On the Yacht "Les Trois Soeurs" and the Fragonard Perfume Launch. This is actually two sections Scott and I cut to streamline the narrative. It contains an account of a very well-known perfume (we've hidden the name, but undoubtedly some of you reading this are wearing it right now) that was declared poisonous and the emergency resuscitation they had to do on it, plus a great battle with the chief of one of the Big Boys, Givaudan, over a secret smell strength chart. The first part of the battle takes place on a yacht off the Côte d'Azur, the second at the HQ of Fragonard in Paris for a perfume launch party. The book gives the contexts.

[Lunch on the yacht "Les Trois Soeurs," owned by Agnès Costa, head of the perfume house Fragonard. Turin is chatting with Geoff Webster, head of the "Big Boy" industrial perfume molecule manufacturer Givaudan-Roure.]

As the yacht's waiters cleared plates and brought dessert and coffee, Turin found Webster giving him inside industrial information that a) he didn't know and b) made him, to his astonishment, reconsider fundamentally the way he thought about smell. And he saw, in fact, that Webster's information, when coupled to his smell prediction algorithm, could make a lot of money in a way he'd never considered.

We're not used to thinking about it this way, but from Webster's (or any industrialist's) viewpoint, a smell has three aspects to it. Most of us only think about the first one, what the smell chemists would call "character" and what we would call "What does this stuff smell like?" But if you're a smell chemist, and you're weaving together a new molecule, you also have to worry about something called "substantivity," which is how long the smell hangs around on your skin (if you're perfuming a fabric softener you want to know a) does the molecule stick to cloth and b) for how long). The third is called "intensity," i.e., strength. In other words: 1) What does the molecule smell of? 2) How long does it last? 3) How strong is it?

Until that moment, Turin realized, he had only ever really thought about character. Sell at Quest had brought up "strength" in the test, and Turin had thought "Hm" but sort of ignored it. In fact all this time he had assumed, he confessed to Webster, that the Big Five more or less only cared about smell "character" (peach blossom? cut grass? incense?). Webster said Uh uh, nope, in fact totally the other way. Deal was, what Givaudan really wanted to predict was, if they manufactured some molecule, how strong would it smell. Why? asked Turin. Webster smiled: Money, he said. Look. Say you have two molecules, both with the same character, Lily of the Valley. Odorant molecules can vary in strength by a factor of 10 to the 7th (100,000,000). Which means that though two molecules might both smell of green papaya, for the customer actually to be able to pick up the papaya smell Givaudan's perfumers might have to use 1 gram of the first but 10 tons of the other. We want the one that's 1 gram, said Webster, because that lowers our costs hugely: Our factory is smaller, we spend less in transport, we pollute less, and toxicologically it's virtually certain that you're going to have fewer problems (since we'll only be using a microgram per ounce, and regulators test toxicity by how much of the molecule the customer actually puts on her skin). If he had to choose between a fabulous new cedar character and just taking an old one and making it 100 times more potent, Webster would go with the old one every time.

Turin sat looking at the Mediterranean's blue and thinking about it. Not only (it appeared) did the smell chemists have no clue about smell character, they had even less of a grip on smell strength. What made this even more interesting was that Turin happened to know about a secret piece of information Givaudan possessed, which now made much more sense. Several years before, Givaudan's chemists had launched a major assault on the enigma of strength. They had built machines with tubes, seated many people in front of them, and sent odors flying up the tubes, recording the reactions, all to try to measure the lowest concentration at which people could smell, say, methoxypyrozine. (They would give subjects a little of the stuff so they knew what it smelled like, then seat them in front of the machine, and play fake-out: A light would come on when the odor was coming through except for the times when it wasn't coming through; the subject had to hit a button when he thought he smelled whatever it was while Givaudan's chemists hovered in the background). They had painstakingly assembled all the data on a big Smell Strength Chart, carefully labeled molecules bunched together in groups according to strength. And then tried to figure out the molecular connection: Where did their strength come from?

Turin knew about the smell strength chart because a young Frenchwoman named Calice Becker, the Roure perfumer in charge of creating it, was a friend of his, and she had told him it existed. She had been in charge of its creation and had waged a war to get the perfumers to use it, which they would not. And in fact during meetings for his ill-fated Vibration testing for Givaudan-Roure, he had actually seen the thing several times, hurriedly, under difficult conditions, specifically on the way to the toilet at Givaudan's headquarters where they had blown up the chart to poster size and hung it on the wall in the corridor leading to the Men's Room. It fascinated him, he found it "phenomenal." Turin used to excuse himself from meetings and loiter in front of it, trying to appear nonchalant while he attempted to memorize the thing. It was filled with information, what molecules smelled strong, and what didn't, and their proximity to each other presumably was based on some molecular property, but no one had the faintest clue what that property might be.

If Givaudan would publish the thing, thought Turin (the scientist), the scientists could jump on it and maybe crack the multi-million dollar secret of odor strength. But no. The usual perfume industry attitude prevailed: The chart was Secret, it was Proprietary. Which was idiotic, since they were not about to crack it any time soon. Turin being Turin, he pointed this out quite directly to Webster at the table. And he added that when anyone (say, Luca Turin, for one notable example) asked the Givaudan people "Can I see it?" they would respond immediately "It's secret," which was ridiculous given that it was functionally useless to them because they didn't understand the first damn thing about it.

Data without understanding, said Turin to Webster, is garbage. Neatly organized grafitti. And so Turin said, "Geoff, you should publish the chart."

Webster responded: "It's proprietary."

Turin said, basically, Geoff, get a clue, besides which every perfumer who ever jumped ship from Givaudan-Roure slipped off to some conveniently remote photocopier first. They've all got the chart, but they can't admit they've got it because you can sue them. But they can't use it because they don't know how the hell the map works- which brings up the question of why they stole it in the first place, but that's another issue. And- yes, yes, said Turin, he realized he was a scientist arguing the scientist's point of view, but in keeping it secret from science, Webster was keeping it secret from himself as well. Publish it, said Turin, let all the scientists out there in Internet-land download it. And someone will solve the puzzle, and you'll have the molecular secret of smell strength.

Webster thought about it and said: We'll do it.

The chart, a pirated copy of which you can get rather easily, cannot legally be reproduced here, but it is basically two axes, the vertical being Vapor Pressure (VP, you put a bunch of the molecules in a canister, stick a tube in the top, and measure the number of molecules that escape per second), the horizontal being the smell "threshold," the more subjective question "What's the lowest concentration people can actually smell?" You find a molecule's strength- Givaudan calls it the "odor value" or OV-by dividing the Vapor Pressure by the Threshold:

VP (vapor pressure) = OV (odor value)
T (threshold)

And then you get a scatter of data points, molecules falling across the plain you've created.

The chart has apparently been used only once, in an emergency involving one of the most commercially successful perfumes on the market. You've heard of it. In fact, there's a chance you're wearing it now. It's sold in a reddish bottle. In a bit of bad planning, after the capital had been sunk into this perfume, after the fragrance had been painstakingly constructed, launched, marketed, and become a sales phenomenon, the toxicity regulators tested it and found that one ingredient, methyloctincarbonate, was toxic above a certain concentration. So this multi-million dollar perfume that crowds were snapping up at department stores around the world, was (at least theoretically) toxic (given the hypersensitivity of regulators, most chemists would argue that it probably was not, and Turin agrees: "That's ridiculous"). But the regulators had spoken, and both the Big Boy that manufactured it and the famous French luxury goods company whose name was on it faced the immediate death of a cash cow, not to mention a public relations catastrophe ("Our perfume is poisonous...."). What to do? The obvious answer was to replace the offending molecule with a molecule that smelled like it, but this turned out to be tricky- and a perfect illustration of the difference between the strength of an odor and its character. There was one molecule-methyleptincarbonate-that had the same "character" as methyloctincarbonate. But it sat in a completely different spot on Givaudan's odor strength chart. Never mind; no one bothered to check odor strength. They just substituted this molecule in. The result was a disaster. The new concoction smelled nothing like the original. The perfume's top notes now behaved very differently, and not in an attractive way.

So they went to the chart. And they noticed there was a molecule that, by itself, didn't really smell anything like the toxin. But it sat at almost the exact same point on the chart. Having no actual idea what that meant in scientific terms, they decided to try it. It worked perfectly. It behaved the same, it hit the wearer with the same force, and somehow it created the same smell note. And so this perfume, which is still one of the most popular fragrances on the market, is being sold with a molecule that smells nothing like the toxic molecule that it replaced, chosen from a chart that hardly anyone uses, which functions for reasons that no one understands.

As for publishing the odor strength chart, Turin never heard from Webster....

[At the Fragonard perfume launch party in Paris]

....Turin was enjoying the mesmeric music, the interplanetary lighting, and a discussion with John Stephen and Jean Guichard, the chief perfumer for Givaudan, when Guichard's boss, Givaudan chief Geoff Webster, joined the group. Turin had not seen Webster since the lunch on Agnès Costa's yacht, and the subject turned (again) to Givaudan's secret odor strength chart, Turin arguing (again) that Webster should publish it. Webster was again refusing: It's proprietary. Guichard finally spoke up to his boss and said Geoff, you might as well publish it, everyone's stolen the damn thing. He also, with tact, repeated Turin's point that by making the thing public, someone somewhere might crack the mystery of odor strength, which would help Givaudan. But Webster wasn't hearing it, and Guichard diplomatically got out of the arena.

Turin, doing tag-team, then clambered back into the ring, and Guichard and Stephen watched as Turin and Webster went at it. It was an almost total communication disconnect. Webster kept doggedly repeating "I'm not gonna help anyone else." He alternated this with a typically American MBA mantra, "Take no prisoners." Turin's argument went, basically, that, A, everyone had the chart already, B, you've never been able to crack this thing so let someone else try because, C, it's in your own interest to have the answer- it'll mean a lot of money to you, no matter who comes up with it- and D, this is going to be of great scientific importance. Webster's response, which essentially failed to acknowledge Turin's logic, went, A, he didn't give a damn if everyone had stolen copies because, B, he wasn't going to help anyone have it officially because, C, he didn't see how it was going to help out Givaudan (Webster just didn't seem to get C) and, D, he didn't care about science, he was running a goddamn company. And besides, said Webster, we absolutely can design molecules through Shape. And at that point he started talking about the way the body recognized drugs, which do work by Shape (but which interact, Webster omitted to note, with a completely different set of receptors), at which point someone dragged Turin away before he could react, which allowed Webster to say to the remaining audience that, look, he was sorry and all but Luca was just wrong and they absolutely could design fragrance molecules by Shape!

Then Webster was whisked off to be introduced to some businessmen and Stephen was called on to talk about his new Fragonard creations with some perfume writers, and it was not until it was all over and Turin and Stephen were in the Citroen driving down Boulevard des Capucines toward a restaurant near the Bastille that they tried to make sense of it all. "Geoff comes out with little phrases like 'molecule design,'" says Stephen, shaking his head, "and then he just stops. As if that's supposed to mean something."

Turin says, "You ask him 'So does your [Shape] theory actually work?' and he says 'Oh, I'm sorry, that's a corporate secret.' Translation: 'No!'" Someone had asked Webster to clarify whether he meant that drug molecules would be delivered through the nose; he hadn't made it clear. "So, what, we're going to be snorting suppositories?" asks Turin. "Or sticking aspirin up our noses? Please."

Stephen: "He was talking about how we were designing molecules in the 1850s." He laughs, incredulous. "In the 1850s, they didn't know what a molecule was."

Turin says that maybe they'd found some hexahydrophthalic anhydride in some Egyptian tomb.

"Is it fair to say," asks Stephen, "that all the traditional odorants were found by pure luck, discovered by people who happened to smell them?"

Turin: "Absolutely."

Webster had said that when it came to designing molecules, "It used to be hit-and-miss."

"And what is it now?" Turin demands of the windshield, "miss-and-miss?" He steers the Citroen toward the Bastille. He is grinning, they're both grinning. "These fragrance companies all have these multi-million dollar NMRs, this huge magnet that pulls the metal out of your teeth at 200 yards, and it's for analyzing molecules." He is enjoying the irony. "And it's so beautiful, gleaming metal and glass, and when you turn it on the lights light up so brightly and the buzzes buzz impressively and the whirrers whir and they spend all this money on it, and then they think 'Well, what are we going to do with it? Hey, let's go suck up some roses in Costa Rica and stick them in this zillion dollar machine.' And then they analyze them, and the rose smell turns out to be... linalool." Stephen and Turin are both in paroxysms of laughter. "Well, we knew that!"

Stephen, wiping his eyes: "The things they go through to find a new molecule."

"They buy a bunch of linalool," says Turin, "and from it they synthesize some new molecule, and then they say well, there's good news and bad news. The good news is this thing smells absolutely terrific and it's really powerful. The bad news is it's got eleven cyclopropane rings in it, all of them interlocked, the Rubick's Cube of molecules, and so Webster puts his chemists on it and six months later they have a milligram of the stuff, with a yield of .01%. So they've started with a ton of linalool and squeezed out a single drop. And then they give it to the process chemists, who slave away for two years and reduce the price from a million pounds per kilo to 800,000 pounds per kilo. At which point they give it to the accountants, and the accountants throw the thing in the trash and make them go back to linalool!"

Turin starts talking with Stephen about aromatherapy. He has just been professionally courted by a woman who does aromatherapy, brought to her mansion in London. "This woman is seven feet tall if she's an inch. She blasts into a room, *boum!" (He throws up both arms violently in the car.) "Wearing a mini-skirt made out of cheese and a fur like a boa constrictor. She's the most alarming looking woman you've ever seen. Next to her, we all look positively vestigial. She lives with her husband in some huge two million pound house in Islington. It's a modern house composed entirely of stark, sterile white angles. You feel like you're about to be operated on. The maid serves you a drink, and you keep expecting to smell anesthetic. The walls are covered with huge pieces of stark modern art. They have a baby. They keep it in a room somewhere in the house. The baby is going to have its head bashed in by a huge piece of modern art."

Stephen shares some interesting views on aromatherapy and then returns to biology. "If only human chemists could be as efficient as nature's chemists," he says. "The enzymes in our bodies catalyze things, at room temperature, with an incredible 99% efficiency."

Turin is mulling over commercial insanity. He sighs and says soberly, "The thing is, Geoff actually believes his own corporate PR."

He parks the Citroen on rue Biscornet, and they go into the Restaurant Saint Amaranthe for dinner.