Journal

From the bench.

Harvest reports, arguments about grading, and the occasional explanation of something the trade would rather leave vague. Written by the person who graded the pearls.

Wet oyster culture panels stacked on the timber deck of a pearl farm work barge

Harvest report

14 March 2026

4 min read

What came up at Tallon Reef, and what we left behind

Every year we take a share of one lease's harvest, sight unseen until the panels come up. Here is what that actually looks like, including the part where we turned down half of it.

The Kimberley harvest is lifted once a year, in the cooler months, and the whole thing happens over about nine days. You fly into Broome, you get on a boat, and then you sit on the deck for a week while panels come up out of eleven metres of tide.

This year's lot ran to just under four hundred pearls across the three lines we buy from. Of those, we took a hundred and sixty-one. That sounds brutal and it is fairly normal: a working lease produces a wide spread, and a house our size can only sell the top of it.

The ones we left were mostly surface. Not badly spotted — just spotted enough that we would have had to write "moderately spotted" on a report, and at these prices people do not want to read that word. A larger house would have taken them and sold them into a lower tier. We do not have a lower tier, which is a limitation as much as a virtue.

What we did take was unusually consistent in colour. The whole strand you can see on this site as the Kimberley Strand came off two panels lifted within an hour of each other, which is why the silver-rose holds all the way round instead of drifting the way an assembled strand does.

One more thing worth saying, because nobody says it: we graded on the boat, in daylight, with the pearls wet. Wet pearls flatter. Everything got re-graded three weeks later on the bench in New York, dry, under a north window, and four of them moved down a grade. Those four are not on this website.

A larger house would have taken them and sold them into a lower tier. We do not have a lower tier.

How it works

2 February 2026

5 min read

Macro photograph of the iridescent nacre lining of a large oyster shell, showing growth ridges

The bit you are buying is about a millimetre thick

Nearly every fine pearl sold today is a bead with a coat on it. The coat is the entire product, and almost nobody will tell you how thick theirs is.

Here is the thing most pearl buying gets wrong. A cultured pearl usually is not solid pearl. It is a small sphere cut from freshwater mussel shell, implanted into an oyster, which then coats it in nacre for two or three years. When you cut one in half you see a bead with a rind.

That rind is what you are paying for. It is what produces lustre, it is what produces orient, and it is what wears through. A thin-coated pearl looks perfectly acceptable in a display case and goes dull and then bald in a decade of real wear, because the coat is thinner than a sheet of paper and skin is mildly acidic.

You cannot see coat thickness from the outside. You need an X-ray, or you need to trust the person selling it. That asymmetry is precisely why so few sellers mention the number, and it is why we put it on every page and every report.

The floors we hold to: 0.4mm per pearl for akoya, which is the threshold the Japanese laboratories use for their highest certification. 1.2mm for Tahitian — French Polynesia scrapped its 0.8mm export minimum in 2017 and we kept the old rule. 2.0mm for South Sea, which is easy, because a South Sea oyster is enormous and slow and lays down two to six millimetres without being asked.

Keshi are the exception and the reason they are worth knowing about. A keshi forms when the oyster rejects the bead but coats the pocket anyway. There is nothing inside it. It is solid nacre, all the way through, which is why the lustre on a good keshi is harder and more metallic than on almost any round pearl — and why the collar on this site costs fifteen hundred dollars rather than nine thousand, despite arguably being the more interesting object.

You cannot see coat thickness from the outside. That asymmetry is why so few sellers mention the number.

An argument

9 December 2025

3 min read

Three pearls on a grey card — a white South Sea, a golden South Sea and a dark Tahitian

Nobody issues AAA. Everybody sells it.

There is no governing body behind pearl letter grades. There is no agreed scale. Every seller assigns their own, and the whole trade knows it.

Diamonds have the GIA, and a GIA report means the same thing in Antwerp as it does in Los Angeles. Pearls have nothing equivalent in general retail use. The GIA does publish seven value factors, and they are excellent, but they are a framework for describing a pearl — not a letter you can print on a tag.

So the trade invented letters. AAA appeared, then AAAA, then AAAA+, then AAAAA, each one arriving because the one below it had been devalued by everyone using it. It is grade inflation with no examiner. A pearl one shop honestly calls AAA is the shop next door's AA, and neither of them is lying, because the word does not mean anything.

The alternative is not complicated. Describe what you measured. "Excellent lustre" can be defined: a reflected light source has a hard edge in the surface. "94% clean" can be counted. "3.1 to 4.0mm of nacre" can be X-rayed and checked by someone who is not us.

That is the whole of our position, and it costs us money. Descriptive grading means writing "lightly spotted" on a report for a pearl a competitor would call flawless, and then watching somebody buy the flawless one. We think that trade is worth making, but we would rather say out loud that it is a trade than pretend it is pure principle.

It is grade inflation with no examiner.

The short version

All of it is on the pieces themselves.

Every argument in this journal shows up as a line on a grading report. That is the point of writing them down.

See the collection How we grade