DANAE THE MISSING PAINTING

Danae was a painting produced at an early stage during J.W Waterhouse's period of high key theatrical paintings based on mythology from Homers Odyssey between 1889 and 1897. Waterhouse in 1893 said that he would rather be remembered for Danae than any other painting.

This is a very unusual composition for Waterhouse. The focus of the work is offset to the right of the image countered only by an area of deep water and a distant island on the left. Though the fishing net that floats to the left and the lines of wave tops do draw the viewer back to the foreground, a compositional device he later reused in Miranda. The closeness of the focus to the viewer compensates for this aspect of the composition. To some extent the weakness of the left contributes to the intensity of focus in the foreground right.

In general Waterhouse's mid career paintings are so strongly designed that they can be viewed by looking at small areas of the painting that each have a self sustaining composition. The composition in this case shows that many of the aspects of the composition of Ulysses and the Sirens have influenced the design. Thrusting the focus at the viewer and placing them in a relationship to the painting that infers they are within its boundaries, is one such aspect of the design developed from Ulysses. The composition in this case is far simpler and one point focussed, Ulysses on the other hand can easily be divided into seven or eight smaller paintings each of which could stand alone.

How effective this composition is has been difficult to understand as the painting was stolen in 1947. Only a black and white photo reproduced in magazines from 1892 to 1909 exists to show what it looked like. I determined to reconstruct the painting to investigate how this simple composition worked.

The following process shows how I proceeded to construct a coloured image that was as close as practicable to the original. The details I had to start with were the size of the painting and the black and white photo. I knew that the painting was produced in the same year as Circe Invidiosa and hung in the year following the hanging of Ulysses and the Sirens at the Royal Academy.

I made a number of assumptions to start with to help in directing my research.

• Waterhouse had been to Capri three years before painting Danae, his colour use and subject matter had been influenced by the colours of the island, it was likely that this would also be true for Danae.
• It was likely that Ulysses and the Sirens, Circe Invidiosa, Danae and Circe Offering the Cup to Odysseus, were all completed using a very similar range of colours as well as paint handling. These paintings appeared in 1891, 1892.
• Ulysses and the Sirens and Circe Invidiosa were both in Australia and able to be viewed.
• The paint and colour notes provided in recent Waterhouse monographs are accurate. A safe assumption as Peter Trippi (the most recent and thorough of Waterhouse's biographers) has engaged the foremost modern conservators in determining colours and how they were applied by Waterhouse.
• The preparation of the canvas and initial painting followed the same model as the later Lady of Shallot. Three oil works are available to trace the development of this painting as well as early design sketches.
• I would not use linseed oil, stand oil or turps in producing the painting but would stick to my current materials, Archival paints and mediums.
• To force myself to be as focussed as possible I would use linen as the cost would not allow me to feel free to just stop/ give up.
• Waterhouse applied colour direct and did not rely on monochrome underpainting to glaze over. This is shown in samples displayed in Peter Trippi's book. It is confirmed by the incomplete paintings that were in his studio at the end of his life, further evidence is given by the three oil versions of Lady of Shallot that are contemporary with Danae.
• Waterhouse did glaze over in very fine areas to provide translucence and reflective depth, also to achieve subtle variations within small spaces. This is particularly evident in the painting of eye sockets, where transitions are incredibly subtle.

So where to start? Find prints of Ulysses and the Sirens and Circe Invidiosa. Easily done by visiting the NSW State Gallery and taking a side trip to look at Diogenes, an earlier work of the 1880s but that reflected some of the later use of thick paint. The prints look odd when compared to the images in books, there is a definite lack of true colour in the prints… equally the books are not renowned for being accurate colour reproductions either. There is only a limited set of choices but I take the most obvious and go to Melbourne for a few days to see Ulysses and the Sirens up close.

In Melbourne I photograph the painting in great detail, to the point where the attendant wanted to know what I'm doing (I included the framing in my photography which looked odd to them). Without a doubt I'm now completely hooked. Ulysses shows signs of incompleteness, changes in canvas size, later paint colours in areas that don't quite match the previous colours, added figures and interesting points like the artists signature being partially painted over by the eyes on the bow of the ship and unplugged nail holes showing where the initial stretcher size had been.

I'm already carrying information from earlier research, such as,  the wings and bodies of the Sirens are eagles at London Zoo that was close to where Waterhouse lived. The concept of the winged Sirens being taken from a Greek pottery, though this is surmised, French artists also used the bird form of the Sirens without ready access to the pottery in the British Museum. The 13 man ship being way too small as most of that small number were lost shortly after passing the Sirens. The face that can be made out on the stern is a mystery. The rollicks are lions faces that Waterhouse has used in a number of paintings and will use again in Danae. There is a use of Vermillion in small areas to draw the eye, from observing Diogenes I think he does this to break up longer lines and reduce the awkward effect they can have on what the eye perceives.

What I add to this is that Vermillion is used along the waterline in the cliff behind for much the same reason as before but I start to see a rhythm in the placement of this colour. This appears to be a quite advanced use of small areas of colour to draw the eye to points on the canvas without exposing a large area of colour interest. In my observation of Diogenes I noted that Waterhouse breaks lines that appear to go too far without interruption and uses small details such as broken twigs under the heals of feet to ensure no jarring angles are present. This type of attention to detail is not present in the work of other artists of the period, nor have I noted this emphasis in later works by other artists.

I'm beginning to appreciate that the task I've set myself is far from trivial. What I gain from Ulysses is:

• An exact match for the lions heads on the box in Danae, though one has to move through 90degrees and have an alternate light source.
• Skin colours for the men. Though this may need to be altered because of different lighting effects.
• Skin colours for the female, but this also derives from Circe Invidiosa and all are in full light, where as the females face in Danae is lightly shaded.
• An understanding of how Waterhouse represented the shingles on the ship and these are probably dealt with in the same way in Danae. What material they are supposed to be made of is uncertain.
• Colours for the deep water can be extrapolated from Ulysses but no shallow water is present, nor is any distant water on the horizon.
• The application of paint, the brush stroking and wetness/ dryness of paint can be gathered. Though this is not reflected from Ulysses to Circe, so I can't be overly guided by the this, if anything it may be a parameter to paint inside.
• How the female hair was painted can be drawn from the female figures. This is however sketchy in approach as none of them are the major figure. Circe Invidiosa is a better guide in this regard. Waterhouse had to guess at what this would look like and is near to failing to carry the illusion.

I take the black and white photo as a value guide. It will become a final arbiter of how correct my colouring is at a tonal level. The black and white photo is used to develop the under drawing. Using a video projector the image is displayed as it appears on my computer. This projection method sounds as though it should be perfect for positioning and detail… but it is not.

My first decision about how to approach the painting is to use a pencil outline with no underpainting. In general underpainting may help a new painting in Waterhouses style, but for a copy it adds nothing to the process.

Parallax errors and distortions resulting from the projector's lense give the drawing a number of faults that will only be found when  the painting is well advanced.

Starting to Paint

To start with the faces and flesh were roughed in, then the water and as a reference the drapery on the central figure of Danae. At this early stage I was daunted by the scale of the painting (it is full size of the original), the lack of paint near the edges is an indicator of my impatience to move ahead. The sea used colours from a very similar waters edge that appears in Miranda of a much later date, they are keyed up to carry the colours of Capri and Waterhouses fascination with these colours.

Next the drapery of the two male figures is roughed in to attempt to establish the colour relationships around the box. At the same time the hair of the female is worked on to experiment with getting the feeling of a light breeze into the painting.

The box is coloured as areas to firm up the visual composition.

Foreground rocks and detailing of the water, plus defining drapery and its details are added. The background also has work done on it. Along the way I modify the flesh colour to adapt to the changed colour relationships

By this time I recognise that the sky colour is the wrong colour and needs to be replaced. A brighter more Mediterranean colour is inserted, modelled from a Leighton painting of Capri. This makes all the other relationships wrong. The people look like corpses rather than physically active robust

A series of attempts to fix the soften the face of the female are undertaken. The hair is reddened. The horizon line is raised and the sea colour changed. The painting is exhibited in this state, but I'm not happy with it.

The sea colour changes again. The hair colour of the female changes and the face is repainted for the sixth time. The skin colour of the men is changed to be closer to the Ulysses males, this is done by glazing the complete skin area of the figures. Feet and hands are reworked. The lion faces are reworked, the box is also reworked to tonally fit the original photograph. The female has the entire figure repainted to improve the arms and torso. Small detail changes are made to the baby.

Struggling to complete the work in the final details. The lower male figure has his clothes repainted. Some detailing of the rocks to improve 3 dimensional look. The frame is now included to show the end work.

This is how the face of Danae looked before I reworked the face to incorporate my study of Waterhouses method of portraying faces. It also shows the problems inherent in using small photos as reference, the position of the head reflects my natural desire to bring the face more upright than it actually was in the original painting.

These changes to the face incorporated my more developed understanding of Waterhouses technique in this area. I had been working on a direct copy of Gather Ye Rose Buds 1908 and finding a great deal of interest in multiple layers of glaze and using multiple soft brushes to achieve effects. In addition I found that Waterhouse taught me a great deal about the shape and form of a face and how this can be represented by the use of colour… tonal representation of a face will always look plastic and unalive.

I'm still disappointed with this figure and will probably revisit it to get a better feal of the position of the head. Hands were also reworked to improve detail. Note the vermillion in the box, which shows well above its normal visibility in the actual painting.

Skin colour and the use of line to accentuate edges is the main reason for this rework. This change led to the work on the figure of Danae.

A cropped version of the black and white photo.

The finished painting before some tonal adjustments to come closer to the photo.

Final after adjustments the final painting in frame.

Not a prefect production but I'm reasonably pleased with the colours and the sense of drama.