One of the enduring memories of my childhood, spent partly in London, is walking in smog, then commonly known as a pea-souper. The combination of dense fog and smoke was so thick I could barely make out street lights, and the streets were for once almost empty, as vehicles could only proceed at walking pace.

This weekend I present a selection of paintings of mist, fog and maybe even a touch of smog on the River Thames, in and near London. Today’s paintings come from the pioneers of the nineteenth century, and tomorrow’s from the twentieth.

Many of JMW Turner’s greatest paintings take advantage of the optical effects of mist and fog. Being a Londoner, he must have experienced these all too frequently.

These peaked in Turner’s famous painting of a Great Western Railway train crossing the River Thames at Maidenhead: Rain, Steam, and Speed, exhibited at the Royal Academy in 1844. The whole image is fogbound and vague, and proved a precursor to the approach of the Impressionists after his death.

Less than thirty years later, when he was taking refuge from the Franco-Prussian War, Claude Monet’s The Thames below Westminster (1871) is less Impressionist. Painted from the Embankment to the north of Westminster Bridge, near what is now Whitehall, the three towers to the south are almost superimposed, and aerial perspective is exaggerated by the mist. The river is bustling with small paddleboat steamers. In the foreground a pier under construction is shown almost in silhouette. Small waves and reflections on the river are indicated with coarse brushstrokes, suggesting this is a rapid and spontaneous work.

A decade later, The Houses of Parliament is Winslow Homer’s faithful representation of the Palace of Westminster when viewed from the opposite bank of the Thames, to the north (downstream) of the end of Westminster Bridge. The tide is high under the arches of Westminster Bridge, and small boats are on the river. This classic watercolour makes an interesting contrast with Monet’s later oil paintings I show tomorrow: Homer provides little more detail, the Palace being shown largely in silhouette, but works with the texture of the paper and careful choice of pigment to add granularity. He provides just sufficient visual cues to fine detail, in the lamps and people on Westminster Bridge, and in the boats, to make this a masterly watercolour.

The following year, Jules Bastien-Lepage paid a return visit to the city, when he painted The Thames, London. This view of industrial docklands further downstream maintains detail into the far distance, except where it’s affected by the smoky and hazy atmosphere typical of the city at that time. It was this section of the river that was also painted on several occasions by James Abbott McNeill Whistler.

Tom Roberts’ Fog, Thames Embankment (1884) is painted from a similar location to Monet’s The Thames below Westminster above, on the Embankment to the north of Westminster Bridge, but is cropped more tightly, cutting off the tops of the Victoria and Elizabeth Towers. The Palace and first couple of arches of Westminster Bridge appear in misty silhouette, with moored barges and buildings on a pier shown closer and crisper. He renders the ruffled surface of the river with coarse brushstrokes, different from those of Monet.

Among six paintings that Camille Pissarro started work on during his visit to England in 1890 was this view of Charing Cross Bridge, London from Waterloo Bridge. For this he made a sketch in front of the motif, then following his return to his studio in Éragny he painted this in oils. This looks south-west, towards a skyline broken by the Palace of Westminster and the familiar tower of Big Ben.

In Frederick Childe Hassam’s Houses of Parliament, Early Evening (1898), the sun has already set, and he is viewing the Palace in the gathering dusk from a point on the opposite (‘south’) bank, perhaps not as far south as Lambeth Palace. The Victoria Tower is prominent in the left of the painting, the Central Tower is in the centre, and the most distant Elizabeth Tower is distinctive with its illuminated clock face. Moored boats in the foreground provide the only other detail. His rough facture gives a textured surface to the water.

Many of the greatest paintings succeed in part because they use optical effects in their paint layer. This takes advantage of the fact that thinner layers of paint aren’t completely opaque, so allow some light to pass through them. In watercolours, transparent paints are often referred to as transparent watercolour, while those that are opaque are known as gouache or body colour. In oils, different pigments result in degrees of opacity, expressed as their covering power.

This cross-section of the paint layer from Honoré Daumier’s The Strongman (c 1865) demonstrates how many oil paintings consist of a series of layers, with several pigments present through their depth. Generally, the most opaque are applied first, and on top of those come a series of thinner transparent layers or glazes. Together these allow complex combinations of reflection and refraction of light, generating optical effects.

Layered oil painting technique is best seen in paintings that have been abandoned before completion.

The British portrait painter Sir Joshua Reynolds received a conventional training in traditional and conservative methods with roots dating back to the late 1600s. He painted in layers, starting with dead colouring, the laying in of shadows and lights, then blending in transitions of shading and colour wet-on-wet. Highlights were then brought out, and shadows glazed, to produce a series of thin layers of oil paint, and a smooth, finished paint surface.

Those early stages are shown well in this abandoned portrait of Mrs Robinson from about 1784, where most of the paint layer is sufficiently thin as to allow the texture of the canvas to show through. When used with ‘lean’ paint, this dried quickly and complies with the longstanding edict of applying ‘fat’ over ‘lean’, so that the lowest layers dry first.

Richard Parkes Bonington’s unfinished Venice: The Piazza San Marco (1827-28) shows signs that it might have been among his best. Its buildings have a golden glow from the setting sun, but those colours would undoubtedly have been enhanced by rich glazes had he lived long enough to complete it.

Occasionally the paint layer develops problems that demonstrate the effects of glazes.

In Abraham Bloemaert’s The Adoration of the Magi from 1624, the cloak of the Virgin Mary appears to use two different blues, with its lower passages painted in the duller hue of indigo, which has faded. The dullest areas are those that had the thinnest ultramarine glazes applied, much of which have now abraded away during subsequent cleaning of the painting. The unprotected indigo has therefore suffered sufficient exposure to fade, as well as losing those rich glazes.

JMW Turner’s Approach to Venice (1844) was painted with very thin transparent glazes over thick white impasto, creating a distinctive flickering effect in its highlights.

Despite the artist’s efforts to get the white impasto to dry more quickly, the glazes dried first, and cracked as they became stressed over the white that was still wet. This hasn’t been helped by the later conservation process of lining, which places an additional layer on the back of the canvas to help the support do its job.

Although a wide range of pigments have been used successfully in glazes, some were developed specifically for the purpose. From the fifteenth century onwards, verdigris pigment was mixed with natural resins for use in glazes. This produces a different pigment from regular verdigris, as the copper combines with the resin acids to form what is known as copper resinate.

A popular technique among many masters to produce an intense green was to paint an underlayer using verdigris, over which several glazing layers of copper resinate were then applied. Although generally reliable and stable, verdigris and copper resinates have a tendency to turn brown on the surface. Thankfully this affects relatively few paintings.

Tintoretto used copper resinate glazes in several of his paintings, most notably the rich, varied, and often lush vegetation in his Saint George and the Dragon from about 1555.

Studies at the National Gallery, London, have found copper resinate in three of the four paintings in Paolo Veronese’s series The Allegory of Love. In the third of these, Respect (c 1575), the pigment was found in the man’s intense green cloak, and the duller gold-brown brocade patterning on the wall behind his hand (detail below). The surface of that wall has superficial brown discoloration of the paint layer.

One of the last major uses of copper resinate is in Arnold Böcklin’s Triton and Nereid from 1874. This is reported as being painted in tempera, but copper resinate glaze appears to have been used to develop the intense green patterns on the sea monster in the foreground. This is consistent with Böcklin adhering to traditional techniques despite working in the late nineteenth century.

During that century, painting slowly in multiple layers with glazes progressively fell from favour. By the end, many oil painters had adopted techniques in which much or all of a painting was made at once, known as direct, alla prima or au premier coup.

Because most fish aim to spend their entire lives underwater, where few artists go to paint, fish are seldom seen in paintings. That contrasts with those who try to capture fish by going fishing, an activity I have previously covered in this series in this article and a second.

Most of the aquatic creatures seen in paintings of myths, including those accompanying the god Neptune, appear to be caricatures of marine mammals including dolphins, or sea-monsters bearing no resemblance to fish.

One exception to this is Joseph Stella’s The Birth of Venus from 1922. As might be expected, his treatment is completely novel and seems to have benefited from visits to an aquarium. Aphrodite is shown at sea, in the upper part of the painting her upper body above the waterline, and below morphing into an aquatic plant underneath, where it finally merges into a helical shell. Matching the birds and flowers above the water are brightly coloured fish below.

Another interesting exception is Joachim Beuckelaer’s depiction of water in his Four Elements cycle from 1569. This shows A Fish Market with the Miraculous Draught of Fishes in the Background, the one place even landlubbers would come across fish, combined with the Gospel story in the far distance.

In the summer of 1872, as a one-off, Gustave Courbet painted an allegorical still life of The Trout, that is “hooked and bleeding from the gills”, a powerful expression of his personal feelings after being imprisoned for damage to the Vendôme Column during the Paris Commune the previous year.

JMW Turner recruits a school of fish for effect in his Slave Ship from 1840. His threatening sky and violent sea put the ship in the middle distance, silhouetted against the blood-red sky. The foreground is filled with the ghastly evidence of the slaves who were cast overboard.

Seen in amongst a feeding frenzy of fish and scavenging seabirds are hands raised from the waves in their final plea for rescue, a gruesome manacled leg, and various shackles used to restrain the slaves when in transit. Further back on the left a vague white form could represent spirits, and on the right is the thrashing tail of a sea monster.

Fish make the occasional appearance alongside legendary mermaids, as in Hans Thoma’s Three Mermaids from 1879. These mermaids are remarkably human in form, lacking fishtails, and frolic with fish under the light of the moon.

Historically the most important fish in Europe has been the humble herring. In the Middle Ages herring fisheries prospered and were the foundation of Copenhagen and Great Yarmouth, and influential in early Amsterdam. They remain strongly associated with the Netherlands and Nordic countries, where they are commonly preserved in brine (soused) or pickled.

Gabriel Metsu’s Woman Selling Herring (c 1661-62) is going from door to door with her fish, here trying to convince an old woman standing with a stick at the door of her dilapidated cottage in the Dutch Republic.

In Hans Andersen Brendekilde’s Home for Dinner from 1917, a young girl holding some fresh fish stands talking to a man with a spade.

JMW Turner toured the West Country as far as Cornwall in 1811, and the Tate Gallery has his ninety-page sketchbook recording many views of the Cornish coast from that visit. He later developed several into fine oil paintings, although it’s unclear whether this watercolour of St. Mawes, Cornwall, from about 1823, had its origins in those sketches and studies.

As with his paintings of other coastal areas, Turner shows a fishing boat coming in to a beach to land its catch, and the great activity in the open air fish market in the foreground. Behind are typical Cornish cottages stepped up from the shore to the top of the coastal cliffs, and the castles of St Mawes (closer) and Pendennis, in Falmouth (more distant, on the other side of this estuary).

Richard Dadd’s Fish Market by the Sea, from about 1860, shows an impromptu open-air fish market, run by the fishermen’s wives, to sell their husband’s catch as soon as it had been landed.

Long before paintings became movable objects of great value used by the rich as investments, artists and the owners of their paintings wanted to protect the paint layer that had been so carefully applied to the ground. From the early Middle Ages onwards, one popular means of doing this has been to apply some form of protective layer, a varnish.

Varnishes have been widely used not only for protection. Careful choice of their composition can enhance the appearance of a painting, through the optical properties of the varnish medium and its smooth, glossy surface. Until the late nineteenth century, the great majority of painters either applied final layers of varnish themselves, or advised their patrons and clients to do so.

Three types of varnish have come into common use:

• Drying oil and resin, in effect a resin-rich transparent and unpigmented paint layer, that usually becomes an integral part of that. Some artists have added pigment, perhaps to make a general colour adjustment. There isn’t any clear distinction between that and a final paint glaze.
• Solvent and resin, from which the solvent will evaporate, leaving a thin surface coat of resin.
• Water-based washes such as egg white, known as glair, vegetable gums like gum arabic, and animal glues.

Resins used in varnishes have rich and sometimes strange histories. Most are exudates from trees in exotic locations, and have evocative names like mastic, sandarac, colophony and dammar. They’re usually highly insoluble, either in drying medium that has to be heated to make oil-based varnishes, or in turpentine or similar organic solvents. A great many recipes have been proposed, and there’s always the lure of the perfect, and inevitably top secret, formula.

The biggest problems with varnishes are their propensity to yellow or grey with age, and their tendency to take up dirt and atmospheric contaminants. Rembrandt’s first painting of Bathsheba at her Toilet from 1643 has sadly lost much of its detail into the gloom of old varnish, which can be almost impossible to clean off when composed of drying oil and resin, without damaging the paint layer underneath.

Any work older than a few decades that has been varnished or had any form of surface treatment is unlikely to appear today with the colours the artist intended. Multiple layers of old varnish and trapped dirt give a misleading impression of what we would have seen soon after the work was completed. Painstaking work by conservation specialists can often restore old paintings to what we presume is their former glory, in full colour again.

Hieronymus Bosch’s Christ Carrying the Cross (1490-1510) is seen above before recent conservation work, and below is the result of thousands of hours of painstaking cleaning and treatment.

Another problem for the conservation specialist is a painting like Edward Poynter’s A Visit to Aesculapius from 1880. Although this is little more than a century old, the evidence from contemporary prints made from this work is that it was originally far from being so dark. Sadly it’s now almost impossible to read as a result of its near-black shadows.

A good varnish should be both colourless and transparent, but painters haven’t always respected that.

When finishing his monumental Raft of the Medusa in 1819, Théodore Géricault is thought to have applied glazes or varnish containing asphalt to give the painting a deep brown tone. Asphalt is not only completely unprotective and almost attracts dirt, but it never fully dries, and can have adverse effects on underlying paint too. It hasn’t helped that this two hundred year-old painting was rolled up and stored in a friend’s studio when it remained unsold, and was then transported to London still rolled up the following year.

Conventional wisdom says that it’s best to leave an oil painting to dry for at least six months before varnishing it. JMW Turner sometimes varnished over paint layers that were far from dry. In the case of The Opening of the Wallhalla, 1842 (1843), painted on mahogany, Ruskin reported that it had “cracked before it had been eight days in the Academy Rooms”, although this overall view shows little evidence of that damage.

Hellen and Townsend attribute this to Turner’s extensive use of Megilp, here a product sold by his colourman containing leaded drying oil and mastic varnish. Used sparingly and with great caution, such medium modifiers don’t necessarily cause serious ill-effects. But Turner has used Megilp to excess, to produce a soft impasto used in the foreground figures, in particular. This has resulted in wide and shallow drying cracks, as the surface has dried quickly and shrunk over trapped layers of liquid paint.

Varnishes do provide mechanical protection to the paint layer, but at the cost of locking out atmospheric oxygen, required for drying oils to polymerise properly in their drying process. Applied too early, varnishes can therefore greatly slow drying of underlying paint layers; the danger is that they may saponify (turn to soap) instead of drying normally.

Despite these dangers, varnishes can, when used with care by those who understand them properly, be valuable beyond simply providing a protective coat. Kirsty Whiten’s The Quing of the Now People (2015) achieves its superbly realistic effect by the skilful combination of conventional oil paint with varnish.

In the late nineteenth century, attitudes to varnishing oil paintings changed markedly, as Impressionists like Camille Pissarro started to prescribe that their works should on no account be varnished. This was to preserve the soft matte surface of the paint as applied by the artist, and became increasingly popular in the twentieth century.

For such paintings, protection can be provided by glass, when necessary. That isn’t of course an option for many extremely large oil paintings on canvas, which will probably need to be varnished and cleaned periodically well in the future, as they have in the past.

Varnishes, usually of the third type containing vegetable gums or animal glues, have also been used extensively on paint layers other than oils.

These are reported in Samuel Palmer’s Tintern Abbey at Sunset, above, and Dante Gabriel Rossetti’s Lucrezia Borgia, below. Gum or glue varnishes can have impressive optical effects when used carefully on watercolours.

Unfortunately, their tendency to yellow can also cause colour shifts. William Blake liked to apply glue varnish to his watercolours and perhaps to his glue tempera paintings as well. In the case of his Penance of Jane Shore in St Paul’s Church from about 1793, this has resulted in a generalised yellow shift and loss of chroma.

Other artists appear to have been more successful: Henry Ossawa Tanner apparently applied varnish to this tempera painting of The Destruction of Sodom and Gomorrah almost a century ago, and it doesn’t appear to have suffered any adverse consequences, yet.

Varnishing has become such an accepted process that major exhibitions have incorporated ‘varnishing days’, although what happens on those occasions can be quite different. In Turner’s day at the Royal Academy in London, Varnishing Day was an occasion for artists to make any last-minute changes, and Turner himself seems to have turned up armed with paint and brushes and continued to work on his paintings.

Varnishing Day in the Paris Salon was completely different, attended normally by the artists’ colourmen, who applied a coat of varnish to the paintings for which they were responsible. The artists themselves don’t seem to have been involved, unless they chose to apply the varnish in person.

We’re remarkably good at perceiving different surface textures, but find it harder to imagine them in two-dimensional images. While the overwhelming majority of paintings, at least until the twentieth century, consist of a paint layer on a flat ground, there’s nothing that requires the surface of the paint layer to be flat and smooth. But if all you look at are images of paintings, you generally won’t see their surface texture, where the artist has applied and shaped paint in thick layers of impasto, or incised into some of the layer in sgraffito. This article looks in detail at four examples where surface texture in the paint layer is important.

Arnold Böcklin’s unusual painting of Sirens from 1875 was made in tempera on canvas, with the ground and paint layers thin enough to let the texture of the canvas weave show through. This image was fortuitously taken with lighting that allows the texture to show.

Another famous tempera painting wasn’t painted on a textured ground, but is one of the earlier paintings to feature impasto as a technique.

The Wilton Diptych was painted on two small panels of oak wood in the final years of the fourteenth century. That wood was first assembled into the panels, then carved down from a thickness of about 2.5 cm (1 inch) to form an integral frame with a recessed painting surface. A smooth gesso ground was then laid on the wood before the gilded areas were laid onto it using thin sheets of gold leaf, and patterned using a range of punches.

Details of jewels and similar objects such as the white hart brooches were raised using thicker areas of lead white paint, to give the impression of enamelling. Coupled with mordant gilding, they mimic the three-dimensional form of jewels and act as point reflectors of light, sparkling as if they really were gems in the paint layer, as shown in the details above and below.

The finest strokes of paint seen here are less than 0.5 mm (1/50th of an inch) across.

From those early days of modern painting until the decline of ‘academic’ painting in the late nineteenth century, patrons, Salon juries and critics expected paint surfaces to be smooth and flat. But there were rebels.

Many of Rembrandt’s paintings from before 1650 have fairly conventional ‘finished’ surfaces, his monumental Night Watch being a good example. By about 1660, though, many of his paintings had quite rough surfaces that significantly alter their optical properties.

One of the best examples of Rembrandt’s use of texture in the paint layer is The Jewish Bride from about 1667, just a couple of years before his death. This is among his works studied by the Rembrandt Research Project.

In this detail, highlights on the sleeve and jewellery have been applied roughly, although it’s still a matter for speculation as to exactly how he achieved that. Lower down, on the red dress of the bride, the duller top layer of paint has been scraped through to reveal lighter lower layers. The end result is a painting that creates its visual effects as much by its surface textures, as by form or colour.

One of Rembrandt’s secrets that have been sought by so many since lies in how he was able to exploit surface texture in his paint. That is the ‘secret recipe’ which Maroger, Redelius, and others claimed to have discovered. Systematic analyses of Rembrandt’s paint layers by White at the National Gallery in London and the Research Project soundly rebutted the ‘secrets’ claimed. In the main, Rembrandt used linseed oil as his binder, occasionally using walnut oil as well, and just once poppy seed oil.

In some passages the oil had been thickened by heat treatment, but this was by no means widespread. Traces of pine resin found in some samples may have been introduced during retouching, and don’t appear to be a feature of Rembrandt’s impasto work; neither is there any evidence that he added wax to his oil paint to give it body, as some had asserted.

As you might expect, JMW Turner was another such rebel.

For example, in his Seapiece with Fishing Boats off a Wooden Pier, a Gale Coming In, possibly from as early as 1801, Turner made extensive use of sgraffito, made using a knife, brush handle, or even his fingernails for all we know.

Later that century, Vincent van Gogh developed a more radical approach, in his initial version of Wheat Field with Cypresses from 1889, the year before his death.

It’s possible that he completed this painting in a single sitting, as this seems to have been intended as an oil sketch for a more finished version which he painted later that summer.

The detail above shows the tops of the wheat towards the lower left of the field, in the foreground. Over his initial thin layers of paint, van Gogh laid thick gestural strokes of highly chromatic paint, orientating those strokes according to the object they show. In the golden yellow of the wheat there are blues and greens, mostly showing through from his underpainting, with superimposed impasto of pale straw, ochre, and pale greens.

This detail, taken from the edge of the wheatfield at the lower right corner of the painting, shows three distinct areas of brushwork: the diagonal strokes forming the standing wheat, swirling loops to form the grasses and weeds below, and shorter marks forming a more random pattern for the heads of the wheat in the upper section.

At the centre of the canvas, from where this detail is taken, impasto blue and white have mixed with the green and yellow of the fields below. This shows that much of the painting was painted wet on wet, either in the same session or on consecutive days. Some of the darker green at the right may have been painted later, onto paint that had by then become touch dry.

An X-ray image of the whole painting shows in white those passages likely to contain the most lead white, and some other pigments which are most radio-opaque. This also reveals the pattern of brushstrokes well.

With the introduction of acrylic paints in the latter half of the twentieth century, painters have been able to apply even heavier impasto, and some have used this to paint what are in effect reliefs.

This is why so many paintings have to be seen in the flesh, up close, and in the right light for their full appreciation.