Before the Line Broke

No edition number tells us whether a sheet of Hokusai’s Great Wave was the hundredth or the thousandth pulled. The picture offers a different kind of sequence. Beside the enormous crest, a narrow blue rectangle encloses its Japanese title. On some surviving sheets the frame is continuous. On others, little stretches of its double border have gone missing.

Those gaps preserve damage to the printing surface. Once a thin ridge of wood broke away, subsequent impressions carried its absence onto the paper. A detail much smaller than a boat can help put the boats, mountain and wave into order.

The sheet above belongs to the Met in New York. Its intact title frame places it among the earliest surviving impressions in a study by British Museum scientist Capucine Korenberg. It cannot tell us which sheet came off the blocks first, or how many were made before a line broke. It gives us a point before the damage, within a history of repeated work.

A view for the shop

Katsushika Hokusai was in his seventies when he designed Under the Wave off Kanagawa, around 1830–32. He was working in Edo, the city now called Tokyo. The picture belonged to his series Thirty-six Views of Mount Fuji, commissioned and sold by the publisher Nishimuraya Yohachi, whose firm was known as Eijudō.

Fuji could be seen across the series from roads, waterways, settlements and places of work. Here it is a small, snow-capped triangle below a much larger form in the foreground. Three oar-driven cargo boats pitch among the waves. They are of a type used to deliver fish to Edo’s markets, although the image does not establish the contents of these particular boats.

The series offered a new attraction in its colour. Prussian blue, an imported synthetic pigment, gave printers a strong blue that could remain vivid where other colourants faded. Eijudō advertised the Fuji series as blue prints. The Great Wave also employed greys, yellowish boat colours and clouds whose pink tones have often become faint.

Eijudō was commissioning a product that could be sold repeatedly. The designer supplied the drawing, but the publisher needed cutters to turn it into printing surfaces and printers to turn those surfaces into saleable sheets. Their names have not been securely identified for the Great Wave. Their work survives in the edges of the foam and in the way one blue passes over another.

A drawing on thin paper guided the cutting of the principal outline block. The carver removed wood around the lines that needed to print, leaving narrow ridges standing. Proofs could then carry the designer’s indications for colours, which required further carved surfaces. Cherry wood was commonly used for this kind of work: it had to hold small details while enduring repeated wetting, brushing and rubbing.

No original blocks for the Great Wave survive. Their arrangement has to be reconstructed from the sheets. Korenberg’s examination indicated at least seven carved printing surfaces for the early design, probably distributed across four boards with several carved on both sides. A printing “block” can therefore mean one working face of a board rather than a separate lump of wood.

Rubbing colour into paper

The printer brushed pigment, water and a little rice starch paste onto the raised surface. Dampened paper was laid against it, its position controlled by registration cuts at the block’s edge. Rubbing the back with a circular pad called a baren transferred the colour. The paper had to return to the right position on the next surface so that the boat, its passengers and the surrounding water agreed.

That left considerable judgement with the printer. The amount of colour, its distribution on the wood and the pressure of the pad all affected the result. A gradation in the sky could be worked into the application of pigment. The paper needed to survive repeated handling and rubbing as its colours accumulated.

Tokyo National Museum’s film below follows a modern demonstration by the Adachi Foundation. It makes the movement of block, brush, sheet and pad easy to follow. The craftsman is using reproduction blocks; these are not tools preserved from Hokusai’s workshop.

A closer look · 4:05 · English on-screen text

Printing the wave, one layer at a time

Play film

The blue itself was more complicated than the arrival of a single new pigment. Scientific analysis at the Met found mixtures of Prussian blue and indigo in the dark outlines. In the hollow of the great crest, a further printing of pure Prussian blue passed over already printed dark stripes and filled the spaces between them. Overlap produced a deeper blue while the single application remained brighter.

Printing also changed the surface of the sheet. Three-dimensional microscopy distinguished untouched paper, areas printed once and areas printed twice at different heights. The darkest passages had been pressed farther into the paper. A photographic reproduction retains the colour relationships while flattening this small relief left by the work.

When a line disappears

Korenberg gathered photographs of 111 original impressions from museums, libraries, dealers, auctions and other sources. She compared defects that recurred in the same positions. A pale or interrupted line on a single sheet might result from uneven pigment or pressure. A break repeated across a group of impressions was stronger evidence that part of the wood was missing.

Three recurring breaks in the title frame, a break in Fuji’s slope and another in a wave line supplied a sequence. Changes to the colour blocks extended it. Korenberg distinguished eight states: an analytical ordering of surviving impressions, rather than eight dated or numbered editions. Twenty-six sheets could not be placed securely because the relevant details were unclear or had been retouched.

The sequence resisted a simple account of steady deterioration. Some lines became progressively less distinct, but the diagnostic breaks appeared abruptly. There was no succession of thinning borders before one double break. Damage may have occurred between printing sessions. The relative importance of printing pressure and the ageing of the wood remains uncertain.

Even a gap could become smaller as a ridge wore down. If a loss was wider at the top than farther inside the wood, lowering the surface exposed a narrower missing section. The visible break was evidence to be compared with other features, not a clock that could be read on its own.

New wood, changed water

Late in the sequence, the makers introduced replacement surfaces for the light blue of the sea and the yellow of the boats. These did not reproduce all the earlier forms. In Chicago’s later impression, the pale-blue shapes are more angular; the yellow printing leaves out parts of the boats that had previously received colour.

The reason for replacing those surfaces is unknown. They may have been damaged or lost. Korenberg raised the possibility that economic difficulties or a change of publisher played a part, but the sheets establish the altered tooling more securely than its cause. Still later impressions used a new surface to print the bodies of the boats in brownish pink, replacing the earlier combination of grey and yellow.

Colour adds another history to the wood. Exposure to light can diminish the clouds and boat colours while leaving the sea comparatively strong. Some sheets have also been retouched. Korenberg noted that the pink clouds in the Chicago impression may have been reapplied during a later intervention. Its present sky cannot simply be treated as a printer’s untouched decision from the 1830s.

Elsewhere, repairs altered the very evidence used to order the sheets. In another Met impression, missing sections of the title border had been drawn back in. Wear in other parts of the picture gave away the inconsistency; examination under magnification confirmed the retouching. A border could look young while the rest of the sheet recorded older blocks.

The distinctions between impressions were already attracting attention when Japanese prints entered Western collections. At a London auction in 1909, Frank Gibson bought a Great Wave for the National Gallery of Victoria for ten shillings. Another impression from the same Happer collection, praised in The Times, went to Charles Davis for £23 10s. The Melbourne sheet would later appear among the early impressions in Korenberg’s study. The two purchases had been made more than a century before that reconstruction of the blocks’ working life.