What Hawkesbury sandstone is, how it shaped the harbour and the city, what makes it so good to carve, why it decays, and where the stone comes from now.
On 15 May 1788, less than four months after the First Fleet landed, Governor Arthur Phillip wrote home to report what he had found underfoot: a freestone “which appears equal in goodness to that of Portland”. He was comparing it with the limestone used for St Paul’s Cathedral, and for a man looking at a beach, some bark huts and a great many convicts, it was a remarkably confident thing to say.
Phillip was right about the stone’s promise, if not about its durability. Within a century it would be the material of nearly every important public building in the colony. It gave The Rocks its name, supplied the sea walls of Circular Quay, and made a trade so skilled that Sydney’s masons could strike for shorter hours and win. It also determined, long before anyone arrived to quarry it, the shape of the harbour.
Geologists call it the Hawkesbury Sandstone. It was laid down in the Middle Triassic, about 247 to 242 million years ago, before there were flowering plants and early in the age of dinosaurs. At the start of that period the Lachlan Fold Belt rose to the south-west, and the Sydney Basin tilted down to the north-east. Rivers came off the high ground carrying coarse quartz sand worn from old quartzites, and spread it across the basin in a sheet more than 200 metres thick in places.
The scene is hard to place. Sydney then lay at about 62 to 64 degrees south, close to the Antarctic Circle, and yet the climate was warm, with little evidence of polar ice. Picture a wide sandy plain crossed by fast, shifting channels, with seed ferns and horsetails along the wet margins. The classic comparison, made in 1975, is the Brahmaputra in Bangladesh. It isn’t settled, however. Other geologists have proposed a drier inland system more like Cooper Creek, or a delta influenced by tides. The rock has been studied for well over a century and its origin is still argued over.
Today it covers about 12,500 square kilometres of the Sydney Basin. It is commonly 100 to 200 metres thick, thinning to 34 metres in the lower Blue Mountains and reaching 240 metres around the Hawkesbury River. Beneath it lies the older, more mixed Narrabeen Group. Above it, a thin transition called the Mittagong Formation gives way to the Wianamatta Group, whose shales underlie the flatter country of western Sydney. The boundary between sandstone and shale is one of the most consequential lines in the city: it largely separates the steep, bushy, harbour Sydney from the rolling plain to the west.
Look at any large exposure and you will see its signature: slanted lines inside the beds, running at an angle to the horizontal layers above and below. This is cross-bedding, and each slanted set records sand avalanching down the front of a dune or bar as it crept along a river bed. Geologists describe the beds as lens-shaped, “resembling flattened pillows”, from 1 to 10 metres thick and 100 to 500 metres across. You are looking at the direction a Triassic river flowed.
About 5 per cent of the formation isn’t sandstone at all. Lenses of shale formed where a channel was cut off and became a still pool, and mud settled out. These pools preserved the fossils. The Australian Museum holds 52 described species from the Hawkesbury Sandstone, among them 28 fishes, 11 insects and 3 amphibians, many from a brick pit at Beacon Hill.
A note on names: W. B. Clarke called these the Hawkesbury rocks in 1878, partly because “Sydney Sandstone” was already taken by a rock unit in Nova Scotia. In the building trade “Sydney sandstone” means something wider, including similar stone from the upper Narrabeen Group around Gosford. That wider sense matters later, because it’s where most new stone comes from today.
Sydney Harbour is a drowned river valley, cut up to 85 metres down into the sandstone, mainly when ice ages lowered the sea. About 17,000 years ago, when the coast lay some 25 to 30 kilometres east of where it is now, the sea began to rise quickly. It reached roughly today’s level about 6,000 years ago and flooded the valley. The people living here would have watched it happen over generations. Geologists still don’t fully agree on one puzzle: the valley is unusually large for the small catchment that drains into it, and the Parramatta River may once have been much bigger.
Much of what makes the place look the way it does comes from how the sandstone breaks down. The rock has two main forms. Massive beds, with little layering, weather quickly into ridges, bluffs and caves. Sheeted, cross-bedded beds weather more slowly and form ledges. Widely spaced vertical cracks, or joints, control the line of the sea cliffs. Together these produce the stepped headlands and overhangs that define the coast from the Heads to the Royal National Park.
The soils tell a similar story. Quartz has almost nothing to offer a plant, and soils on the Hawkesbury Sandstone are naturally low in nutrients, phosphorus above all. Native plants here are so adapted to this poverty that in one study most of them died once soil phosphorus rose above 200 milligrams per kilogram. So the sandstone also shaped the distinctive bush on it, and it’s part of why early settlers found so little farmland around the harbour and headed for the shales and river flats to the west.
For Aboriginal people the sandstone was shelter, workshop and gallery for thousands of years before Phillip’s letter. Its overhangs gave shade and protection and its high platforms long views, and flat rock near water was used for sharpening: the quartz grains are hard enough to grind stone tools, and the grooves are still visible in outcrops, often beside creeks. One account describes the region’s people as having “no good stone”, because rock for cutting edges and spear points had to come from elsewhere. What the sandstone offered was different.
It offered a surface. On rock platforms around Sydney and the Hawkesbury, people engraved animals, tracks, people, tools and ancestral beings. A 2023 NSW Government history counts more than 600 documented sites with over 4,000 figures, while other tallies are far higher, because they count differently. Engraving is usually described as a process of outlining a figure, pecking holes along the line and rubbing them into a groove. Some engravings show European ships and so can be dated to the last two centuries. Others are thought to be far older, though dating them is difficult, and people are known to have re-cut grooves during ceremony.
Many sites are deliberately not publicised. The Bulgandry engravings in Brisbane Water National Park, shown here, are a managed, public Aboriginal Place that NSW National Parks asks visitors to treat as sacred, without walking on or touching the rock. Sources disagree about which nation’s country this is, so we don’t name one here. The land beneath the Pyrmont quarries, the source of the stone for much of the city, is Gadigal Country.
A freestone is a stone with no strong planes of weakness, which can be sawn and carved in any direction without splitting along a layer. This is what Phillip noticed, and most Sydney stone has some of the quality. The trade recognises six kinds of Sydney sandstone: grey, yellowblock, white, quartz-rich, colour-banded and ordinary surface Hawkesbury stone. The quartz-rich kind is by far the most abundant. The one that built the Victorian city is the yellowblock.
Yellowblock has less quartz and much more clay than its common relatives, and it contains a small amount of an iron carbonate mineral called siderite. Those differences account for almost everything people love about it, and for most of its problems. The clay binds the sand more softly, so a carver can work fine detail: acanthus leaves, faces, lettering. State of Stone, the NSW Government’s 2023 account of the state’s stonework, calls it a stone-carver’s dream, and credits it with allowing architects like Mortimer Lewis and James Barnet to specify the columns and friezes they did.
The golden colour isn’t the colour of the rock. Fresh yellowblock is a pale, creamy grey. Once exposed to air, its iron minerals are drawn toward the surface and oxidise, and over a matter of weeks or months (sources vary) the block turns yellow-brown. The conservation literature calls this self-colouring. A new block in a wall starts visibly paler and sharper than the old stone around it, and catches up with time.
The change is more than cosmetic, because the same process forms a hardened skin on the stone, known as case hardening. It’s a mixed blessing. It protects the softer stone behind it, which is why rubbing a dirty facade back to a clean surface can do lasting harm, letting water and salt in. Other Sydney stone behaves differently. Grey Wondabyne stone from the Hawkesbury doesn’t oxidise to gold, and is reported to have been chemically stained to match when used in repairs.
Quarrying began almost at once. A convict map of June 1788 already shows a quarry, and through the 1790s digging around Bennelong Point and Millers Point cut back the foreshore. Governor Macquarie opened a government quarry and had convicts trained as masons. In 1810 convicts cut a bench from a ledge in the Domain for his wife, the rock seat still called Mrs Macquarie’s Chair. It’s a modest thing, and a good place to see the natural rock and worked rock side by side.
The big convict works came later. On Cockatoo Island from 1839, convicts hewed about twenty grain silos straight down into the rock, and their stone went into the works at Circular Quay. From 1843 convicts began the Argyle Cut, a road through the ridge of The Rocks, abandoned half-finished and completed by the City Council with explosives in 1859. The stone taken out went into the Quay’s sea walls. Fort Denison, in the middle of the harbour, took around 8,000 tonnes from Kurraba Point on the north shore. (Sources differ on the dates.)
The stone that made Sydney look as it does came mostly from one small peninsula. In the 1850s Charles Saunders opened yellowblock quarries in Pyrmont and Ultimo. His Scottish workmen named them Paradise, Purgatory and Hell Hole, apparently for how good the stone was or how hard the going, though nobody recorded exactly why. His son and grandson, both named Robert, ran the business until about 1931, and it was the first in Australia to use steam for quarrying, lifting, cutting and dressing stone. By 1928 more than half a million cubic yards had come out of over twenty Pyrmont quarries.
Pyrmont stone is credited with the Town Hall, the Australian Museum, the Art Gallery, Sydney Hospital, the Great Hall of the university, the General Post Office and St Mary’s, though attributions like these are not all equally well documented. The Queen Victoria Building is variously credited to Pyrmont, Waverley and Maroubra. It travelled too: the Stone Store at Kerikeri, the oldest stone building in New Zealand, appears to have used it in the 1830s, and later shipments went to Melbourne, Adelaide and Brisbane. Yellowblock was shown at the 1862 International Exhibition in London.
A trade this central had leverage. On 18 August 1855 the Operative Stonemasons’ Society told Sydney’s employers that in six months its members would work only eight hours a day. Masons on two church sites, the Garrison Church among them, moved early and won, and held a victory dinner on 1 October. Most of the city’s masons had the eight-hour day by early 1856, at a cut in pay. The Melbourne masons’ action of April 1856 is often cited as the first industry-wide win, so “first in the world” is contested for Sydney. Even so, it was among the earliest anywhere, and in Australia the trade that did it was the one cutting this stone.
What ended the great age was not exhaustion alone. By the early twentieth century, the Pyrmont supply was exhausted or too costly to work. New quarries in the eastern suburbs supplied the State Library and extensions to the Museum and Gallery from the 1920s. After the Second World War, steel, concrete and glass took over, and Sydney built with something else.
I do not know a single stone building in this colony 80 years of age which is in anything like good condition.
Vernon built in this stone all his life, and he knew it. The properties that make yellowblock carvable also make it vulnerable. Its clay-rich binder holds water, and the iron that colours it goes on reacting. A study of Sydney’s heritage buildings found that the kaolinite clay holding the grains together deteriorates over about a century of exposure, its structure altered as iron from the stone substitutes into it. The surface turns orange and brittle and flakes off, then the next layer goes the same way, so a block crumbles from the outside in.
Salt does much of the rest. Salt-laden water moves through the stone and evaporates just behind the face, where crystals grow in the pores and force the grains apart. The salts come from sea wind, pollution, rising damp, even detergents and fertiliser. Freezing, the main culprit in European stone, is irrelevant here, and Sydney’s threats are heat, ultraviolet light and the sea. That difference has practical consequences. Chemical consolidants developed for European sandstone penetrated much less deeply into Sydney’s clay-rich stone in tests by CSIRO and UTS, and couldn’t be relied on to hold a flaking surface to the sound stone behind. Today conservators prefer to draw the salts out where they can, because that keeps original stone in place.
Fort Denison shows how unforgiving this can be, and that there is no single best stone. There, most yellowblock replacement stone failed in the salt spray before 2007. Since then the repairs have used a quartz-rich stone with little clay from Wilton, near Appin, and it has lasted. Precisely the softness that makes yellowblock a joy on a Macquarie Street cornice makes it a poor choice in the harbour.
Yellowblock can no longer be ordered from a working quarry in any quantity, and Sydney’s old buildings keep needing it. In the 1990s stone from Western Australia and Queensland was tried and, according to the Government’s own account, “was never a suitable match”. So the city found its supply beneath itself.
It began almost by accident. In 1985 construction of the Maritime Services building in Kent Street was held up for nearly two years so that 800 cubic metres of yellowblock could be taken from the site. In the mid 1990s, Public Works staff noticed that basement car park digs in Pyrmont and the city could be harvested with excavator-mounted cutting wheels. From late 2000 they spent eight months lifting 4,000 cubic metres from the old Paradise quarry itself, at Jacksons Landing, and that stone went on to repair the Great Synagogue, St Mary’s, the university and the Town Hall.
Since 2008 the City of Sydney has required developers who excavate to test the rock and, if it’s good enough, cut it out in usable blocks. A State stockpile held about 3,500 cubic metres at last report. In 2023 a replacement unicorn’s head for the coat of arms at the old Darlinghurst Gaol was carved from stone dug out at Alfred Street, Circular Quay, and masons expected it to oxidise to match. So every new tower digging its basement in the city is potentially also a small quarry for the buildings around it.
New building stone still comes from the wider family, mainly from the upper Narrabeen Group around Gosford, more than 50 kilometres north. It’s excellent stone, but it’s not the same stone.
Sydney sandstone has now been in continuous use for more than 200 years, and it is still quarried. In 2015 four geologists and conservators formally nominated it as a Global Heritage Stone Resource, the first such nomination from Australia.