Painters describe watercolor by how thick it is: tea, coffee, milk, cream, butter. Joseph Zbukvic pairs each one with how wet the paper is. Drop watery paint into a half-dry wash and it blooms into a pale blot with a dark, ragged rim. Drop thick paint and it stays put. Why?
The short answer
Water always moves from wetter to drier: from the brush into the paper, or from the paper into the brushstroke.
Drier paper pulls harder. Paper holds water in the gaps between its fibres. As it dries, the water is left only in the finest gaps, and fine gaps suck the hardest. So water flows from wherever there is more of it toward wherever there is less.
A brush wetter than the paper blooms. Water flows out of the stroke into the half-dry wash. It pushes the wash’s loose pigment ahead of it and drops it where the water stops: a pale middle with a dark rim. Painters call it a bloom, a backrun or a cauliflower.
A brush drier than the paper stays put. Water flows into the stroke instead, so it softens at the edges but doesn’t spread. Thick paint helps twice: it brings little water, and the gum and pigment make it far too sticky to run.
So the drier the paper, the thicker the paint. On dry paper anything goes: the water has nowhere to flow, and every edge comes out hard.
Drop paint into a wash
A dab of burnt sienna, 5 mm across, into an ultramarine wash at some stage of drying.
Pick how wet the wash is and how thick the paint is, then watch it dry. Watery paint on damp paper blooms. Thick paint on damp paper holds its shape. On wet paper everything spreads, and on dry paper nothing does.
Seen from above, 22 × 22 mm
All twenty combinations
Each one dried in the same model. Click one to watch it.
Every paint on every stage of paper. The blooms sit where the brush is wetter than the paper.
Rows run from wet paper at the top to dry at the bottom; columns from thin paint on the left to thick on the right. The words are the model’s own verdict on each swatch.
1The kitchen scale
The scale comes from the Australian painter Joseph Zbukvic’s book Mastering Atmosphere & Mood in Watercolor[9]. Stephen Berry’s notes on it give the five consistencies [1]:
Paint
What it’s like
Used for
Tea
Very thin and watery, not much pigment; runs freely if tilted
Skies, clouds
Coffee
A stronger mix; still runs, but leaves a wash of color
Distant hills, gentle shadows
Milk
Dries close to its color on the palette
Middle ground, foreground, dry brush
Cream
Sludgy; doesn’t bead, behaves more like gouache
Darks
Butter
Straight from the tube; won’t move even upside down
Final accents
Zbukvic sets these against four stages of drying paper: wet (glossy), moist (the shine just going), damp (no shine, cool to the touch) and dry, and lays the pairs out on a clock face [1]. The advice that runs through it: the drier the paper gets while it’s still wet, the drier and thicker the paint should be. On damp paper, too much water on the brush makes back-bleeds. Once the paper is dry, any consistency works [1]. Painters sum it up as “wet moves toward dry” [8].
2Why water moves from wet to dry
Watercolor paper is a mat of cotton fibres with gaps of many sizes between them. Water in a narrow gap is pulled along by surface tension, with a suction of 2γ/r for a gap of radius r (γ is water’s surface tension). The narrower the gap, the harder it pulls: a 1 µm gap pulls with about 140 kPa, more than an atmosphere, and a 50 µm gap with about 3 kPa.
As paper dries, the big gaps empty first, and what water is left sits in the fine ones, pulling hard. Wet paper has its big gaps full, and those pull weakly. Put wet next to damp and the damp side’s fine gaps win: water flows across. That is “wet moves toward dry”, in physics.
How fast it flows depends on how wet the paper is. Water can only move through gaps that are full, and full paper has wide open paths, so wet paper passes water along many times faster than damp paper. Soil physicists describe exactly this with the Richards equation: water spreads like heat, but with a conductivity that grows steeply with wetness [5]. Thicker liquid moves slower: in Washburn’s law for a liquid filling a narrow tube, the distance grows as the square root of surface tension over viscosity [7].
Why drier paper pulls harder
Surface tension in a gap, and how the model lets water move.
Left: the Laplace suction 2γ/r of water (γ = 72 mN/m) in a gap that it wets fully. Sized paper wets less readily, which weakens every value by the same factor. Right: the mobility this page’s model uses, ∝ w³ inside the paper and rising faster once water stands on top. The exact curve for real paper has to be measured, sheet by sheet [5].
3Blooms: water pushing pigment
Curtis and colleagues describe backruns this way: “When a puddle of water spreads back into a damp region of paint, as often happens when a wash dries unevenly, the water tends to push pigment along as it spreads, resulting in complex, branching shapes with severely darkened edges.” And they note when it happens: “Backruns occur only when a puddle of water spreads slowly into a region that is drying but still damp” [2].
Both conditions show up in the model. The paper must still be damp, because a damp wash still has pigment lying loose in its water, ready to be moved. On dry paper the pigment has stuck to the fibres, so nothing gets pushed. And the brush must be wetter than the paper, or no water flows out. The flowing water picks up loose pigment as it goes and leaves it where the flow stops, which is why the middle comes out pale and the rim dark. MacEvoy tests how readily each paint does this, which he calls blossoming: a dry wash rewetted, its pigment “moved by the capillary movement of the water as it evaporates” [4].
The ragged, cauliflower edge comes from the paper. In the model, water finds some paths through the fibres easier than others and runs ahead along them. A second cause may help: thin water pushing into thicker, gummier water is unstable, and breaks into fingers (the Saffman–Taylor instability) [6]. This model doesn’t include it.
4Why thick paint stays put
Thicker paint brings less water, so it has less to push out. But it is also far more viscous. Watercolor is pigment and gum arabic in water, and both thicken it. Packed particles raise a liquid’s viscosity steeply as they crowd together, until it stops flowing at all (the Krieger–Dougherty relation) [3]. Paint straight from the tube is a paste that holds its shape. In the model, viscosity rises exponentially with the pigment in each drop of water, so undiluted milk is about three times as viscous as water, cream about twenty-five times, and butter barely moves.
On wet paper, thick paint gets thinned by the water already there, which is why even cream softens and “melts” into a wet wash, while staying more compact than milk [1].
5Reading the clock
Wet paper takes tea and coffee for soft, spreading skies and graded washes. Thicker paint softens without running away.
Moist paper takes milk and cream for soft edges that still hold a shape. Watery paint runs further and starts to push the wash aside.
Damp paper is the dangerous stage. Only cream and butter are safe; anything wetter blooms, because the wash’s pigment is still loose and its water is scarce.
Dry paper takes anything, with a hard edge every time. The pigment underneath has stuck, and the edge of the new stroke is pinned to dry paper, which is also what makes the dark edges of page 05.
Blooms on purpose. Dropping clear water or tea into a damp wash is a standard way to make textures: foliage, clouds, rocks. The model says the size of a bloom grows with how much water you drop.
Wetting agents in some paints, such as ox gall, make them shoot outward on wet paper faster than this model shows, which MacEvoy attributes to the soaplike additive [4].
6What this model leaves out
Illustrative numbers. The amounts of water and pigment in each consistency, the mobility curve and the drying rate are chosen to behave like paint, not measured. The model is right about directions (wet to dry, pigment rides the flow) and only roughly about distances.
Flat, level paper. No tilt, no gravity, no brush pressure. Real paper also has sizing, which slows water entering dry fibres.
Mixing is a simple spreading term, fast in standing water and slow through full pores. The real mixing comes from the drop landing, currents and surface-tension differences, including wetting agents.
Sticking. Pigment sticks to the fibres at a rate that grows as the paper dries, the same for every pigment. Real paints differ, which is part of what MacEvoy’s staining and blossom ratings measure [4].
No edge flow. The evaporation-driven flow of the edge darkening page isn’t included here, and neither are sinking grains (granulation).
The math
Water. Each 0.3 mm cell holds water w, in units of what the paper’s pores hold: 1 is full, more than 1 is water standing on top. Water moves between neighbouring cells down its gradient, ∂w/∂t = ∇·(D∇w), a form of the Richards equation [5]. The diffusivity is D = k·M(w)/μ: M = w³ inside the paper and 1 + 3(w − 1)² once water stands on top; k is a log-normal factor per cell for the paper’s unevenness; μ is the viscosity.
Viscosity. μ = exp(1.6 p/w), with p the loose pigment in the cell, a stand-in for the steep rise of a suspension’s viscosity with crowding [3].
Pigment. Loose pigment moves with the water, at the concentration of the cell the water leaves. It also mixes into the water around it: fast in standing water, at a rate proportional to the water’s depth above full, and slowly through full pores, ∝ w³. It sticks to the fibres at a rate proportional to (1 − w)², so it is mobile in wet paper and fixed in dry. Water evaporates at a steady rate everywhere.
The stages. Wet: w = 1.35, all the wash’s pigment loose. Moist: 1.0, 85% loose. Damp: 0.45, 60% loose. Dry: 0, all stuck. The paints, as water and pigment added under the dab: tea 2.2 and 0.12, coffee 1.3 and 0.25, milk 0.65 and 0.45, cream 0.35 and 0.7, butter 0.12 and 1.0. The wash holds 0.25 of ultramarine.
Verdicts. Bloom: the wash’s pigment cleared by more than 15% in the middle and piled to more than 1.3× in a rim. Spreads: 90% of the dab’s pigment ends up more than 2.5 mm beyond the dab’s edge. Hard edge: the dab’s pigment falls by more than 60% of its peak within one cell. Anything else is a soft edge. Colors are Kubelka–Munk layers of both pigments on paper.