
A stone fixture is usually approved from a photograph and then judged on site, and those two moments ask completely different questions. The photograph answers whether the object is beautiful. The delivery answers whether the object is what was ordered, whether its parts meet properly, whether it can be serviced, and whether the person receiving it has any way to tell a natural variation from a defect. Without a checklist, that last question is settled by argument, and the argument is usually won by whoever is more tired.
This article is a receiving and selection checklist for stone lighting, written for the buyer rather than for the showroom. It assumes the fixture has already been chosen and asks what can actually be inspected, in what order, and what evidence should exist before the crate is opened. It is deliberately about checks rather than taste: a checklist that depends on taste cannot be delegated, filed, or used to settle a claim.
Three failures account for most of the trouble with stone fixtures. The first is approving a single styled image and treating it as a specification. The second is rejecting a good batch because natural variation was read as damage. The third is accepting a real structural flaw because it was described as natural. A workable checklist has to do two jobs at once: give the receiver permission to accept variation, and give them a reason to stop a shipment.
- 01Why Stone Needs Its Own Receipt Checklist
- 02First Check: Is It Stone, and Which One
- 03Reading the Surface Under a Raking Light
- 04Where the Parts Meet: Fit, Seams and Movement
- 05Stability, Mass and the Question of Support
- 06The Electrical Layer Inside a Stone Body
- 07Normal Variation Versus a Claim
- 08Proportion, Finish Consistency and the Set
- 09The Documents to Request Before Dispatch
- 10What a Good Receipt Looks Like on a Hotel Floor
Why Stone Needs Its Own Receipt Checklist
Most fixture checklists in a project manual were written for metal and glass, and they inherit the assumptions of those materials: that two units of the same model are identical, that a scratch is a scratch, and that the main thing which can go wrong between factory and site is transport damage. Stone breaks all three assumptions. Two pieces cut from the same block are not identical. A mark that would be a defect on a painted steel plate is a characteristic of the material on a stone face. And the failures that matter most are not the ones visible in a photograph: an uneven section, a joint that was forced during assembly, a lamp that runs hotter than the stone tolerates, a driver sealed behind a finished surface.
There is also a commercial asymmetry that a checklist exists to correct. The supplier inspects the fixture at the point where it is easiest to inspect, in a workshop with the right light and the parts still loose. The buyer inspects it at the point where inspection is hardest: in a corridor, under a temporary work light, with a schedule running late. A receiving checklist moves some of that inspection back to the factory, by turning specific observations into documents that must be produced before dispatch, and it makes the site inspection short enough to actually happen.
The order of checks matters more than their number. Identity comes first, because everything else depends on knowing what the material is. Surface comes second, because surface is where most disputes begin and where most of them can be closed on the spot. Fit and stability come third, because they are the failures that only appear once the object is assembled. The electrical layer comes fourth, because it is invisible until something stops working. Documents come last in the sequence but first in leverage: a document requested before production is a specification, and the same document requested after delivery is a complaint.
Two practical notes before the checks begin. Handle stone with clean hands or clean cotton gloves, because oils transfer to a porous face and show up later as a darker patch that cannot be attributed to either party. And photograph everything against its label at the moment of unpacking, because a claim made after installation has to explain how the damage occurred, while a claim made at unpacking does not.
First Check: Is It Stone, and Which One
The first question is not whether the fixture is attractive but whether the material is what the order says. Stone is imitated well enough that appearance alone is not evidence, and the imitation is not necessarily dishonest: cast resin and composite materials can reproduce veining convincingly, and a buyer who has only seen a screen has no way to tell the difference. Four observations, used together, close most of the gap.
The first is pattern behaviour. Natural stone varies across a face and does not repeat between pieces; a set of units with an identical pattern in an identical position is a strong sign of a moulded material. The second is touch. Stone reads cooler than resin at room temperature, which makes a hand test a useful first screen, though it is only a screen: a thick resin casting in a cold room will also feel cool. The third is documentation. The written description should state the material in plain terms, and a page that describes the object only through mood words is not a material statement. The fourth is photographic evidence at magnification: a genuine stone face shows inclusions, small mineral deposits and a structure that continues below the surface, while a printed or moulded imitation tends to show a pattern that sits on top of the material.
Then comes the harder question, which is which stone. The names in this category overlap badly. Alabaster is used for two different mineral families; one of them is also sold under names borrowed from onyx and travertine; marble and travertine appear in the same sentences as though they were finishes rather than materials. A buyer does not need to resolve the geology, but does need the name on the order to be unambiguous, because the two alabaster families differ in hardness, in how they respond to a damp room, and in how they are cleaned. The written statement should give the mineral family and the surface treatment, not just the trade name, and it should be the same statement that appears on the packing list.
Where identity is in doubt, the tests belong to the supplier and to an offcut, not to the delivered face. A scratch test on a finished surface, or an acid drop on a polished panel, converts a question into damage. What a buyer can reasonably ask for is the supplier’s own identification, a sample of the raw material, and a statement of the treatment applied. Our own guide to what alabaster is and how to tell real stone covers the mineral side in more detail; the point here is procedural. If the material statement cannot be produced, the checklist stops at the first line, because every later judgement depends on it.
Reading the Surface Under a Raking Light

Surface inspection is usually done with the light behind the inspector, which is the least informative arrangement possible. A face lit straight on shows colour and hides topography. Move the light to the side, so that it rakes across the surface at a shallow angle, and the same face becomes a map: undulations, tool marks, filled pits, the transition between a polished face and a sawn edge, and any repair that was made before dispatch. This is the single most useful technique in the checklist, and it costs nothing but the willingness to hold a torch at an angle.
What the raking light reveals has to be sorted into three groups. The first is structure: veining, colour drift, mineral inclusions, the cloudy areas that make the stone legible when it is lit from within. None of this is damage, and all of it should be expected to vary between units. The second is workmanship: tool marks left on a face that should be finished, a filled pit that was not levelled, a polish that stops short of an edge, a seam that was opened and re-bonded. This group is legitimate ground for a conversation with the supplier, and it is best raised with photographs taken under the same raking light. The third group is damage: a crack that runs across an edge, a chip that removes material, a star at a fixing point where a screw was over-tightened.
Telling a vein from a crack is the skill that decides most claims, and it can be taught in a sentence. A vein follows the structure of the stone, keeps a consistent width, and does not step: run a fingernail across it and the nail passes over without catching. A crack has a step, catches the nail, and usually runs to an edge or to a fixing point. A crack that reaches an edge is a structural question rather than a cosmetic one, and it should be held rather than installed, whatever the schedule says.
Finish deserves its own line in the inspection record. Polished, honed and matte faces are three different products made from the same block, and the difference is not a matter of degree: a polished face returns light in a defined direction and shows every fingerprint, while a honed or matte face scatters light and hides marks but reads flatter. A set that mixes finishes by accident will look like a set that was assembled from two orders. The check is simple: put the units side by side under the same light and look along the row, not at each one individually. Any unit that reads brighter, darker or glossier than its neighbours should be set apart and photographed before installation.
| What the raking light shows | What it usually is | What to do |
|---|---|---|
| Cloudy fields, colour drift, dark mineral spots | Structure of the material, present in every genuine piece | Record as expected variation; do not treat as a fault |
| Faint line that the nail passes over | A vein following the stone | Accept, and note its position on the receiving sheet |
| Line with a step that catches the nail | A crack, often from handling or an over-tightened fixing | Hold the unit, photograph with the light at the same angle, raise before installation |
| Dull patch inside an otherwise polished face | Incomplete polish, or a repair that was levelled and not re-polished | Ask for the finish reference and compare; this is a workmanship question |
| Chipped edge or missing corner | Impact damage, sometimes old and filled | Reject the unit; a filled edge on a lit face will read as a dark mark once switched on |
Where the Parts Meet: Fit, Seams and Movement
The joint between stone and metal is where most of the engineering in a stone fixture actually happens, and it is the part of the object that a photograph does not show. Three interfaces deserve attention: the collar or gallery that carries the shade, the backplate or canopy that carries the fixture to the wall or ceiling, and any point where a threaded part passes through stone. At each of them, the question is the same: does the joint close evenly, and can it be opened again without damaging the stone.
An even gap is the visible signature of a joint that was machined rather than forced. Look along the seam at eye level and then from below; a gap that narrows at one end means the parts are not concentric, and the stone is probably being loaded on one edge. Look for adhesive squeeze-out at the same time: a bead of adhesive on a visible face means the joint was assembled dry and then filled, which is a repair rather than a fit. Look for the marks of a tool that slipped, since these appear at exactly the moment when a collar is being tightened onto a brittle material.
Serviceability belongs in the same check, and it is routinely forgotten. Ask whether the shade can be removed for cleaning or lamp replacement without taking the fixture off the wall, and whether the lamp is reached from below, from the top, or only by dismantling the assembly. A stone shade that has to be lifted off a collar with two hands while standing on a step is a maintenance problem waiting to be discovered by housekeeping at the least convenient moment. The same question applies to the driver and to any control gear: if it sits behind a finished surface with no access panel, the first failure becomes a building repair.
Stone also moves, and the joint has to allow for it. The material expands and contracts with temperature and moisture far less than timber, but it is not rigid in the way metal is, and it does not forgive a thread that is tightened until the stone stops it. Well-made assemblies use a compliant interface: a gasket or a soft washer between stone and metal, set screws with soft tips rather than a direct metal point on the stone, and a collar that locates the shade rather than clamping it. When a unit arrives with a star crack radiating from a fixing point, the cause lies almost entirely in this interface, and the fix is a design change rather than a replacement part.
Two questions to the supplier close this section, and both are answerable in writing. Was the fixture trial-assembled before finishing, and again after finishing, with a record of what was checked? And which parts are intended to be removed in service, with what tool? A supplier who can answer both has a process; a supplier who answers with reassurance has a catalogue.
Stability, Mass and the Question of Support
Stone is dense, and density is the reason a stone fixture cannot be specified from a photograph of its silhouette. The relevant input for fixing design is the tested weight of the finished item, not the shipping weight on the carton and not an estimate based on the size of the box. A table lamp and a wall sconce fail in different ways: the lamp tips, the sconce pulls out of the wall. Both failures trace back to a fixing decision made without the real number.
The mistake that recurs in projects is transferring experience between materials. An alabaster body of a given envelope is lighter than a marble body of the same envelope, and a fixture with a thin stone shade is lighter again than the same shape in solid stone. Treating them as one category produces two errors at once: over-built supports where they are not needed, and under-built supports where the stone is heavier than the installer assumed. The sound instruction is to fix each item according to its own tested weight, and to let the person who signs for the structure make that judgement.
On site, the stability check is physical and takes seconds. Stand a portable fixture on a level surface and press alternate corners; any movement means the base is not flat, and a stone base that rocks will eventually chip at the contact point. For a wall fixture, check the backplate for flatness against a straight edge before it goes up, because a twisted plate will pull the stone when the fixings are tightened. Then look at the proposed fixing position: the question is not whether the wall is strong, but what the wall is made of, and whether the fixing reaches something structural rather than a hollow void behind plaster.
Everything about load paths, anchors, box ratings and structural adequacy belongs to a qualified person working from the actual weight and the actual substrate, in accordance with local regulation and a structural calculation. That is not a disclaimer bolted onto the end of the article; it is the correct division of responsibility, and a checklist that pretends otherwise is dangerous. What the buyer can do is make the inputs available: the tested weight per item, the fixing method the manufacturer intends, and the position on the drawing. Our own note on the checks that should happen before a heavy fixture arrives sets out the same sequence from the contractor’s side.
The Electrical Layer Inside a Stone Body
A stone fixture is a lamp inside a mineral envelope, and the envelope changes the electrical requirements rather than sitting outside them. The first consequence is thermal. Stone holds heat and does not dissipate it the way an open metal frame does, so a source that runs hot will raise the temperature of the material around it, dry it, and over time change its colour near the lamp. The check is not a temperature measurement by the buyer but a specification question: which lamp type and rating is the fixture designed for, and is that stated in the documents rather than implied by the showroom sample.
The second consequence is optical, and it interacts with heat. A source with good colour rendering and a warm tone flatters the mineral, which is why the intended lamp is part of the design rather than an accessory. A colder source drains the warmth out of the stone and makes it read grey. A source that is too compact and too close to a thin wall of stone produces a bright core and a darker rim, which is a property of the geometry rather than a fault in the material. Our article on how thickness changes the glow of a stone shade deals with the optical side; here the point is that the lamp stated on the order and the lamp installed on site should be the same lamp.
The third consequence is maintenance. Three items fail in service: the lamp, the driver, and the connection. Each needs a route. Ask where the driver sits, how it is reached, and whether it is replaceable as a unit rather than spliced. Ask whether the terminal block is accessible with the fixture in place, and whether the cable entry is at a position that allows the fixture to hang or sit naturally rather than fighting its own supply. A stone body with no service access is a sealed object, and sealed objects are replaced rather than repaired.
Finally, dimming. A stone fixture is often specified on a dimmed circuit, and dimming compatibility is a system property rather than a fixture property: the lamp, the driver and the control gear have to agree. Ask for the compatibility statement in writing, and test one unit on the actual circuit before the batch is installed. The failure mode here is familiar and expensive: a fixture that works perfectly on a bench and flickers at the low end on site, in a room where the stone is the sole light source.
Normal Variation Versus a Claim

The most expensive argument in stone lighting is not about quality. It is about category: whether a given phenomenon is a characteristic of the material or a defect in the product. That argument is expensive because it is usually conducted after installation, by people who did not see the goods at unpacking, about an object that is now fixed to a wall. The way to avoid it is to agree the categories before production and attach them to the purchase order.
The categories are not subtle once they are written down. Natural structure, colour drift, mineral spotting and variation between units are characteristics: they are the reason the material was chosen, and a set with none of them is probably not stone. Workmanship faults are the supplier’s responsibility and are correctable: an incomplete polish, a filled pit that was not levelled, a joint that does not close. Damage is a claim: cracks with a step, chips that remove material, stars at fixings. The checklist works when the receiver can place an observation in one of the three columns within a minute, and when the columns were agreed before anyone had an incentive to disagree.
Two refinements make the table usable in practice. The first is that the category can depend on the face. A mineral inclusion on the outside of a shade is a characteristic; the same inclusion sitting directly in front of the lamp, where it reads as a dark shadow through the lit stone, may be a reason to re-select the piece, because the choice of which part of the block becomes which face is a manufacturing decision. The second is that the category can depend on the position: a filled area on the back of a base is invisible and irrelevant, while the same fill on a lit face is not.
| Phenomenon | Category | How to handle it |
|---|---|---|
| Veining, cloudy fields, colour drift between units | Characteristic of the material | Accept; record on the receiving sheet; explain in the handover note |
| Small mineral inclusion on a visible face | Characteristic, unless it sits in front of the lamp | Accept on outer faces; re-select if it reads as a shadow through the lit stone |
| Patch where the polish stops short | Workmanship | Raise with the supplier with a raking-light photograph; re-polish or replace |
| Adhesive visible at a stone-to-metal joint | Workmanship | Treat as a fit question, not a cosmetic one; ask how the joint was assembled |
| Crack with a step, or a star at a fixing | Damage or a design fault in the interface | Hold the unit; do not install; raise as a claim with photographs taken before unpacking is complete |
| Chip or missing corner | Damage | Reject the unit; a filled edge on a lit face will read as a dark mark |
Proportion, Finish Consistency and the Set
Single fixtures are judged as objects; sets are judged as rhythm. A hotel corridor with twenty sconces is not twenty decisions, it is one decision repeated, and repetition exposes inconsistencies that a single sample hides. The first of these is proportion: within a family, the ratio of stone to metal, the depth of the shade and the thickness of the edge should scale together. A larger version that simply enlarges the stone while keeping the same collar looks heavier in a way that is hard to name and easy to see.
The second is finish consistency, which has two layers. The polish itself should be consistent across the set, and the edge treatment should be the same on every unit: a chamfer on one piece and a square edge on the next reads as a mistake from three metres away, even when nobody can say why. The third is batch: pieces cut from different blocks will not match in tone, and under warm light the difference becomes more visible, not less. For a project, the practical instruction is to order the full quantity plus spares in one production run, and to record the batch on the delivery documents so that a later addition can be matched or deliberately contrasted.
Then there is the on-and-off question, which is specific to this material and worth a line of its own. A stone fixture has two appearances, and both are part of the specification. Switched off, it is an object with a surface, a colour and an edge. Switched on, it becomes a light source whose internal structure is visible. A set that matches when switched off can still mismatch when lit, because the transmission depends on the section and on the density of the veining in that particular piece. The check is to view the set in both states, in the intended space if possible, and to accept the set in both.
Finally, the spare. Stone is repairable but not replaceable in the way a metal part is: a piece cut later from a different block will not match. Ordering a spare unit at the same time as the batch is the cheapest form of insurance in the whole package, and it is the item most often removed when a budget is trimmed. A spare held in the building also shortens the response to accidental damage, which is the failure mode that actually occurs in hotels.
The Documents to Request Before Dispatch
Documents are the part of the checklist that survives staff changes, which is why they matter more than any single inspection. What should exist before the crate is closed is a short, specific set, each item answering a question that will otherwise be answered by argument on site.
- Material statement: the mineral family, the treatment applied, and the trade name used on the order, in the same words on the packing list.
- Finish reference: the agreed sheen and edge treatment, with a retained sample that production was matched against.
- Tested weight per item: the number the fixing design will be based on, not the carton weight.
- Lamp and control specification: the lamp type and rating the fixture is designed for, the driver, and the dimming compatibility statement.
- Service instructions: how the shade, the lamp and the driver are reached, and with what tool.
- Care card per material: what to use, what not to use, and who to call when the surface changes.
- Packing method: how the piece is supported inside the crate, so that a damage claim can distinguish packing failure from handling failure.
Each document has a reading discipline. The material statement is read against the order, not against the invoice description. The finish reference is read under the same light as the delivered piece, because a sample viewed in a showroom and a fixture viewed in a corridor are not the same comparison. The tested weight is read as an input to someone else’s calculation, and copied into the fixing schedule rather than kept in an email. The service instructions are read before installation rather than at the first failure, because the decision they affect, which is access, is made by other trades.
Two of our existing notes support this step directly. The seven fields that belong in a fixture schedule describes what a schedule needs in order to be usable on site, and what to do when a specification contradicts itself covers the common case where the drawing, the data sheet and the sample disagree. Both are worth reading before the order is placed, because a document requested at that point can still change the product.
What a Good Receipt Looks Like on a Hotel Floor
The theory becomes concrete in a corridor with forty cartons and a schedule that has already slipped. A workable routine has five steps and takes about the same time as an unstructured inspection, which is the main reason it survives contact with a site.
First, unpack in a clean, lit area rather than at the point of installation, and keep the packing materials until the unit is accepted. Second, photograph each unit against its label before removing the protection, so that the record links the object to the batch. Third, run the raking-light pass on every visible face and note anything that falls into the workmanship or damage columns; this is the step that gets skipped under pressure and the step that prevents the argument later. Fourth, test one unit fully: assemble it, switch it on, dim it if the circuit allows, and time how long a lamp change takes. Fifth, sort by batch and quarantine anything held, with the label still attached.
The signatures matter as much as the steps. Three people have an interest in the outcome: the receiver, who owns the record; the electrical contractor, who owns the fixing and the circuit; and the designer’s representative, who owns the appearance. A receiving sheet that all three have seen is a different document from one that circulates after a problem appears, because it establishes what was visible at the moment of delivery and what was decided about it.
None of this requires expertise in mineralogy. It requires a sequence, a torch held at an angle, and the discipline to write down what was seen before the crates go up in the service lift. The reward is not a perfect batch; natural material does not produce perfect batches. The reward is that the conversation about the batch, when it happens, is about the right category, and that the answer is already on file.
Should I reject a stone fixture because it differs from the sample? Not on colour or pattern alone. Compare it against the agreed finish reference and the material statement, and judge variation as a characteristic unless it changes how the piece reads when lit. What justifies rejection is a workmanship fault or damage, and both are easier to establish at unpacking than after installation.
How can I tell a crack from a vein without damaging the piece? Run a fingernail across the line. A vein follows the stone and the nail passes over it; a crack has a step and catches. A line that reaches an edge or a fixing point should be treated as structural until someone qualified says otherwise.
Is a heavier stone fixture the better one? No. Weight is an input to fixing design, not a quality signal. What matters is that the tested weight is known, that the fixing is designed for it, and that the joint between stone and metal does not load the stone on one edge.
Why does the same model look different in two rooms? Stone is viewed in two states and under two kinds of light. Switched off, the surface and the ambient light dominate; switched on, the section and the density of the veining decide what the face looks like. A set should be approved in both states.
Do I need a spare unit? For a project, yes. A piece cut later from a different block will not match the installed set, and a spare in the building is the fastest response to accidental damage.
We build stone fixtures with the material statement, the finish reference and the tested weight issued as one package, so that the receiving sheet has something to check against. Send the fixture schedule and the positions, and we will return the documents and a sample set for approval before production rather than after delivery.
See the alabaster collection · How an alabaster lamp is made · Talk to us about a project













