triangulation survey aimed at establishing survey control over an extensive region to produce a map correctly depicting the main terrain features. His expertise in such an approach to survey and mapping came initially from his 16 years of military service where he fought in the Peninsular War (the Peninsular War (1807/08-1814) described that phase of the Napoleonic Wars fought on the Iberian Peninsula, where the French were opposed by British, Spanish, and Portuguese forces, for control of the Iberian Peninsula, the peninsula in southwestern Europe, occupied by Spain and Portugal. Arthur Wellesley, 1st Duke of Wellington, KG, GCB, GCH, PC, FRS (1769-1852) was leader of the British force and eventually won against Napoleon at the Battle of Waterloo in 1815). Rightly or wrongly then, Mitchell’s survey work left no ground marks so only descriptive evidence exists as to its locations.
In his publication Thomas Mitchell : Surveyor General and Explorer, Cumpston (1954) quoted part of a letter from Mitchell to his mother, dated 1 February 1828,…I measured a base of a mile on the smooth sandy beach at Botany Bay yesterday for the purpose of making a grand survey of the whole country...I came, as it were, accidentally on a brass plate fixed in the rock marking the first spot where Captain Cook had landed on these shores. These few lines indicate that a one mile triangulation baseline was established near today’s Kurnell where the site of Cook’s landing is recorded.
Under Secretary for Lands, AR Jones in 1952 in the Goulburn Evening Post newspaper said in his Lake George History:…It was as early as September 1828, that Mitchell reported that three trigonometrical baselines had been laid down in the Colony; two in the vicinity of Botany Bay and one at the north end of Lake George, a mile in length. In respect of the Botany Bay baselines, Foster (1985), in his book Sir Thomas Livingston [sic] Mitchell and his World 1792-1855, stated:…In April 1828 Mitchell measured two bases, each of 832 yards on the sandy shore of Botany Bay. Brock (2006) stated something similar:…Mitchell measured two baselines in April 1828 as the foundation for his triangulation survey, each of 832 yards (760.781 metres) in length, along the sands of the shore at Botany Bay.
Even though Mitchell’s and the others’ dates did not align it seemed highly unlikely that Mitchell set down a baseline in January 1828 and then two baselines later in April 1828; all for the same purpose and all at Botany Bay. Also, why two baselines of equal length? There seemed to be no advantage of going to the extra effort of making their lengths equal when the amount of following survey work would be the same no matter what the length of the baselines. The length of 832 yards was also puzzling; why not 880 yards which was half a mile (1 mile is 1 760 yards). A plausible explanation was sought and is detailed at Annexure D, A Plausible Reconciliation of Mitchell’s Botany Bay Baselines. From the Botany Bay baseline(s) (1), Mitchell used a basenet to create a more substantial baseline formed by the line between the Sydney Lighthouse and a small hill on the south side of the bay. The terminals of this baseline offered views to the peaks of Mount Jellore in the south west, various prominent peaks in the Blue Mountains to the north west, and Mount Warrawolong to the north.
Sometime in the period between February-April 1828, a baseline was measured, by Mitchell and Robert Dixon, at the north end of Lake George (2). Please refer to Figure 8 below. The baseline at Lake George was some 5 kilometres south of today’s town of Collector between today’s highway and the escarpment and was 1 mile in length. The measuring apparatus used was two tent poles of English Deal (Fir or Pine wood) about 2 inches in diameter; one 10 feet 2 inches and the other 9 feet 11 inches long.
Figure 8 : Section of Mitchell’s 1834 map showing the area where he located his 1828 1 mile baseline
(courtesy Hunter Living Histories, University of Newcastle).
In late May 1828, while Mitchell was selecting triangulation stations, he stood on Mount Jellore outside of Mittagong. From Jellore he noted that in the clear Australian atmosphere visibility of between 90 and 110 miles was achievable as he could see Mount Warrawolong, Mount Hay in the Blue Mountains and the Sydney Lighthouse. The following year, when observing at Mount Warrawolong on 14 July 1829 he saw Jellore.
Figure 9 : Section of 1: 1 000 000 scale map showing (i) the locations of Sydney Lighthouse, Mount Jellore, to its south west outside of Mittagong; Mount Hay, to its north west, north of Katoomba; and Mount Warrawolong, to its north and south west of Newcastle, forming the primary 1828-29 triangles; (ii) the site of the Botany Bay and Lake George North baselines; (iii) the 60 mile check line Oxleys Pic to Mount Wambo.
The first series of Mitchell’s triangles extended from the baseline at Lake George to Sydney Lighthouse, and was the result of observations made in June, July, and August, of 1828. Notably this triangulation extended to Mount Hay and Mount Tomah, which are the highest points of the Blue Mountains (note that the Blue Mountains’ peaks of Mount Hay, Mount Banks (previously King Georges Mount), and Mount Tomah, all mentioned in the literature are relatively close to one another so appear to have become confused in some reports).
Triangulation operations were continued across the northern portion of the colony in June, 1829. Ultimately Mount Jellore, Mount Hay and Mount Warrawolong became the principal points on the triangulation, since from them may be seen the summits of all the higher ranges of the present Colony. Please refer to Figure 9 above. The triangulation was connected to the observatory at Parramatta by a traverse with a Gunter’s chain. Stations in the triangulation could now be located by their geographical coordinates of latitude and longitude relative to that of the observatory.
At the northern extent of Mitchell’s triangulation, on the Liverpool Plains (3) in the region of today’s Quirindi, in 1831 Robert Dixon established a three mile base of verification. The measuring apparatus is not specified. Dixon’s work included a connection to the triangulation from the south. From this baseline and basenet observations, the distance between stations on Oxley's Pic (Pic being French for Peak; station today is named Wereid), south west of Quirindi, and Mount Wambo, outside of Singleton, was calculated. This same distance was also determined through the triangulation from the Lake George baseline. A comparison of the two distances showed the difference to be a few yards. The triangulation was subsequently carried northward beyond the Liverpool Plains to the region of today’s New South Wales border with Queensland.
Mitchell however, did not only rely on baselines to check the triangulation. Stations in the triangulation had calculated geographical coordinates of latitude and longitude relative to that of the Parramatta observatory. On stations in the triangulation near the parallels of 30° and 35° south, and also close to having the same calculated longitude, Mitchell observed astronomical latitude. To remove any origin inconsistencies in the different determinations of latitude, Mitchell then took the difference between two stations calculated latitudes and the same two stations astronomically derived latitudes. Over a meridian arc of some 350 miles (550 kilometres) these two differences coinciding, without any sensible difference…have been considered as affording proofs of a sufficient degree of accuracy for the present purpose.
Figure 10 : Diagram showing The Nineteen Counties mapped from the control established by Mitchell’s triangulation.
By 1834, Mitchell had prepared a map, commonly called the Map of the Nineteen Counties, based on this triangulation covering the nineteen counties then existing around Sydney. Please refer to Figure 10 above.
Among Mitchell’s surveyors were Robert Dixon and Robert Hoddle. In addition to Dixon’s work already described above, Dixon also established a baseline in 1839 south of Ipswich (4). At this time what was to become Queensland was still New South Wales, and the Moreton Bay region was seen as suitable for settlement. Later in 1849 James Charles Burnett (1815-1854) measured an approximately north-south baseline south of Jondaryan, please refer to Figure 31 further on. The later Jondaryan baseline was established in this same area except it ran approximately east-west. The Ipswich and Jondaryan baselines are discussed in more detail in the later section on Queensland. Robert Hoddle later became associated with surveys of early Melbourne and is further discussed in the Victorian section below. The locations of the baselines established during Mitchell’s administration may be viewed via this link.
The need to establish the colonial boundary between South Australia and New South Wales saw Charles James Tyers under the direction of Surveyor General Mitchell sail to Melbourne from Sydney in 1839 (at that time what was to become Victoria was known as the Port Phillip District of New South Wales and Melbourne was yet to be formally named). From Melbourne, Tyers observed a triangulation west to enable the South Australia-New South Wales border meridian at longitude 141° east, to be marked on the ground. A few years earlier in 1836, Robert Russell, the senior of a three man party, was sent from Sydney to Melbourne with specific survey instructions. Part of this party’s initial work was a small triangulation survey. The work of Tyers and Russell is discussed in more detail in the later section on Victoria.
The manuscript Trigonometrical Survey of Port Jackson : commenced as a military survey by order of General Darling and continued as civil duties permitted or required, by Lieutenant Colonel Sir TL Mitchell, Surveyor General of New South Wales, was published in 1853, and contained the first accurate maps of the area. Around that time Mitchell had written to Frederick Peel, Under Secretary for the Colonies, that in 1829, by instruction of Governor Darling he had commenced a military survey of Sydney Harbour 'a work which, amidst many civil duties I could only perform at intervals of leisure’.
This 1853 manuscript also contained two panoramic sketches based on angular observations taken at two points “A” and “B” which are shown in Figure 11 below. The angular observations at “A”, located south of Point Piper in today’s Bellevue Hill, were taken in 1829 based on the bearing to Sydney Light, set as 0 degrees; the angular observations at “B”, on Dobroyd Hill, in today’s Balgowlah Heights, were taken in 1853 based on the bearing to “A” set as 0 degrees. These views show that Sydney Light and North Head were visible from “A” as was Dawes Battery and the tower of Fort Macquarie in the region of the Botanic Gardens. There is no indication in this publication of any baselines despite it being recorded that in 1828 Thomas Florance (1783-1867) measured two baselines; one baseline located at North Head; the other in the vicinity of the Botanical Gardens. It is concluded that these two baselines were used only to control this local Port Jackson mapping survey.
Figure 11 : Section of 1857 map of Port Jackson by British Admiralty, showing the relativity of locations reportedly used by Mitchell/Florance for his/their baselines and angular observations
(courtesy National Library Australia, nla.obj-233809197).
Requiring an accurate, detailed map of Sydney, the City Commissioners in 1854 authorised a triangulation survey of the city. This triangulation was controlled by a baseline at Waterloo Swamp (today encompassing the suburbs of Waterloo and Zetland), 5 miles from Sydney, and the second located at Paddington. The Waterloo Swamp baseline of 3 250 feet was measured eight or nine times with a 100 foot chain and afterwards with 20 foot rods, tested before and after each day's work. The Paddington baseline of 13 192 feet, was measured with rods provided with adjusting screws at the ends. Development has meant that any sign of these two baselines disappeared long ago.
Despite all Mitchell’s and his colleague’s work, and perhaps because it was undertaken with what equipment was available, which was not necessarily fit for purpose, and as his survey points and baselines were not permanently marked, New South Wales started its triangulation again in 1867.
New South Wales from 1867 to 1912
In 1865, George Robarts Smalley, Government Astronomer, had written to Colonial Secretary, (later Sir) Charles Cowper (1807–1875), regarding the Sydney Observatory of which he had charge. Please refer to Smalley’s letter at Annexure A. Smalley had previously communicated with the British Astronomer Royal, George Biddell Airy (1801-1892), who is best known for his establishment of Greenwich as the location of the prime meridian. Airy, who also established parameters for the figure of the earth in the northern hemisphere suggested to Smalley that:…the province of New South Wales, and the continent of Australia generally, may be the scene of important geodetic and hydrographic operations, either for the purpose of territorial survey of a high order, or for the scientific measures of arcs of meridian and arcs of parallel. Works of this class should originate from the Sydney Observatory as starting point.
Smalley therefore proposed that:…as there has never been any triangulation of this colony upon which sufficient reliance can be placed; and it is equally certain that such a work, properly carried out, is of the highest public importance. Smalley was thus given charge of commencing the triangulation of New South Wales, starting from a new baseline. In accordance with the instructions he received dated 16 January, 1867, Smalley then met with Surveyor General, Philip Francis Adams (1828–1901). Please refer to Annexure E. Later in July 1867, Adams and Smalley met District Surveyor Edward Twynam (1832-1923) at Lake George, and after a careful examination, unanimously agreed that no better site [for a triangulation baseline] could be conveniently selected.
By June 1869, Smalley reported that he had selected a line of about 5 miles 64 chains 72 links (30 672 feet or about 5.81 miles), the ends of which were on hills supposed to be not liable to inundation. A stone column had been completed at one end, and at the other end, construction of a column was in progress. Smalley passed away in September 1970. Adams then took charge of the work but found that the country had been recently devastated by the greatest flood then ever known. The waters of the lake had risen to nearly a metre over the baseline’s northern terminal, submerging about one and a half miles of the baseline. He thus altered the direction of the baseline so that the southern terminal was now further east. However, by May 1871, Adams had reported slow progress and thus by the end of 1872 it had been determined that another baseline should be begun.
The work of the third baseline was commenced on the 2 January, 1873, this line being about half a mile east of the second line, beginning at the lowest portion of the swamp, north of Deep Creek, which was then only eighteen inches above lake level. Please refer to Figure 12 above. Arthur Charles Betts (1843-c1890) and Leonard Abington Vessey (1847–1880) undertook measuring operations during 1873, completing their work by 31 January 1874. The Lake George (16) baseline’s length was reported to be 29 286 feet (about 5½ miles or 9 kilometres). Betts and Vessey used a set of three wooden measuring bars with a nominal length of 10 feet. Smalley had recovered the well-seasoned pine bars belonging to the Government of New South Wales, from the Lands Office in Melbourne where they had errantly been deposited.
Far from performing as first Denison in 1857, and later Smalley had forecast, the wooden bars reportedly acted in a most incomprehensible manner, and for some time defied all efforts to account for certain spasmodic contractions and expansions, which caused the loss of several days' work. It turned out however that the days upon which the insular action had been observed had been preceded by a low minimum temperature during the night, and if 60 degrees was reached or a lower temperature, the bars did not recover their normal length until many hour’s exposure to a temperature of 90 degrees or upwards. After this discovery the bars were carefully put to bed every night, blankets being thrown over them, and two small kerosene lamps left burning under them every night. The wooden measuring bars were also compared every day with the New South Wales 10 feet wrought iron standard OI4, mentioned above. Such was the care taken that the possible error was estimated at one inch and three-eighths for the 5½ miles (35 millimetres or less than 4 millimetres per kilometre). The difference between the lengths of the base as found from the measurement and remeasurement was 0.542 inches (14 millimetres or 1.5 millimetres per kilometre).
While Adams was instrumental in overseeing the completion of the Lake George baseline, he had earlier established another baseline. Adams baseline was established in 1859 near the Ellerslie Triangulation station some 20 kilometres north of Albury (10) with a length of 48 chains or a little over half a mile. The measuring apparatus was not unspecified. Please refer to Figure 13 below.
It was reported in 1880 that from Lake George triangles had been laid down for a considerable distance, extending as far as Albury. Presumably this triangulation closed on the Ellerslie baseline. If so, it was the last time this baseline was incorporated as later triangulation bypassed it.
Figure 13 : The location of Ellerslie Triangulation station (green pin) as depicted in the County of Goulburn, Parish of Huon map; with insert of 8226 Walbundrie 1: 100 000 scale map showing Ellerslie Triangulation station would today be in close proximity to ELDERSLIE Homestead.
The New South Wales triangulation was also being extended west by John Sofala Chard (1853-1911). The Bullenbung baseline (17) was established by Chard with a steel band in 1875. The southern terminal was at Tollendool Hill; the northern terminal however was not identified by name. From the terminal coordinates the baseline was calculated to have been 11 454 metres (7.1 miles) long. Named after the Parish of Bullenbung, the baseline was located between Wagga Wagga and Lockhart some 40 kilometres west of Wagga Wagga. Please refer to Figure 14 below.
The location of this baseline as depicted in the section of the combined County of Mitchell, Parish of Bullenbung and County of Mitchell, Parish of Ashcroft map sheets, may be viewed via this link.
District Surveyor Edward Twynam (1832-1923, Surveyor General, or equivalent, New South Wales 1887-1901) reported to Surveyor General Philip Francis Adams on 30 September 1879, that he had completed his work measuring the Newcastle baseline (18).
In his handwritten report (Twynam, 1879), Twynam wrote:…In accordance with your instructions I have measured a baseline by means of steel riband within Australian Agricultural Company’s Estate of two thousand acres, to meet the requirements for the immediate trigonometrical survey of the environs of Newcastle, in anticipation of the general trigonometrical survey reaching that place…I found the baseline laid out by Mr Brownrigg under your direction and cleared of timber and other obstructions; the conditions of the ground are generally favourable, and the site of the base appears to be judiciously selected for securing well condition triangles…For convenient reference the base was divided by proper marks into four sections and the computation of measurement and remeasurement…from which it appears that the greatest difference in any section is 0.046 feet, whilst from the fact of the errors nearly balancing, the difference on the total measurement and remeasurement is very minute being 0.001 of a foot in a distance of upwards of 7 200 feet [some 1.4 miles]…The steel riband used comprises three lengths of 100 feet, 66 feet and 200 feet respectively…the steel ribands 100 feet and 66 feet have been tested at the Observatory…The length of the 200 feet riband was…adjusted to the length laid down on section 57-58 of the baseline; this length had originally been measured with the 100 feet riband.
The Australian Agricultural Company’s Estate of two thousand acres, was shown on the County Northumberland, Parish of Newcastle 1959 plan. The extent of the Australian Agricultural Company’s holdings was transferred to the 1892 Map of the country around Newcastle, New South Wales by Major Thomas Samuel Parrott, please refer to Figure 15 below. Today this area is covered by suburban Newcastle. The exact location of Twynam’s baseline within this area is unclear but at some point, it did cross the 1879 road to the racecourse which is possibly the road shown in the 1892 map. Presumably, like the Ellerslie baseline, the triangulation north from Sydney closed on the Newcastle baseline, and if so, it was the last time this baseline was incorporated as later triangulation bypassed it.
Figure 15 : The Australian Agricultural Company’s Estate of two thousand acres, indicated by the black boundary, on a section of the 1892 Map of the country around Newcastle New South Wales by Major Thomas Samuel Parrott
(courtesy National Library Australia, nla.obj-229953448).
The New South Wales triangulation was now being extended to Cooma in the south, Wagga Wagga in the south west; to the triangulation station Rocks, just west of Bathurst in the north west; and to Bowen Mountain to the north west of Richmond near the town of the same name. The terrain forced William Jacomb Conder’s (?-1890) triangulation network from Lake George to travel by Boorowa and Cowra to the vicinity of Bathurst where it could then turn east towards Richmond. The direct chain from Lake George to Richmond apparently gave some difficulty, with several stations abandoned, and was not completed until 1881. Conder established the 7 mile Richmond base of verification (19) in 1879-80, at Ham Common. Please refer to Figure 16 below.
The location of this baseline as depicted in the section of the combined County of Cumberland, Parish of Ham Common and County of Cumberland, Parish of Castlereagh map sheets may be viewed via this link.
An 1880 report on the Richmond baseline, stated that:…The distance was measured twice over, first from north to south with the wooden bars, then back again from south to north with simple steel bars which Mr Adams has had constructed for the purpose…The standard bar is kept in a vaulted cellar, constructed for the purpose, in which the temperature is as equable as can well be. Every night on finishing the day's work, and every morning before commencing, each bar - of which there were three, each ten feet in length - was carefully compared with the standard; an account of the temperature of the bars, ascertained by means of thermometers let into them, was also carefully kept at intervals during the day. The steel bars are also three in number, and each ten foot in length; each bar is padded with felt and enclosed in a wooden box, and the box itself is similarly padded and enclosed in an outer wooden box with square and true faces. About half an inch of each bar projects from the ends of the boxes, and on each of these ends there is embedded a small disc of silver, and the distance between the centres of the discs should be exactly ten feet, when the iron is at a certain specified degree of temperature. If the metal is above that degree of temperature, the bar will have expanded, and it will measure a trifle more than ten feet. On the other hand, should the temperature be below the specified point, the metal will contract, and the length of the bar will be under ten feet. It is necessary therefore that the temperature of the bar should be constantly watched, for, this being known, the length of the bar can be at once predicted. For this purpose there are thermometers attached to each bar, the bulbs of which are in close contact with the metal, and the temperature they indicate is constantly recorded…At the north end of the Richmond base, in consequence of doubtful foundation, the concrete foundation was placed some eight feet below the surface. At the south end the terminal dot is almost level with the surface…the amount of discrepancy between the measurement by means of wooden bars and the remeasurement by steel bars has not been completely investigated; but sufficient is known to show that the error cannot exceed two or three inches in seven miles, and will in all probability be reduced to within one inch.
An interesting aside is that in November 1879, Charles Todd, CMG (1826-1910), then Postmaster General and Government Astronomer, South Australia, attended a conference in Sydney. Whilst in Sydney Todd had the opportunity of visiting Richmond…, and examining the operations for the measuring of a baseline of verification in connection with the trigonometrical survey of the colony [of New South Wales].
For his measurement of the Richmond baseline, Conder used the same wooden bars that had been used for the Lake George baseline measurement. The difference between the two determinations of length, by wooden and steel bars respectively, was 0.662 of an inch in the total length of 7 miles (17 millimetres or 1.5 millimetres per kilometre). When the triangulation connected the Lake George baseline with the Richmond baseline, it was found that there was a discrepancy of only one and two thirds inches against the length of the Lake George baseline (42 millimetres or 5 millimetres per kilometre).
The Richmond baseline was remeasured with Invar band(s) in 1927 showing a difference of 0.2709 feet (about 85 millimetres or better than 10 millimetres per kilometre) between the 1927 and 1880 measurements.
The New South Wales triangulation now headed towards Bourke where a baseline was proposed to the south of the town along the railway. The marking of the base termini was completed in September 1915 and although the Bourke baseline had been selected, marked and connected to the network by October 1916 no measurement had been made and it was decided to partially suspend field operations due to war.
Chard’s Bullenbung baseline near Wagga Wagga was reported to have been abandoned. Nevertheless, the Bullenbung baseline was shown as part of the New South Wales triangulation in 1912, along with the Lake George, Richmond and Bourke baselines. Twynam’s Newcastle and Adam’s Ellerslie (Albury) baselines, however, are not shown on this plan of triangulation.
Triangulation Baselines pre 1912
South Australia became a separate colony in 1834 with Colonel William Light (1786-1839), the first Surveyor General of South Australia (1836-1838), appointed in 1836. By mid 1837, Light had selected the site for the city of Adelaide and completed a triangulation survey of eight stations across some 75 square miles. The first baseline for the triangulation of South Australia was established in 1840 by Captain Edward Charles Frome (1802–1890). Over the next three years Frome observed a network of triangles extending over the hilly country to the east, south, and mid north of Adelaide. His triangulation stretched from the Fleurieu Peninsula in the south to around Orroroo in the north.
Figure 17 : Section of map (nla.obj-231421482), The District of Adelaide, South Australia; as divided into country sections, from the trigonometrical surveys of Colonel Light, late Surveyor General, showing the triangulation stations at the north west corner of the city and Masters Hill in Section No. 164, with the red line indicating Frome’s 1840 baseline.
South Australian surveyor Charles Hope Harris (1846-1915), under the auspices of then Surveyor General George Woodroffe Goyder (1826-1898) produced the South Australian Field Service Handbook for Government Surveyors, (Harris, 1880 & 1887). In this publication, Harris stated…A baseline of about three and a quarter miles was measured, near Adelaide, during the year 1840…it was measured six times, the greatest difference being +0.83 feet, and it extended from the north-west corner of the city to Masters' Hill, on Section No. 164, the mean length being stated at 17,462.2 feet….A heavy steel chain, 100 feet in length, and a lighter chain 66 feet long were used to acquire the six measurements. A brass yard standard was used for reference.
The map, The District of Adelaide, South Australia; as divided into country sections, from the trigonometrical surveys of Colonel Light, late Surveyor General, showed not only the county sections but also the locations of the triangulation stations in and around Adelaide. Please refer to Figure 17 above for the relevant section of that map. A triangulation station was identified at the north west corner of the city and one in Section No. 164, shown as red dots in Figure 17. These locations transferred to a modern map as accurately as possible produced an equivalent distance, confirming that this was the most probable location of Frome’s Adelaide city 1840 baseline (6).
Figure 18 : Plaque indicating the location of Colonel Light’s Trig Station A, from where he commenced his survey for the city of Adelaide; the plaque is located on the corner of North and West Terrace; XNATMAP image.
During this investigation it was found that the triangulation station at the north west corner of the city was at the intersection of North and West Terrace and thus was undoubtedly Colonel Light’s Trig Station A, as shown in Figure 18 above. The original Masters Hill triangulation station in Section No. 164, is today located in the grounds of Adelaide Airport and appears to have been where the hangar for REX Regional Express now stands.
Harris continued…In course of extending the triangulation from time to time several other lines have been laid down, both for checks and as independent bases of operations for mapping localities, several hundred miles apart. No special value attaches to any them of but it may be of interest to indicate their positions, which are as follows :
Reported as No.2, Base of Verification, the circa 1850 Burra baseline (7) was listed to be between Bald Hill and Mount Bryan (also Mount Bryant) north of Burra near the town of Mount Bryan. The surveyor, apparatus used and actual length of the baseline was unrecorded. It is most likely however that the baseline was measured using 66 feet/100 link chain (this was the standard measuring apparatus of South Australian Government surveyors as described in Harris (1880)). The distance between the peaks of Bald Hill and Mount Bryan was calculated to be nearly 17 miles (27 kilometres).
Reported as No.3 base, James Brooks chained a baseline from Hurd Hill (14), south westerly for 8 miles 336 links (42 461.76 feet or some 13 kilometres) in 1861 (the use of the word chain indicated that he probably used a 66 feet/100 link chain as given by Harris (1880)). Hurd Hill is some 20 kilometres almost due east of the town of Olary on the Barrier Highway in South Australia. Olary is about 120 kilometres south west of Broken Hill, NSW. Brooks’ triangulation extended from Black Rock to the eastern boundary of the then province.
Reported as No.4 base, Samuel Parry measured a baseline of 23 189.70 feet (4.4 miles or 7 kilometres), near Hookina (9) between two small hills called Castor and Pollux, from which the distances in the surrounding country were calculated. The area of the 1858 triangulation carried out by Parry was 8 000 square miles. Again, it was likely that this baseline was measured using the standard measuring apparatus of South Australian Government surveyors, a 66 feet/100 link chain.
Reported as No.5 base, John Moyle Painter (c1817-1863) commenced his survey by measuring a baseline at Mount Serle (8) in 1857, again most likely using the standard measuring apparatus of South Australian Government surveyors, a 66 feet/100 link chain. Mount Serle is some 50 kilometres east of Leigh Creek. From his baseline Painter worked northward to Mount Distance and Mount Hopeless towards the head of Lake Frome. George Woodroffe Goyder, then Assistant Surveyor General, reported on 9 July 1857, that due to the adverse character of the country in the region of Mount Serle the measured baseline was limited to two miles. This line extended southwards between Mounts Serle and McKinlay, on a bearing of 7° 29' 25" west of the true meridian. The baseline was measured by Painter and Lee (his assistant) with Goyder’s oversight, three times with…the three measurements varying but three quarters of an inch in the entire length, and the mean measurement adopted as the correct distance. The baseline was subsequently extended by a basenet, to Arcoona Bluff to the north, and Mount Rowe to the south, giving a total length of baseline of 7.55 miles (12 kilometres).
Reported as No.6 base, George Woodroffe Goyder (1826-1898) measured a baseline with an ordinary chain (again likely a 66 feet/100 link chain), at Termination Hill (11) some 50 kilometres north west of Leigh Creek in 1859. It was reported that this baseline ran 20 miles (105 861.90 feet or some 32 kilometres) true north of Termination Hill to the highest point of the Alromba Range, which intersected the line. This high point in the range was named Twentymile Hill.
Not included in the above list but of the same era was another baseline established by Goyder in 1859. The 58 423.20 feet (11 miles or 18 kilometres) baseline was again measured with an ordinary chain (66 feet/100 link chain), and was near Mount Margaret (12) some 150 kilometres north east of Coober Pedy.
Figure 19 : Section of Cornish’s plan of triangulation in the region of his baseline at Thalepittinna Claypan (green line); in the bottom right corner it is noted that the baseline’s length was 9 107.71 feet.
The last pre 1912 triangulation baseline in South Australia appears to be that of Geodetic Surveyor William Henry Cornish (1850-1888) in 1880. Cornish measured his baseline situated in latitude 27° 35', longitude 138° 22' (today Thalepittinna Claypan) (20) eight times with Deal bars, 10 links in length, and twice with a steel bar also 10 links in length. All bars were standardised to the brass yard standard mentioned above. Two sets of measurements were made in summer and two in winter, corrections for temperature being carefully applied. The wooden bars instability finally led to the length of the baseline, as given by the steel rod alone, of 9 107.71 feet (1.7 miles or about 3 kilometres) being adopted (please refer to the annotation on the plan showing Cornish’s basenet at Figure 19 above. This length was considered to be correct to about 2 inches in a mile (around 30 millimetres per kilometre).
Thalepittinna Claypan is located to the west of Lake Howitt on the Birdsville Track about 250 kilometres north of Marree. Cornish’s triangulation extended from north of Hergott Springs, today Marree, to the northern boundary of South Australia, and eastward to the eastern boundary of the then province, encompassing Cooper Creek country from Innamincka to Lake Eyre. This base was used for the calculation of triangles extending about 150 miles east to the border with Queensland and 590 miles west as far as the boundary with Western Australia.
Using these existing baselines, as shown in Figure 20, South Australia continued its triangulation surveys to the north west and west of Adelaide. In May 1892 South Australia reportedly ceased such surveys.
Figure 20 : Map showing the locations and length of the South Australian baselines to scale in green.
Triangulation Baselines pre 1912
Victoria separated from New South Wales in 1851. Prior to separation, however, the need to establish the colonial boundary between South Australia and New South Wales saw Charles James Tyers (1806-1870) under the direction of Surveyor General of New South Wales, Thomas Mitchell sail to Melbourne from Sydney in 1839. From Melbourne, Tyers observed a triangulation consisting of fourteen stations and two astronomically determined bases being, Mount Eckersley, north of Portland, to Mount Sturgeon, near Hamilton, of 49.628 miles, and Cape Sir William Grant, on the coast at Portland, to Mount Eckersley, of 21.652 miles. Please refer to Figure 21 below. Suffice it to say that the result was that…this work forms no part of what is regarded as the trigonometrical survey proper of Victoria.
Figure 21 : March 1841 map of Trigonometrical survey of part of the country between Melbourne and The River Glenelg by CJ Tyers Surveyor & TS Townsend Assistant Surveyor
(courtesy National Library Australia nla.obj-232547618).
A few years earlier in 1836, Robert Russell (1808-1900), the senior of a three man party, was sent from Sydney to the Port Phillip Settlement, later to be named Melbourne, with specific survey instructions. Six months later in early 1837, New South Wales Governor Sir Richard Bourke (1777-1855), who had sent Russell and his men, along with Surveyor Robert Hoddle (1794-1881) arrived to inspect the Port Phillip Settlement. They found that Robert Russell had made a small triangulation survey, by means of which he had been able to prepare a plan showing illegal occupations. Bourke was dissatisfied with Russell’s progress and appointed Hoddle Surveyor-in-Charge of the Port Phillip District of New South Wales. Disgusted with Hoddle’s appointment, soon thereafter, Russell resigned his position in the Colonial Service. Hoddle is then credited with the design and layout of Melbourne city. Russell was later to recollect that to maintain the quality on his survey that a standard was brought down from Sydney (he thought it was a brass rod about eight feet long), and there were pegs left in the survey yard to check and maintain the chain to its correct length.
Now a colony in its own right, in October 1852, by direction of Victorian Governor Charles Joseph La Trobe (1801-1875), Clement Hodgkinson (1818-1893) commenced what is recorded as a detailed survey of Melbourne on trigonometrical principles between points on Batman's Hill, Point Ormond, at the mouth of the Yarra, on the high ground in Royal Park, and in Studley Park. No record of Hodgkinson’s baseline or subsequent triangulation has ever been found but it was believed to be of a high standard. The baseline on the Melbourne swamp (now occupied by the facilities associated with the Swanston and Appleton docks) was measured by Hodgkinson using tested pine rods properly compared with a standard yard measure belonging to the Customs Department.
What was initially to be a limited triangulation but ended up covering some 400 square miles, was commenced early in 1854 by George Christian Darbyshire (1820-1898). The constrained extent of the project meant that the triangulation baseline was not measured with appropriate care other than that the hundred link (66 feet) surveying chains used were adjusted with reference to the standard laid down at the Melbourne Government Offices. This standard however, may not have been derived from a strictly accurate source. Nevertheless, it was regarded as sufficiently precise for a triangulation of the initially small area. Over the extensive area finally surveyed the error was magnified and the final survey results were thus rendered mostly worthless.
Part of the Wimmera triangulation was also shown to be of suspect quality. It was found that the work of Surveyor Thomas Burr (1813–1866) was reported to have so many discrepancies that it appeared that Burr had read only two angles in each triangle, and found the third by taking their sum from 180 degrees…it was evident that in his triangulation Thomas Burr had no regard for the figure of the earth (if the triangles are sufficiently large then their three internal angles may sum to more than 180 degrees).
The actual Geodetic Survey of Victoria, by triangulation, was commenced in 1860, but in its earlier stages consisted more of the definition of meridian and parallel (chordal) lines embracing squares of one tenth of a degree than of triangulation. As will be seen this approach was still controlled by a baseline but the triangulation was undertaken along specific corridors.
Mid September 1858, saw the determination by triangulation of the difference of longitude between the Williamstown Observatory and a point selected in the Royal Park, to the north west of the Melbourne CBD, as the starting point of the first chief meridian. This chief meridian was at longitude 145 degrees east and was selected as it was then within two miles of the Williamstown Observatory. The Supplement to the Victoria Government Gazette of Friday, March 1, 1861 (Government of Victoria, 1861), published the position of (Williamstown Timeball) tower as 37° 52' 08" S, 144° 58' 30" E (9h 39m 54s) meaning the meridian at 145° east would be 1.5 minutes of arc further east or less than 2 miles in distance. Please refer to Figure 22 below. The chief meridian at 145° east was described as running from the waters of the Bay (near the Port Melbourne Railway Station, now a light rail stop) to the River Murray.
Figure 22 : Section of 1943 1: 253 440 scale SJ55-05 Melbourne map sheet showing the relative positions of the Williamstown observatory and the then 145 degree Meridian.
(It needs to be remembered that today the longitudes in Victoria are some 7 kilometres different from those used in the 1850s. Initially observatories determined longitude from their own astronomical observations completely independent of other observatories. The first determination of the difference in longitude between the Observatories of Sydney and Williamstown by the telegraphic exchange of clock signals, took place in 1861. Relative to Sydney, Williamstown was then at 144º 57’ 59” E (9h 39m 51.94s) longitude. The longitude of the Williamstown Observatory was, however, determined by observations of moon culminations in the years 1860, 1861, and 1862, resulting in the adopted value being 144º 54’ 42” E (9h 39m 38.8s) longitude, as it was found that Sydney Observatory had its own longitudinal error.)
Accurate triangulation enabled the difference in longitude between the Williamstown and Melbourne Observatories to be computed as +4 minutes of arc (16s). When this difference was applied to the longitude of Williamstown it gave the longitude of the Melbourne Observatory as 144º 58’ 42” E (9h 39m 54.8s) which was adopted until 1883. In the 1930s when the triangulation between the Sydney and Melbourne Observatories was complete and a common datum adopted, the old Williamstown Observatory coordinates were more accurate than those of Melbourne. It was apparent that, despite best efforts at the time, the Williamstown-Melbourne triangulation had introduced an error of around 15 seconds of arc (1s of time) resulting in the longitude value for the Melbourne Observatory being in error by that amount. A 1968 map last quoted the coordinates of the Melbourne Observatory as 37° 49' 52.47" S, 144° 58' 33.36" E, prior to all Australian geographical coordinates being standardised to AGD66).
The baseline for this initial triangulation from Williamstown to Royal Park was of two miles measured on the rails of the Williamstown and Melbourne Railway with a standard [100 link/66 feet] chain. The meridian was then traced, cleared, and measured with many observations of intermediate points that it was considered that in the forty miles the error of deviation from the true meridian did not exceed one inch. It was not until March, 1859, when the measurement of the first standard parallel (37° 48’ S latitude at that time) was commenced, starting from the first chief meridian near Flemington. Later, when the chief meridian had been extended as far as 37° S latitude (at that time near Puckapunyal) and subsequently to the Murray, that determinations had been made astronomically, and checked by careful triangulation, of the geographical position of Mounts Macedon and Warrenheip, also the intersection of the 37th parallel with the 144th meridian (near Maryborough) as well as the 145th. A most careful chain of triangulation was also carried along the parallels to fix the intersections of the secondary meridians.
It soon became evident that connecting distant localities with the chief meridian by running and measuring meridians through great expanses of country not required for survey was not cost-effective. Primary triangulation was the only way to rapidly reach the distant districts and connect the work generally. Robert Lewis John Ellery (1827-1908) Government Astronomer, who had been appointed to the position of Superintendent of the Geodetic Survey, in 1858, had been obliged to follow the meridian and parallel (chordal) line design for Victoria. When, in 1860, the work of defining meridian and parallel (chordal) lines was abandoned in favour of the traditional triangulation Ellery seized the opportunity to standardise the geodetic triangulation of Victoria.
Ellery measured a baseline on the Werribee Plains (13) some 30 kilometres south west of Melbourne between January and May 1860, using three iron bars each with a length of 10 feet (please refer to Figure 6 above). The southern end of the baseline was a point on the Melbourne to Geelong railway reserve, about 2 miles east of the Werribee railway station. The baseline extended 26 091.826 feet, or 4.941.16 miles to a point in direct line with Green Hill (today Eynesbury) on the western side of the Werribee River. Over 2 000 observations were made in the process of determining the length of the baseline. The measured base was subsequently extended by a basenet, a further 5.651 miles to Green Hill, its total length then being 55 931.65 feet or 10.593 miles. Please refer to Figures 23 and 24 below. The north and south ends of the measured baseline were permanently marked with sunken masonry piers, having in their upper surfaces a piece of brass carrying a platinum dot indicating the termini of the measure. These marks were then covered with heavy cap stones. Please refer to Figure 25 below. The mark at the end of the extension to the north, on Green Hill, consists of a sunken bluestone block with a projecting iron spike.
Figure 23 : Section of 1943 1: 253 440 scale SJ55-05 Melbourne map sheet showing the location of the Werribee baseline (green line); south base-north base-Green Hill.
Figure 24 : Section of 1943 1: 253 440 scale SJ55-05 Melbourne map sheet showing the basenet configuration for the Werribee baseline (green line); south base-north base-Green Hill; points E, W, and Western Auxiliary are only located by scaling.
Figure 25 : Site of the South Base terminal at Werribee.
The three iron bars (identified as I, II and III) used by Ellery were constructed in Victoria. The bars were assumed to be 10 feet long but before being used for measurement were standardised against the 10 foot standard lent by the New South Wales Government, namely OI4, as described above. Victoria subsequently obtained its own standard from England being OI6, also described above. The three iron bars were fitted with steel ends, one flat and the other rounded to form a section of a sphere of 5 feet radius. They were used by being placed in series, with distances of about ¼ of an inch between the spherical end of one and the flat end of the next. The precise distance between these two terminals was then obtained by carefully inserting a graduated wedge between them until contact with each side was made. The graduations on the wedge then allowed the size of the gap to be read. The wedge was of hard bell metal, 7 inches long by 2 inches wide and the inclination of the faces was 30 minutes of arc. The iron bars were kept level during the measurement. A remeasurement of the southern 11 174.29 feet (2.11 miles) with the bars following the general inclination of the ground instead of being level as during the first measurement, was also undertaken. A difference of only 0.308 inches (about 8 millimetres), or about 0.15 inches per mile (better than 2 millimetres per kilometre) was recorded. This result indicated that the measurement of the Werribee baseline was the most accurate to date.
It would appear from Ellery’s own 1891 report that the measuring bars used at Werribee underwent a further calibration when Victoria received its own 10 feet (120 inch) standard OI6 in March 1862; their actual lengths in terms of the Melbourne edition of the British 10 feet standard were as follows :-
No. I : 119.99957 inches ;
No. II : 119.99318 inches ; and
No. III : 119.99997 inches, at a temperature of 62°F.
There is evidence that Ellery intended to remeasure the Werribee baseline with Colby type bars; a second measurement will shortly be commenced with some new rods, I have got made here - they are of Iron planed on the upper surface forming a compound bar…the iron and copper being permanently fixed together at one end, the other ends being free with marks for determining the expansion and interval between the rods. No physical verification of the rods or the remeasurement, however, have ever been found.
Immediately on completion of the baseline, triangulation was extended over the whole of Victoria with the exception of the north western mallee section. The north east region triangulation was by Alexander Black (1827-1897) and the south west work by Alexander Charles Allan (1831-1901). From 1862 to 1874, William Turton (1828-1906) undertook triangulation in the Gippsland area ranging from Bass Hill in the west through to Cape Howe and the New South Wales Border in the east. Before being sent to Gippsland, Turton had been involved in the latter stages of the measurement of the Werribee baseline during the first six months of 1860. Over the years Turton and Ellery both developed a profound respect for one another.
Figure 26 shows the locations of some of Turton’s stations including Baldhead, Baw Baw, Ben Cruachan, Boisdale, Elizabeth, Fatigue, Hoddinotts, Lakes Entrance, Merrimans, Seacombe, Shallow Inlet, Stockyard, Taylor, Toms Cap, Useful, Wellington and Woodside. At Seacombe, ti-tree was cleared in gaps about two to six hundred metres long and up to fifty metres wide; at Taylor, two thousand six hundred metres by fifty metres and at a station east of Lakes Entrance, three thousand two hundred metres by forty metres was cleared to established the necessary lines of sight to the other stations! The extent of this work prompted Turton to obtain Ellery’s approval for the equivalent of a reconnaissance party being formed, to carry out the clearing and station establishment independently of the main observation party. Considerable time could be saved by not having to return, sometimes on several separate occasions, to the same station to complete all the necessary work.
Figure 26 : Section of map showing the locations of Turton’s Gippsland triangulation stations.
On the extension of the primary triangulation westward to Mount Gambier and nearby Mount Schanck, in South Australia, Allan in July, 1864 reported that between these two mountains was an admirable site for a base of verification and that his proposed site was a much better one than that which had been selected for the Werribee Base. Allan recommended a length of 5.5 miles for the new baseline. While the intention to measure a baseline of verification was agreed, the actual work, however, was never performed.
During the twelve years of the Victorian survey, 209 primary stations were established and observed. The whole area of the colony of Victoria, with one exception, Ellery observed in 1873, had been triangulated with a primary triangulation almost equal in accuracy to any in the world, from a baseline, which was measured with all the refinement, skill and instrumental means then available. Ellery continued, pressing for two baselines of verification; one on the western boundary of Victoria, and one in Gippsland, to close and verify the triangulation. The end of the Victorian triangulation came without either baseline being established, which would have been a great disappointment to Ellery, for there was no independent check on the work of his department. The quality of the Victorian triangulation was only revealed in the 1930s.
The demarcation of the straight line, essentially from the source of the Murray River to Cape Howe, portion of the boundary between Victoria and New South Wales, in 1870 meant that the Victorian triangulation survey ranged north into southern New South Wales. Stations were established on Mount Pilot and Mount Kosciusko, and at Cape Howe. The line now known as the Black-Allan Line, after Alexander Black and Alexander Charles Allan assisted by Turton, previously mentioned, then being indicated by marks on the ground. When the New South Wales triangulation survey later reached that area the coordinates of Mount Pilot, Mount Tingaringy (today’s spelling), Mount Delegate (also Delegete) and Mount Buckle derived independently from both surveys could be compared. For the period, the overall quarter of a second of arc difference in latitude and around 8 seconds of arc difference in longitude were acceptable. The difference in longitude came from the fact that the longitude origin for the surveys came from independent observations at the Melbourne and Sydney observatories and until then the accuracy of one longitude determination against another had never occurred. This longitude difference was resolved in later years as technology developed.
Also associated with Ellery was a very small specific purpose triangulation in Melbourne in 1863. Some ten years earlier, at what was already a naval depot and signal station at Point Gellibrand, now Williamstown, a time ball was set up on the flagstaff there as a time signal facility for the port’s shipping; when the time ball dropped daily at usually 1PM the ship’s masters could adjust their clocks. Ellery was made Superintendent of the Observatory overseeing the time facility.
As the port developed it impacted the operation of the observatory equipment. The advent of the railway to Williamstown in 1859, such that trains were now travelling within 200 metres of the facility, dictated a new site be found. Due consideration of many factors saw the Melbourne Botanical Gardens selected as the site of the new facility. So as not to lose some ten years of observations establishing the longitude and hence time at Williamstown, a small triangulation was observed between the Williamstown and new Melbourne observatory. Please refer to Figure 27 above. The triangulation found that the difference in longitude between Williamstown and Melbourne was 4.0 minutes of arc. Thus, the longitude of the new Melbourne observatory was calculated to be 16.0 seconds of time greater than that of Williamstown.
It is recorded that with the establishment of the Melbourne Observatory, Ellery used the Victorian standard OI6 to set out a baseline of 100 feet to enable the Surveyor’s 66 feet/100 link and Engineer’s 100 feet chains, to be accurately calibrated. The observations for this baseline were performed by Ellery in 1892-3. Terminal coordinates were also found for an Observatory baseline. This baseline was 750.3 feet (228.7 meters) long and ran approximately north-east to south-west to the west of the Observatory in the Botanic Gardens and outside the Observatory precinct. Please refer to Figure 28 below. It is uncertain if this baseline was part of the control for the Williamstown-Melbourne Observatory triangulation but the baseline was certainly not part of the survey control for the construction of the Shrine of Remembrance and was destroyed when the Shrine was built and landscaped.
Figure 28 : Map of the Melbourne Observatory environs in the late 1800s showing the relative locations of (A) the Melbourne Observatory baseline and (B) the Chain Standard baseline.
Triangulation Baselines pre 1912
Before Queensland separated from New South Wales, in 1859, Robert Dixon (1800–1858) established a baseline in 1839. Dixon’s baseline, on what was known as the Normanby Plains, was in the region of today’s Harrisville, 20 kilometres south of Ipswich, near Warrill Creek. Please refer to Figures 29 and 30 below. The Ipswich baseline (4) of some 3 miles was measured with three Deal rods each ten feet long, tipped with brass. The survey from this base led to the first accurate map of the region. In May 1849, James Charles Burnett (1815-1854) measured a north-south baseline for his triangulation network. Burnett’s Jondaryan baseline (5) was 2 miles and 10 chains (170 chains) long and was situated about midway between Mount Maria and Mount Irving on the plain south of the town of Jondaryan on the Warrego Highway between Oakey and Dalby. Burnett used the same 3 Deal rods that Dixon had used in 1839. As is described below, Mount Maria and Mount Irving later became the terminals for the longer 7 mile, approximately east-west, Jondaryan baseline.
Figure 29 : Section of April 1840 map (nla.obj-2005473077), Trigonometrical Survey of the country at Moreton Bay, by Robert Dixon, showing the location of his baseline. Note that the peak shown as Mount Forbes by Dixon is today named Mount Walker with Mount Forbes a lesser peak to the north.
Figure 30 : Section of 1955 1: 253 440 scale SG56-14 Ipswich map sheet showing the location of the Dixon’s Ipswich baseline (dark green) to the south of a line (light green) between Flinders Peak and Mount Walker.
Given Surveyor General Gregory’s thoughts, detailed above, on the quality of Mitchell/Dixon’s initial baseline and triangulation, it would appear that Gregory wanted a new start. Government Surveyors Robert Hoggan (1852-1929) and Archibald McDowall (1841-1918) thus measured a baseline between Mount Irving and Mount Maria of 7 miles in 1883. Please refer to Figure 31 below. The now Jondaryan baseline (21) was measured using two 100 foot steel bands. Each band was standardised against a steel bar floating in mercury, which itself had been carefully standardised by measurement against the standard bar (OI4) of New South Wales. Three separate readings were taken with each band, so that six independent readings were obtained and could be checked against each other. Overall, the difference between the means of the three measurements from each band amounted to 0.117 inches (3 millimetres). The mean length of the Jondaryan baseline was accepted as 37 029.7549575 feet (7 miles or 11 kilometres). The completion of this baseline then allowed triangulation of a large tract of Queensland; including the Darling Downs, Moreton, Wide Bay, and Burnett districts, stretching south to Wallangarra (beyond Stanthorpe), north to Mount Perry (north of Mundubbera), and eastward to the coast.
In mid 1939 then Lieutenant Lawrence FitzGerald (1903-1988), RA Survey used Invar bands to remeasure the Jondaryan baseline as part of a program to consolidate the triangulation in the eastern states. There is more discussion on this aspect later in the paper.
Figure 31 : Section of R502 1: 250 000 scale SG56-14 Ipswich map sheet showing the location of the Jondaryan baseline (approximately east-west, thick green line); the other line is the location of Burnett’s 1849 baseline (approximately north-south, thin green line).
The Surveyor General of Queensland then instructed staff surveyor Henry Wood to carry out a triangulation survey of Brisbane city in the late 1880s. A minor baseline of 1 923.748 feet was established, with 100 feet steel bands, along the river bank on the eastern side of the Brisbane Botanical Gardens. Please refer to Figure 32 below. This baseline was measured twice with one of the bands used on the Jondaryan baseline, and once with the other band used at Jondaryan; a fourth measurement was made with a new 100 feet band. A basenet extended from this baseline to finally include the side of a primary triangle formed by the line between Mount Petrie and Eildon Hill. Please refer to Figure 33 below. The length of the Mount Petrie and Eildon Hill baseline (29) was about 9 miles 25 chains. When the length of the minor baseline in the gardens was then computed from the Jondaryan baseline through the intervening primary and city triangulations, in 1889, the length was found to be within 0.01 inches (about 3 millimetres). Triangulation in Queensland was discontinued in 1891.
Figure 32 : Section of 1893 map (nla.obj-692765192) Plan illustrating Trigonometrical Survey of Brisbane by R. Hoggan etc, showing the location of the Brisbane CBD minor baseline.
The survey of Brisbane also comprised a standard traverse of South Brisbane made during the first half of 1893, by Cecil Twisden Bedford (1850–1917) and Alan Alfred Spowers (1857-1937), staff surveyors. This standard traverse was run along the main streets, and the stations permanently marked by iron castings set in concrete. The survey was connected with the triangulation survey and was executed with a high degree of accuracy.
Triangulation Baselines pre 1912
Under the administration of South Australia for mapping the positions of gold and tin claims, in 1870, George McKay measured a baseline (15) in the area of McMinns Bluff (after South Australian Government Surveyor Gilbert Rotherdale McMinn) near Pine Creek some 200 kilometres south of Darwin. His survey area reportedly covered an area of 1 000 square miles.
By direction of the South Australian Government, a triangulation survey of the MacDonnell Ranges, Northern Territory, was commenced at Alice Springs in 1880. Piles were erected on suitable hills for triangulation stations situated 20 miles west of the telegraph line, and 100 miles east, but detailed observations never reached the head office.
Around 1884, David Lindsay (1856–1922) chained the Roper River (22) triangulation baseline somewhere between All Saints Well on the then Overland Telegraph Line in the west and Todds Bluff on the Roper River in the east. His survey area was some 500 square miles.
Figure 34 : Map showing the approximate locations of the McKay, Lindsay and Wells baselines in the Northern Territory of South Australia.
During the years 1906-1908, Lawrence Allen (Larry) Wells (1860-1938) measured the Victoria River triangulation baseline (33) somewhere between the then Overland Telegraph Line in the east and the eastern boundary of Western Australia in the west. Wells’ survey area was some 20 000 square miles.
There is no information about the length of the above baselines or their means of measurement. It is most likely however that they were all measured with a standard 66 feet/100 link chain, the standard measuring apparatus of South Australian surveyors.
Based on the information provided above, the approximate locations of the McKay, Lindsay and Wells baselines in the Northern Territory of South Australia are shown at Figure 34 above.
Triangulation Baselines pre 1912
The city of Perth began as the Swan River Colony in 1829, founded largely because the British feared the French would establish a colony in Western Australia. In 1873 the colony of Western Australia was founded.
Figure 35 : Map of Western Australia by the Department of Lands and Surveys, showing for 1906-07 the locations of the De Grey baseline and triangulation and Jennaberring baseline east of Perth
(courtesy State Library Western Australia, slwa_b2244687_1.jpg).
John Septimus Roe (1797-1878) was the first Surveyor General of Western Australia and in 1829 immediately set about making preliminary surveys of the harbour, river and surrounding land. The sites of Perth and Fremantle were chosen on his recommendation, and he was responsible for laying out these towns. Alongside general settlement, the finding of gold was a driving factor for surveys in Western Australia.
The Perth Observatory was connected to the triangulation in July, 1901. The Observatory’s adopted latitude and longitude was then used as the datum for computing the positions of all the triangulation stations in the State. That work involved a complete reduction of all the triangulation surveys and the elimination of differences between baselines and azimuth determinations.
Between 1882 and 1887 the essentially coastal triangulation between Fremantle and Geraldton had been observed. The remainder of the triangulation, to the south and north of these locations was then completed by around 1900. The 1911 Department of Lands and Surveys, Map of Western Australia : Showing the Triangulation System, shows triangulation baselines at or near the following localities or on the respective lake beds; Albany, Busselton, Cape Leeuwin, Cervantes, Esperance, Hardman Range (Ord River region), Lake Brown, Lake Deborah, Lake Moore, Marble Bar and Toodyay. The locations of these baselines are shown on the map that may be viewed via this link. At that time, however, it was reported that six baselines had been measured with steel wires 66 feet long; near Perth (23), near Geraldton (24), on the Gascoyne River (25) opposite its confluence with Dalgety Brook, near Globe Hill (26) on the Ashburton River, on the Fitzroy River (27) and near Wyndham (28). The locations of these baselines are shown on the map that may be viewed via this link. The Wyndham baseline was found to have been established by Harry Frederick Johnston (1853-1915, and Surveyor General 1896-1915), between 16 April and 14 May, 1885, on the plain between the East Bastion and Quoin Hills, a distance calculated to be some 9 miles or 14 kilometres. Please refer to Figure 36 below.
Figure 36 : Section of R502 1: 250 000 scale SD52-14 Cambridge Gulf map sheet showing the location of Harry Frederick Johnston’s baseline (green line).
Doubts arose as to the accuracy of the standard length used by Western Australia around 1902. Two three quarter inch steel bands were subsequently obtained and compared with the British standard and stamped by the British Board of Trade as being of the standard length of 66 feet at a given tension and temperature. One of the bands was then enclosed in a wooden trough in the basement of the main building in Perth, and adapted for accurately testing chains before further use. The second band was carefully stored for use in the case of any accident to the other. The new standard showed a very close agreement with the one previously used.
Figure 37 : Map showing the locations of the Western Australian baselines established between 1899 to 1912.
Baseline establishment from 1899 to 1912 involved the 1899 triangulation baselines at Lake Carey (30) and Lake Raeside (31). Both baselines were measured with an Invar 500 link band. The survey work covered the nearby Mount Margaret goldfields. Lakes Carey and Raeside and the goldfields are south of the now Leonora-Laverton Road.
Three triangulation baselines (32) were measured with an Invar, probably 500 link, band in 1904, 1905 and 1907 in the Kimberley region. The baselines were located east of Derby; one on the Meda River and two, in a quadrilateral arrangement, on the Fitzroy River. These baselines and the earlier Wyndham baseline were all connected by triangulation.
The De Grey baseline (34) was measured with an Invar 500 link band in 1906. This baseline was just south of today’s Shay Gap and north of the De Grey river.
Nearer to Perth, both the Jennaberring (35) (the location of the Jennaberring baseline is shown in Figure 35 above) and Rockingham (36) baselines were measured using an Invar, probably 500 link, band. In the annual report for the year ending 30 June, 1911, the Surveyor General stated that…An important trigonometrical connection was completed during the year between the coastal triangulation and that carried out by Mr HS King between Mount Bakewell and Mount Stirling in 1890…By means of this connection a comparison was obtained between the Rockingham base measured in June, 1908 and the Jennaberring base measured in August, 1906. The computed values of arc HK 211 to County Peak:- From Rockingham base equals 118,237.76 links; from Jennaberring base equals 118,234.82 links [about 40 miles or 25 kilometres]; difference equals to about 2-10th of a link per mile [0.13 inch/mile or 2 millimetres per kilometre].
An Invar 500 link band was used in about 1911 to establish the 48 500 link (6 miles or 10 kilometres) Irwin baseline (37). Apparently named after the nearby Irwin River, the terminal coordinates locate the baseline some 30 kilometres east of Dongara. Please refer to Figure 37 above and Figure 38 below.
Figure 38 : Section of 1943 1: 253 440 scale SH50-05 Dongara map sheet showing the location of the Irwin baseline.
A map showing the locations of the triangulation baselines for the period 1840-1912 and the triangulation work reported by the respective Surveyor’s General at the 1912 conference, may be viewed via this link.