White specks and hard lumps: undissolved wax, or wax that seeded on a cold jug

Hard white specks are wax held below its melt point, or wax reseeded on a cold jug. The 45 C test, a straining rescue and the melt hold that prevents it.

A hard white point that a fingernail will not smear out is not shea grain, and it is usually not a grown wax crystal either. It is wax that never went into solution, or wax that came back out of solution on something cold between the pot and the container. Those two faults look identical in the tin and need opposite corrections, so the first ten minutes of work are diagnostic rather than remedial.

• Beeswax 62-65 C
• Candelilla 68.5-72.5 C
• Carnauba 80-86 C
• Hold 10-15 minutes
• Strain 106-212 um

Two specks, two different mistakes

Undissolved wax is a survivor. It is a fragment of the flake, pastille or hand-broken block that went into the jug, and it has been solid continuously since it left the supplier's bag. It is usually larger than anything the batch grew on its own, it is scattered unevenly, and it is present in the very first container you filled.

Reseeded wax dissolved properly, then met a surface cold enough to take it back out of solution: the wall of an unheated jug, a stirrer fetched from a drawer, a stack of tins at room temperature. It forms as a thin film that breaks up into flat, glossy flecks, and it appears partway through a filling run rather than at the start.

The distribution tells you which one you are looking at, because solid wax is denser than hot oil. Beeswax sits around 0.96 g/cm3 at 20 C and carnauba near 0.99, while a liquid oil blend is about 0.91 at room temperature and closer to 0.87 at 70 C. Undissolved fragments therefore sink. If the last three tins in a filling run are visibly worse than the first three, you have settling solids and the melt was never hot enough. If the first tins are clean and the fault arrives halfway through, the jug cooled while you worked.

The third and fourth rows send you elsewhere. Grown needle crystals are the subject of gritty balm from hard wax, and segregated triglyceride is grainy shea butter. Applying a straining rescue to either of those wastes an afternoon and changes nothing.

Melting point, drop point and what your pot is actually at

Three different numbers get quoted for the same wax, and mixing them up is how people melt to a figure that was never a dissolution temperature. A melting range (USP general chapter 741) is the interval over which a capillary sample goes from first collapse to fully clear. A drop point (ASTM D127) is the temperature at which the first drop falls from a heated cup. A congealing point (ASTM D938) is taken on the way down. The methods are compared in melting point methods; the practical consequence is to melt to the top of whichever range your supplier publishes, plus a margin.

Two rows in that table conflict with the standard instruction to melt at 70 to 80 C and treat 85 C as a ceiling. That instruction is written to protect beeswax colour and tocopherol, and it is right for a beeswax formula. It is simply incompatible with carnauba wax or castor wax, both of which finish melting above it. The resolution is a split melt rather than a hotter batch: put the high melting wax into a quarter of the liquid oil, take that portion past its range until it is optically clear, then build the rest of the batch on top and let the whole lot settle back into the normal window. The general technique is described under melting methods.

The other half of the problem is that you probably do not know what your melt is at. A bain-marie caps the heat source near 100 C, but the contents of a jug standing in it commonly equilibrate 10 to 25 C below the water, depending on fill level, vessel wall and whether you are stirring. An infrared reading makes it worse, because it reports a thin surface skin that is radiating to the room, and it reads badly wrong off polished stainless. Use an immersed probe, and if two instruments disagree, work through thermometer readings disagree before you change the formula.

Hold above the highest melting component, not until the last lump vanishes

The commonest melt error is treating the disappearance of the last visible lump as the endpoint. It is not. A lump stops being visible when it drops below roughly a tenth of a millimetre, which is still ten times the size a lip can detect once it grows back. Meanwhile the jug carries a temperature gradient, so material near the surface can be several degrees cooler than the probe reading at mid depth, and a film of wax clings to the wall above the liquid line where it is not being heated at all.

Hold for ten minutes above the melting range of the highest melting component in the formula, and fifteen for a rescue. Time at temperature is doing the work. Stir through the hold rather than at the end of it, both to even out the gradient and to wash the wall film back into the bulk. If the formula contains more than one hard wax, the hold is set by the highest of them and not by an average.

The cold surface test, and why it beats looking at the jug

Before you pour, drip a few drops of the melt onto a chilled stainless spoon or a piece of dark tile straight from the freezer. It sets in about five seconds into a film thin enough to read. Tilt it to the light at a shallow angle. A clean melt gives an even, slightly waxy sheen. An incomplete melt gives glinting points scattered through the film, and you can count them.

This works because the problem is one of scale. Particles suspended in 200 g of hot oil are invisible; the same particles concentrated in a film a few tenths of a millimetre thick are obvious. Run it at the same point every time, immediately before the heat sensitive phase goes in.

Cold spot seeding: where a clean melt turns lumpy

A melt that passed the tile test can still deliver lumps, and when it does, the wax came out of solution on the way to the container. Four cold spots account for nearly all of it.

1. The jug itself, once it leaves the bath. A stainless jug standing on a worktop loses heat fastest through the wall, so a skin forms up the sides and around the pouring lip. Stirring drags it back in as flakes rather than redissolving it, because the bulk is now below the wax melting range.
2. Cold tools. A room temperature spatula, pipette or probe dropped into a 68 C melt takes wax out of solution on contact. The mark is a smear of white on the tool.
3. Unheated containers. A stack of tins or tubes at 18 C is a heat sink in intimate contact with a thin film of balm. The film sets before the bulk arrives, giving flow lines, an unfilled corner and flecks along the container wall.
4. A cold room. Below about 18 C ambient, a 100 g melt loses the top of the pour window in a couple of minutes, and everything else on this list gets worse at once.

Pour temperature is the other half of the same lever. Pouring at the bottom of the window into cold containers is exactly the combination that produces flecks and flow lines, and the windows in pour temperatures assume a container that is not fighting you. Filling tubes is the harder case, because the balm has a long narrow run past the elevator disc with cold plastic on both sides of it.

Telling a speck from grain in five minutes

The discrimination test is a warmed smear, and it works because the melting ranges do not overlap. Spread a sample from the middle of a tin thinly on a glass slide or the back of a spoon, float it on a water bath, and check with a probe rather than guessing.

At 45 C, butter grain clears, because beta form StOSt has a clear point of about 41 to 43 C. Every balm wax survives, because the lowest of them starts at 62 C. That step is the standard test and the full protocol sits on the wax grit page. What is worth adding here is a second step. Take the same smear to 70 C and look again. Specks that vanish between 45 and 70 C were beeswax or candelilla wax, which means an under-melt of the ordinary kind. Specks that survive 70 C are carnauba, rice bran wax, sunflower wax or castor wax, which means the batch never approached the temperature that material needed. And specks that survive 90 C are not wax at all.

When the speck is not wax

That last case matters, because straining and remelting will not touch it. Four candidates account for most of it.

Unfiltered beeswax carries hive material: propolis, pollen, wax moth debris and fragments of cocoon. Cosmetic grade wax is filtered during processing, but wax bought direct from a beekeeper often is not, and the particles are dark, opaque and will not melt at any temperature you can safely reach in oil. The fix is upstream: melt the wax on its own, hold it liquid so the debris settles, then pour off the clean fraction before it meets a formula.

Powdered actives are the second group. Zinc oxide agglomerates and does not dissolve in oil at all, so an unwetted clump stays a clump. Allantoin is barely soluble in oil and stays particulate by design. Clays, starches and silica clump if they are tipped into a hot melt rather than dispersed into a cool oil portion first. None of these is a melting problem, and heating harder only degrades the rest of the batch.

Third, crystalline aromatics. Menthol crystals added below their 41 to 44 C melting range, or camphor added to a batch that has already begun to thicken, sit as discrete grains that feel exactly like undissolved wax and read cold on the lip. Dissolve them in a warm oil portion before they meet the batch, keeping to the limits in menthol, camphor and actives.

Fourth, and least often admitted, contamination: a fragment of dried balm left on a jug from a previous batch, or a crumb of label adhesive. If a speck is the wrong colour for the formula, it came from outside the formula.

Rescuing the batch

Undissolved wax is fully recoverable. Nothing has degraded, so a rescue puts you back where you started rather than at a compromise.

1. Consolidate the whole batch into one vessel. Scrape the tins out, including the wall films. Reprocessing container by container guarantees an inconsistent result and doubles the work.
2. Remelt to 80 to 85 C and hold fifteen minutes, stirring. If the formula contains carnauba, rice bran wax or castor wax, that is not enough: take it above 86 C, or split the melt so only the hard wax and a quarter of the oil go that high.
3. Run the cold tile test before you go any further. Glinting points mean the hold was too cool or too short. Add temperature first, then time, then consider straining.
4. Strain hot, through preheated mesh. A mesh at room temperature blocks in seconds because the balm sets on the wire. Rinse the mesh with boiling water and shake it dry, or hold it over the vessel in a warm oven, and strain in one continuous pour.
5. Add the heat sensitive phase after straining. Flavour, essential oil and tocopherol go in below 50 to 60 C, once the batch is back in the pour window.
6. Pour into preheated containers and cool them uncovered. A rescue that ends with cold tins reproduces the fault it just corrected.

Two cautions. Straining removes material selectively, so the strained batch is no longer the formula on paper: filter out half a percent of a 2 percent carnauba load and the heat resistance goes down the sink with it. And whatever you strain through has to be fit for the product, which rules out workshop paint strainers for anything going on lips. Food or cosmetic grade stainless mesh is the honest choice, and it lives with the rest of the useful kit listed under equipment.

Preventing the next batch

Buying decisions help more than technique does. Pastilles and flakes dissolve far faster than a hand-broken block, because dissolution is a surface area problem before it is a temperature one, and a block also makes it impossible to see what is left. Where a formula needs heat resistance, a small carnauba load inside a well behaved base wax dissolves more reliably than a large one, and the trade is worked through in waxes compared. Solvency matters too: a polar oil in the blend, castor being the obvious one, holds the more polar waxes in solution better and shortens the time needed at temperature, for reasons set out in solubility in anhydrous formulas.

The decision rule when it comes back

Strain once. If specks return in the batch after that, stop treating it as a melting fault. A properly held melt that reseeds is telling you the formula is carrying more hard wax than the oil phase can hold at the temperature the product will actually live at, and the answer is compositional: cut the hard wax, move the weight to a lower melting wax, or add a more polar oil. The wax substitution calculator will rebalance the blend without changing the total wax load.

Two honest limits sit under everything above. There is no published solubility data for cosmetic waxes in cosmetic oil blends, so the bulk melt targets in the table are melting data plus a working margin rather than measured saturation points, and they will be conservative for a castor-rich blend and optimistic for a very non-polar one. And detection is subjective: the threshold at which a lip registers a particle sits somewhere around 20 to 30 micrometres, but it varies between people and it varies with how the product is applied. Test on the people who will use it, from a container they filled from, three days after pouring rather than on the day.

Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.