What natural latex in a sofa cushion actually is

Natural latex names an origin. It does not name a purity level. The phrase tells you the foam in a cushion began as sap tapped from a rubber tree, and then it stops.

Two facts decide what you are buying. The first is the share of the blend that is tapped tree latex. The second is the process that cured the block. A product page can carry the phrase and print neither figure.

The short answer: a percentage and a process

Ask a brand what natural latex means on its spec sheet and you should get two numbers back. One is a blend percentage. The other is a cure method, either Dunlop or Talalay.

Without the percentage, the phrase covers a wide range of compounds. Without the cure method, you cannot predict how the block carries weight. This page takes each in turn, then sets out what Covelle publishes about its own latex core and what it leaves unpublished.

Where does the latex in a cushion actually come from?

A shallow cut is scored into the bark of a Hevea brasiliensis tree. Sap runs down the cut and collects in a cup over several hours. The tree is not felled, and the same trunk is tapped again on a cycle that runs for years. That is the basis of the renewable description you see on spec sheets.

What collects in the cup is a liquid, a long way from a seat cushion. Fresh sap begins to coagulate within hours of leaving the tree, so it is treated with ammonia at the plantation to hold it stable while it travels. At the foam plant it is centrifuged to raise its rubber content, then whipped with air until it froths, then mixed with a curing package of soaps, sulphur and a vulcanising accelerator.

Only then does it go into a mould. What you sit on is that froth with its cell structure set permanently by heat.

Covelle's latex is tapped from Hevea rubber trees in Sri Lanka. Natural, renewable. The species and the country are both published. They belong to a small set of facts about this layer that anyone can look up.

What does the natural in natural latex leave out?

Commercial latex foam is usually a blend. Tapped tree latex goes into the compound. So does styrene-butadiene rubber, a synthetic elastomer that costs less and behaves more predictably from batch to batch.

The industry names the two ends of that range natural and blended. Between the ends, the words on the label hold still while the contents move. A cushion described as natural latex might be almost entirely tree latex. It might be a good deal less. The description reads the same at either end.

Which makes the buyer's real question arithmetic. What share of this block is tapped tree latex?

A brand that prints that share has told you what is in the cushion. A brand that prints only the phrase has named an ingredient without quantifying it, and that is the answer the category gives by default.

None of which makes a blend the wrong choice. The synthetic component buys consistency, and consistency is worth something in a seat that has to feel the same on both ends of a three-seater. The point here is narrower than good or bad. A number you can read is a specification. A word you have to interpret is not.

Dunlop or Talalay: what the two cure processes actually produce

The same tapped latex becomes two quite different foams depending on how it is set.

Dunlop is the older method, and the simpler of the two. The frothed compound is poured into a mould, filled to the brim, and cured in a single pass. Solid particles drift downward while the block sets, so it finishes denser at the bottom than at the surface. Dunlop foam runs firmer and heavier for a given thickness, and it costs less to produce.

Talalay inserts two steps before the heat. The mould is filled part way and sealed, then air is drawn out, so the compound expands under vacuum until it reaches every corner of the cavity. The expanded foam is flash-frozen before curing. That fixes the cell walls while they are still open and even. Talalay foam comes out lighter, with a more consistent density from top to bottom.

Neither process is the better one. They are two answers to the question of how a seat should push back, and a sound cushion can be built from either. Both produce an open cell structure with connected air paths through the block. That structure is why latex returns to shape quickly once a load lifts off.

How a cushion built from either one behaves under you is a separate question, and the cushion feel guide covers that ground.

Why does a latex block have holes through it?

Cut a latex core open and you will find a grid of round channels running through the slab. They are not a defect, and they are not decoration.

Latex conducts heat poorly, so a solid block of it would cure unevenly, setting at the outside while the middle stayed soft. The moulds carry rows of metal pins that leave those channels behind, and the pins let heat reach the centre of the slab during the cure. The channels are usually called pincore holes.

They earn a second job once the cushion is built. Air moves through the channels as you sit down and stand up, so the block is never fighting trapped air. Pin diameter is also a lever for tuning firmness across a single slab. A core can be softer under the seat and firmer at the edges with no change to the compound at all.

Why an ILD number settles less than you were told

ILD stands for indentation load deflection. It is the force needed to press a foam sample down by a quarter of its thickness, measured on a sample of stated size.

That final clause carries the trouble. Change the thickness of the sample, or change its footprint, and the same foam reports a different figure. An ILD means something against another ILD measured the same way on the same geometry. Set beside a figure from a different test setup, it compares nothing.

Two cushions can therefore share a number and sit unalike. A lone ILD on a product page, with no sample geometry printed beside it, is a figure without a basis for comparison. Ask what the sample was before you let the number decide anything.

There is a second reason the figure travels badly. A finished cushion is a stack, so the surface you meet sums a comfort layer, a support layer below it and whatever batting wraps the assembly. A firmness figure drawn from one layer in isolation describes that layer and not the seat.

What Covelle publishes about its latex core, and what it does not

The Covelle materials page publishes a four-inch natural latex core, tapped from Hevea rubber trees in Sri Lanka. Read that page and you have the published record on this layer, in full.

Five things are missing from it. No cure process is named, so neither Dunlop nor Talalay appears. No blend share is given. No density figure is given. No ILD is quoted. No GOLS certification is shown, because Covelle does not hold a GOLS transaction certificate for its latex, so the claim is unavailable and is not made.

The GOLS absence deserves the most attention. GOLS is the standard that would put an independent body behind this layer, and Covelle cannot produce one. Naming the gap is more useful to you than writing around it, and writing around it is what a page like this usually does.

What Covelle does publish alongside the core is the assembly chemistry. Water-based adhesives and finishes run throughout the assembled sofa, with no solvents, no lacquers and no added formaldehyde. The support layer beneath the latex is a different material with its own specification, taken apart in the piece on reading a foam disclosure.

So the instruction this page gives about Covelle is the instruction it gives about anyone. Ask for the blend share and the cure process in writing before you treat a natural latex line as a specification.

Four questions that settle a natural latex claim before you commit

What share of the latex is tapped tree latex, stated as a percentage?

Which process cured the block, Dunlop or Talalay?

What is the density, and if a firmness figure is quoted, on what sample geometry was it measured?

Is any of that printed on the page, or does it surface only when you ask?

Every material documented. Every active certification verifiable through its issuing body. Where a fact is not published, the useful move is to say which fact is missing. The aim of all of it is a sofa you can explain to anyone who asks what is actually in it. The rest of the stack is set out on the transparency page.