On a sunny afternoon last June, the French perfumer Jean-Claude Ellena arrived at the offices of Hermès, the luxury-goods maker, in Pantin, just north of Paris, to present his first essais—or olfactory sketches—for the company’s next perfume. Ellena, who is fifty-seven years old, had recently been named Hermès’s first in-house perfumer by Jean-Louis Dumas Hermès, the chairman of the company. Dumas Hermès wanted to fix a delicate problem: Hermès had an elegant perfume collection that included classic scents like Calèche and 24, Faubourg, yet they sold only modestly. Chanel, one of Hermès’s chief rivals, made ten times as much money on perfume. (Led by its eighty-three-year-old warhorse, Chanel No. 5, the company’s 2003 sales totalled $1.2 billion.) It might be possible for Hermès to make one of its older scents chic through advertising, but the family had chosen a more daring strategy: it would adopt Chanel’s approach, and set up its own perfume laboratory. Ellena’s mandate was to invent an intimately related family of scents that embodied the aesthetic of Hermès—a distinctly Parisian firm, founded as a saddlery concern on the Rue Bassedu-Rempart in 1837, that is known for its craftsmanship.

The scents sold by fashion houses such as Donna Karan and Christian Dior are not made by Donna Karan and Christian Dior. They are created by independent companies, such as Givaudan, in Switzerland, and Quest International, in the Netherlands. Estée Lauder has long been celebrated for her perfumes, but she did not create them—they were created by professional perfumers. (White Linen, for example, was created by Sophia Grosjman, a senior perfumer at International Flavors & Fragrances, a company based in New York.) Lauder was a discerning and involved client, but saying that she created her own scents is like saying that Pope Julius II painted the Sistine Chapel.

Hermès knew that Chanel’s in-house approach had its disadvantages. The house’s fragrance collection was limited by the creativity of one man—Jacques Polge, the company’s perfumer. Chanel couldn’t tap a brilliant new perfumer at, say, Firmenich, a Swiss company. Then again, a fashion house that outsources perfume creation may come up with individual top sellers, but it will find it difficult to amass a collection with a coherent identity. Chanel had a perfumer with a consistent aesthetic, institutional knowledge, a sense of tradition. The Hermès family had taken note when Polge, in 2001, created Coco Mademoiselle, another multimillion-dollar hit.

Soon after entering the Hermès offices, Ellena was directed to a room with a large conference table. Hélène Dubrule, the company’s international-marketing director for perfume, greeted him. Dubrule, who is thirty-nine years old, has an almost English crispness, and wears tailored clothes. Ellena was wearing his uniform: sports coat, button-down oxford, no tie, khakis. “Ellena” means “the Greek,” and he looks the part, although his family is thoroughly French. He is not tall, but he has the confidence of a man who is conscious of being handsome.

Forty-five minutes later, Véronique Gautier, the president of the perfume division, walked in, dressed entirely in Hermès. After ordering tea and coffee from her assistants, Gautier, an elegant woman in her forties with dark hair, chatted briefly with Ellena, careful not to refer to the small glass spray vials that she knew he was carrying in his pocket. The presentation of an essai is a vulnerable moment for a perfumer. Ellena’s submissions have been greeted with kisses and exclamations of joy. At other times, executives have hurled his creations back at him with fury: “This is shit! Get out, Monsieur Ellena! We have nothing left to say to you!”

Finally, Gautier said, in French, “Good. So what do you have?”

Ellena grinned and reached into his sports-coat pocket. “Three,” he said. He placed three spray vials, labelled AG3, AD2, and AD1, on the table. He picked up several touches—paper smell strips—sprayed them with scent, and handed them to Gautier and Dubrule.

The women held the touches under their noses, and breathed deeply. After a moment, Gautier broke the silence. “One of them I like,” she declared. “One I don’t like at all.”

Most perfume houses are based in France, and, as a result, the French dominate the industry. It is an insular and secretive business that remains governed by the solemn idea of the “purity of art.” This is spoken of with equal parts pride and cynicism. “French perfumers come from the Sixteenth Arrondissement, and they all have degrees in poetry and commerce from some chic school,” one Parisian perfume executive told me. “They consider that what they create is great art, and that, because they are French, the world should come on bended knee and think itself lucky to be blessed with their creations. You talk to a French perfumer, and it’s ‘My perfumes are wonderful, they lost five million dollars, but who cares, they’re objects of art that will live forever and conform to my immortal, pure aesthetic.’”

The market for perfume has been sluggish in recent years. Since 1999, the French market has grown anemically, America’s has been flat, and Germany’s has shrunk. At the same time, the amount of money spent on perfume advertising has increased: the launch of a new scent often costs tens of millions of dollars, sometimes even more. Every fashion house wants a blockbuster like J’adore, a scent created for Christian Dior by Calice Becker, a perfumer at Quest International; in 2000, the year after its launch, it topped a hundred and twenty million dollars in sales. Yet trying to create the next J’adore is an expensive gamble, for the number of failures greatly exceeds the number of hits. A French perfumer rattled off for me the names of several recent “disasters”: Champs-Elysées, by Guerlain; C’est la Vie, by Christian Lacroix; Kingdom, by Alexander McQueen. Moreover, it isn’t clear how much profit a success like J’adore actually yields, considering the high marketing and production costs. The perfume industry’s accounting methods rival Hollywood’s in ingenuity. “You can’t always tell a flop, and no one can find out exactly how much a perfume lost, because companies consolidate their figures, although everyone whispers guesses,” the perfumer said. “The strategy is that you spend in the first year on ads what you expect to gross in that year.” Often, the ratio between advertising and sales fails to balance out. “Chanel’s Égoïste, I heard, is a flop, given the amount they spent on advertising,” he said.

These days, the creation of a perfume typically begins with a brief: a conceptual description of an imagined new scent, provided to the perfumer by the fashion house. When developing J’adore, executives at Christian Dior told Becker to create something “sexy like a stiletto and as comfortable as a pair of Tod’s.” Some people blame the brief system for the industry’s decline; because the marketers who write briefs now commission and approve scents, they have acquired substantial control over the perfume-creation process. The French executive told me, “Basically, it’s ‘We want something for women.’ O.K., which women? ‘Women! All women! It should make them feel more feminine, but strong, and competent, but not too much, and it should work well in Europe and the U.S. and especially in the Asian market, and it should be new but it should be classic, and young women should love it, but older women should love it, too.’ If it’s a French house, the brief will also say, ‘And it should be a great and uncompromised work of art,’ and if it’s an American brief it will say, ‘And it should smell like that Armani thing two years ago that did four million dollars in the first two months in Europe but also like the Givenchy that sold so well in China.’”

The briefs at Hermès are more reserved. Every year, Dumas Hermès comes up with a theme for the fashion house. In 2003, it was the Mediterranean Sea. A year before, Gautier had discussed the project with Ellena, who was then working for Symrise, a German company. He “won the brief”—industry parlance for securing a contract. Gautier and Ellena travelled to Tunisia and paid a visit to the summer home of Leïla Menchari, who designs the window displays in Hermès boutiques. He created a scent that was inspired by Menchari’s garden—it suggests warm sunlight splashed with cool water. The perfume was named Un Jardin en Méditerranée, and it became the first scent in a new Hermès collection: the Jardins. For 2005, Dumas Hermès had chosen “river” as the house’s theme. At first, Gautier had considered requesting a scent that conjured a garden in the Amazon Basin. She then shifted her imagination to the Ganges. Finally, after consulting Dumas Hermès, Gautier chose the Nile, and a title: Un Jardin sur le Nil. That was the entire brief.

In early May, Gautier, Dubrule, and Ellena flew from France to Egypt. “When I am in the process of creating a perfume, I never know how things will start,” Ellena said. “So when they say to me, ‘You’re coming with us up the Nile,’ I find that agonizing. Because, right there, they’re delimiting my space. I experience it as a loss.”

Gautier, who evinces certainty when picking up a fork, had complete faith in Ellena. She had called him and said, “The first worked well. You’ll do the same for the second.” Nothing, he replied, is guaranteed.

Before the trip, Ellena had engineered a perfume in his head. Egypt inspired inevitable associations: heavy smells, such as incense, thick jasmine, and wood smoke. Yet, he wondered, what does a real Egyptian garden smell like? Dubrule had learned of a large garden in Aswan, called Kitchener, and she decided that they would make a visit during the trip.

Upon arriving in Aswan, the Hermès group checked into the Old Cataract Hotel, whose slightly shabby rooms featured elegant wood porches. Sitting outside, Ellena sniffed the air of the Sahara and found it disconcertingly blank; he was so nervous that he couldn’t sleep that night.

In the morning, they went to Kitchener, where there were few flowers in bloom. Gautier, Dubrule, and Ellena were disappointed, but they nevertheless started smelling. They sniffed nasturtium, a salad green with an anodyne watercress scent. Ellena ate some. They tried lantana, a perennial whose flowers smell, rather limply, of banana and passion fruit. They avoided looking at one another. Ellena smelled the flowers of the acacia tree, which have a soft, frangipani-like scent. He turned to Gautier and said, “That’s not our story.” She smelled it and agreed, dismissing it in her forceful manner. As Ellena walked through Kitchener, the perfume he had built in his head disintegrated and blew away; now he had nothing.

The group walked around Aswan. The markets were full of spices, and Ellena smelled lotus roots; when macerated in water, the root produces a smell halfway between peony and hyacinth. He also found some jasmin sambac, which is full of indoles, molecules that smell overwhelmingly animalic. Feces are rich with indoles, he explained to Gautier and Dubrule, and so are decomposing bodies. It’s feminine, the smell of death. Calvin Klein’s Eternity is a heavily indolic perfume—the name must have been ironic, he joked. But indoles were not their story, either.

He wrote his observations down in a small orange notebook. Ellena later told me, “The painter learns to see, the pianist learns to listen, I learned to smell. But it’s a question of the brain, not of the nose, and you learn it simply by experience. Everyone can smell everything I can smell, but they don’t know how to understand it, distinguish elements, or how to speak about it. That’s why I’m a perfumer.” He added, “I would say it took ten years to know, twenty to master.”

One morning, Ellena and his companions went for a trip on the Nile in an aged wooden motorboat; ancient ruins on the surrounding rock cliffs loomed over them as Ellena steered the party upstream. The Nile has an opalescent black hue that, in shallow depths, becomes transparent. It has a fresh smell. They motored past wild reeds and feluccas—narrow boats with tall triangular sails—until they reached a small island. Walking ashore, they began following a street that led to a Nubian village. It was during this stroll that Ellena saw, hanging low in the trees that lined the street, plump green mangoes.

The fruit has a complex, authentically exotic smell: it is rich and fresh simultaneously, a rare combination. The scent is also ephemeral. The fruit exudes an odor only when it is on the tree. Once you pick it, the smell deteriorates; within sixty seconds, it is essentially gone. Ellena was beguiled by this elusive fragrance. Green mango, he suggested to his companions, could form the base of Nil.

Dubrule pressed her nose into the branches, finding a hint of apricot and grapefruit. At one point, Gautier frowned; she detected the smell of nail-polish remover. Indeed, green mango contains acetone, the solvent’s active ingredient.

“You will, above all, not put nail-polish remover in the perfume!” Dubrule later commanded Ellena.

“Above all!” Gautier concurred.

Ellena promised the women that he wouldn’t, knowing full well that he would. Acetone is often used in perfumery, he told me; it provides a lightning-like jolt. He would fold in some acetone, he explained, “but in such a manner that you won’t feel it.” Ellena said that he wasn’t worried about ignoring his clients’ demands. “There’s always a décodage between what they say and what I am actually constructing,” he said.

In the conference room, Gautier slapped the touche marked AG3 onto the table. “I reject AG3—very clearly,” she said.

Ellena was unruffled. All three concoctions, he said, had been inspired by the Aswan mangoes. “AD2 is more lotusy, whereas AD1 is more woody,” he explained. And the rejected AG3? Ellena smiled. “AG3 has magnolia,” he said.

Gautier’s brow furrowed. “Ah, that doesn’t surprise me,” she said.

Gautier kept picking up AG3 and saying, “No!,” then throwing it down once more. Ellena said of AG3, “I put in a lot of incense.” He added with a grin, “And there’s something in all of them you’ll never guess.” The women narrowed their eyes. “Something we talked about in Aswan,” Ellena hinted.

“Papyrus?” Dubrule asked.

“Non,” Ellena said, dismissively.

After some silence, Dubrule said, “La carotte?”

“Oui!”

When Dubrule had held a green mango outside the Nubian village, she had detected a carroty tinge. “A fruit that smells of a vegetable!” she had exclaimed. Yes, Ellena had replied, you’re smelling molecules that are common to both.

Gautier was pleased that Ellena had remembered to include this element in the perfume. “Jean-Claude’s specialty is atypical things,” she explained to Dubrule, recalling that Ellena had once put essence of tomato leaf in a perfume. She stood up sharply. “Now, on the skin,” she said.

Ellena rolled up his sleeves. Dubrule sprayed his forearms with puffs of his three viscous assemblages; they settled onto his skin in small slicks.

“They smell completely different,” Dubrule said, hovering over Ellena’s outstretched arms.

“Completely,” Ellena said.

“Much less mango,” Dubrule said.

The women smelled the three fragrances several more times, glancing at each other. “It’s strange,” Dubrule said grimly. “Almost the reverse of the touche.” This phenomenon, in fact, causes significant difficulty for perfumers. The great Guerlain perfumes—Aimé Guerlain created the first, Jicky, in 1889—were all tested exclusively on human skin, never on paper. They were expressly built to blossom and fade, over time, on the body. Today’s customers, however, don’t want five fragrances on their body at the same time; they prefer to sample perfumes on paper strips. As a result, most perfumes today are constructed to smell good, for a few seconds, on a paper strip—which is a perversion, unless you happen to be made of paper. Indeed, many of today’s perfumes don’t last (Kenneth Cole’s Black vanishes as quickly as a picked green mango), and they often clash with the body’s natural smells.

Gautier’s nose moved once more over Ellena’s warm skin. “I prefer AD2 on skin,” she said.

“It’s AD2,” Ellena agreed, lightly.

Dubrule was definitive. “It’s AD2,” she said. After a pause, she added, “I’m finding a lot of rose in it. It’s not bad—it will just need adjusting.”

“It’s the lotus,” Ellena said. “I’ll fine-tune that.”

“This works perfectly for Hermès,” Gautier concluded. She did have one concern: would men be able to wear it? Both she and Ellena wanted Hermès to dispense with the archaic division between masculine and feminine scents—a mere marketing device designed to make heterosexual men comfortable with the idea of wearing fragrance. Though Gautier was, wisely, cautious of being too radical for the market, she nevertheless had decided that the Jardins collection would be unisex.

Dubrule reminded Ellena that Hermès was aiming to have settled on the formula of the “juice”—the industry term for a scent—by July 10th. (The European Union has a list of banned toxins and allergens and is constantly adding perfume ingredients to it.) Meanwhile, Gautier stared at the transparent essai. “Do we need to color it?” she asked.

Dubrule considered the question. “No,” she said.

“Good,” Gautier said. “We’re pretty happy—non?” She didn’t really mean it as a question, but it nevertheless sounded like one.

After the meeting, Ellena returned to his home, in Grasse, on the Côte d’Azur, and went to his former office at Symrise. (Hermès had not yet provided him with a lab, and Symrise had agreed to let him rent his old space.) He soon received a call from Gautier. “AD2 sent bon,” she said—it smelled good. Yet, she continued, changes were required. She and Dubrule liked the scent’s spiky freshness, but Dubrule thought that it was a bit harsh—too much like grapefruit. Both women thought that la persistance, the amount of time the fragrance lasts on skin, needed to be lengthened. And they wanted the smell of green mango to be more present on the skin. Good luck, Gautier said. Ellena was encouraged. “I have no anguish once I’ve got to the ‘Ça sent bon,’” he told me.

At Symrise, Ellena had access to more than a thousand ingredients—some natural, some synthetic. He used only a fraction of them. Ellena is a minimalist in materials and a maximalist in thought. Over the years, he has refined a sort of Bauhaus School approach to perfumery: clean scents made from deceptively simple chemical formulas.

Ellena’s best-known fragrances are Eau Parfumée au Thé Vert, for Bulgari, and First, for Van Cleef & Arpels. Just before joining Hermès, he had created L’Eau d’Hiver for Frédéric Malle’s élite collection, Éditions de Parfums. The scent was inspired by an aspect of the great 1906 Guerlain perfume Après l’Ondée. He said, “The problem—well, you can’t say there’s a problem with Après l’Ondée—but, bon, voilà, it is too opulent. Guerlain is baroque: put this in, and this, and this.” On the other hand, he said admiringly, the Guerlain scent had a marvellous sillage—the olfactory wake that trails behind a wearer of perfume. Someone once defined sillage to me, rather metaphysically, as the sense of a person being present in the room after she has left. Creating a sillage that is potent but not overpowering is tricky. With L’Eau d’Hiver, Ellena said, he wanted to pay homage to the Guerlain scent’s sillage—“but in enlightened form.” He selected elements from Après l’Ondée that were “soft, comfortable, light.” One of these was the natural essence of hay. He took some aubépine, an olfactory blend of finger paint and the wax used to clean linoleum floors, and added it to methyl ionone, a synthetic whose smell suggests iris. He then added a few more ingredients, including a natural distillation of honey. It took him two years to perfect his formula, which in the end contained twenty ingredients—very few, for a perfume. L’Eau d’Hiver smells, delightfully, of ground white pepper and cold seawater, with a touch of fresh crab. And it has a sillage worthy of Guerlain.

A master perfumer like Ellena has memorized hundreds, if not thousands, of recipes for manufacturing smells. Many complex natural scents can be conjured with only a few ingredients. The scent of freesia, he explained, is created by combining two simple molecules: beta-ionone and linalool, both synthetics. (To give freesia a cold, metallic edge, a touch of allyl amyl glycolate is added.) The smell of orange blossom is made by combining linalool and methyl anthranilate, which smells like Concord grapes.

In my presence, Ellena once dipped a touche into a molecule called isobutyl phenal acetate, which has a purely chemical smell, and another touche into vanillin, a synthetic version of vanilla. He placed the two paper strips together, waved them, and chocolate appeared in the air. “My métier is to find shortcuts to express as strongly as possible a smell,” he explained. “For chocolate, nature uses eight hundred molecules. I use two.” He handed me four touches—vanillin plus the natural essences of cinnamon, orange, and lime. The combined smell was a precise simulation of Coca-Cola. “With me, one plus one equals three,” Ellena said. “When I add two things, you get much more than two things.”

Even though Ellena’s perfumes often evoke the smells of nature, he believes that scents containing only natural materials are not, fundamentally, perfumes. The art of perfumery, Ellena believes, is the art of gracefully combining different chemicals, some natural, some synthetic. The first perfume synthetics were created in the nineteenth century. Aldehydes, which were synthesized in the eighteen-eighties, are the key to Chanel No. 5, giving the scent its powdery, soapy, luxurious signature. Synthetics such as ambroxan, which boosts wood and amber notes in perfumes, and karanal, which adds a strong woody accent, are regularly used in fragrances.

Ellena is proud to be an illusionist. “Picasso said, ‘Art is a lie that tells the truth,’” he told me. “That’s perfume for me. I lie. I create an illusion that is actually stronger than reality. Sketch a tree: it’s completely false, yet everyone understands it.” The point of Un Jardin sur le Nil, he said, was not to reproduce the scent of a green mango but, rather, to create a fantasy version of green mango.

On the flight back from Aswan, Ellena had jotted down a formula of thirteen ingredients, which had become his rough sketch for AD2. A natural essence of bitter orange, he had decided, would simulate the freshness of the green mango. And a synthetic grapefruit would evoke mango’s acidity. (Perfumers don’t use natural grapefruit, because it contains many sulfur atoms, which disintegrate to form malodors; the synthetic also has better persistance.) He would also add rosin, the resin that musicians rub on violin bows. Of his original thirteen ingredients, he eventually eliminated two. Opopanax, a synthetic that he had expected to produce a resinous smell, ended up evoking mushrooms. Another chemical, lionone, was supposed to help convey the smell of mango, but it interacted with the other materials to create the illusion of apricot. He had replaced the lionone with carrot, and it had worked: AD2.

Now he began responding to Gautier’s criticisms of AD2. He added several elements that, when combined, would heighten the scent of mango: hedione, a synthetic that simulates jasmine; methyl anthranilate, which is used prominently in Calvin Klein’s Eternity; and neroli oil, which is derived from sour-orange flowers. Ellena documented the formula of his updated AD2: he gave the name and product code for each material; he detailed how many millilitres of each ingredient were used in making the juice; and he listed the price of each ingredient, per litre.

He then created three additional variants of the original AD2, which his lab assistant mixed and lined up on his desk in tiny vials: AJ1, AJ2, AJ3. He smelled them on the touches, and was dissatisfied. All were too citrusy. He lowered the grapefruit synthetic in all of them; to all except AJ3, he added varying amounts of hexanal trans 2, a synthetic that smells simultaneously of golden apple and glue paste. He smelled his iterations again. “At the moment, I like AJ3,” he told me. “It’s the freshest.”

On June 11th, Ellena drove to Laboratoires Monique Rémy, a small company in Grasse that is one of the most rarefied suppliers of natural perfume ingredients. L.M.R. supplies Chanel’s perfume division with dozens of exclusive materials that no other house can obtain. For the wider perfume market, the company produces a distillation of tuberose—a flower that blooms on agave plants—that is one of the most beautiful scents ever created. If you need a basil, the company can supply you with a Basilic Essence, for thirty dollars per pound, or a Basilic Grand Vert Absolue, at five hundred dollars per pound. The company’s most expensive ingredient, Iris Naturelle Absolue, costs twenty thousand dollars per pound.

“This is the first time I come here as Hermès,” Ellena said, parking his Citroën outside the L.M.R. factory. He squinted at the building. “We’ll see how they treat me.” For years, the company has supplied Hermès with several of the ingredients for 24, Faubourg, including distillations of orange blossom, rose, and iris.

Frédérique Rémy, Monique’s daughter and the firm’s commercial director, is an attractive, direct young woman with dark hair. “Félicitations!” Rémy said, smiling. She has known Ellena for years; they are both natives of Grasse. They put on heavy protective glasses—corrosive solvents are used in the distilling of perfume essences—and chatted animatedly as she took him on a tour of the factory floor. (Rémy despises the glasses, which are required by French law; she kept taking them off, and at one point said, “Jean-Claude, can’t you get Hermès Eyewear to do something in these?” Ellena said that he’d look into it.) The machines were huge. Some have blades to hash grains and roots; the essence of iris, for example, is obtained from the root, not from the flower. Other L.M.R. machines make essences: odorant molecules are distilled from a flower or a fruit rind with steam at two hundred and twelve degrees Fahrenheit. It takes ten tons of oranges to create a thousand pounds of bitter-orange essence, which Rémy typically imports from the Ivory Coast. Still other machines create absolutes: smells are obtained with volatile solvents at around eighty-six degrees. The two methods extract two somewhat different groups of molecules. A rose essence includes the material’s top notes. A rose absolute gives you the base notes.

Ellena began his career among these machines as a teen-ager, extracting jasmine. His boss had a Ph.D. in chemistry and taught Ellena the science of scent. By taking just one sniff of a jasmine essence or absolute, he can tell you not only the flower’s country of origin but what kind of machine distilled it—stainless steel, aluminum, or steel.

A few companies, such as Chanel and Malle, continue to demand costly materials—the juice for Chanel No. 19, which contains an absolute of iris, supposedly costs a thousand dollars per pound—but the quality and expense of the materials that other fashion houses are willing to use have been plunging in recent years. The average cost of a perfume formula has dropped considerably in the past decade or so. Fashion houses set the maximum price of a juice when submitting their briefs. According to perfume experts I spoke with, the standard price for ingredients used to be about a hundred dollars per pound, but clients now often require that the juice cost no more than fifty dollars per pound.

Hermès is committed to giving Ellena creative liberty, and he would be allowed to decide which supplier’s materials would be used in the company’s new perfumes. Ellena was considering adding narcissus to the base for Nil, and he had heard that L.M.R. had created a new version of narcissus, called absolue narcisse de distillation moléculaire. (The company tends proprietary fields of narcissus in the center of France.) Rémy took him to the stockroom, a long and low-ceilinged room that was lined with refrigerators. She brought out the absolute of narcissus and set it before him. He leaned over and inhaled. She watched him carefully. It was a beautiful scent, with a raw green hint—and it was, he realized, wrong for Nil. It was not sufficiently tender. He got in his car, went back to his lab, and thought, That won’t work. Now what am I going to do?

Throughout the summer, Ellena continued to experiment, accenting one ingredient, eliminating another. He was pensive but not worried. “For the moment, I’m floating,” he said. “Maybe the answer is already in the perfume. Often, you just need to turn up something already in there.”

At the same time that Ellena, Gautier, and Dubrule were trying to perfect the juice for Nil, they were attempting to answer a grander question: What is an Hermès perfume? If you are Stella McCartney, you do not worry about history and tradition and craftsmanship. You launch Stella, which is a pleasant, millimetre-deep fashion fragrance—it has no persistance, no sillage, and no reference to anything other than its marketing team’s sense of the cultural pulse. But, at Hermès, Gautier had an entire collection, not a single fragrance, to worry about. The existing collection lacked a signature—an over-all stylistic coherence. Its scent Bel Ami was imbued with leather, which made sense, but what did the citrus notes of its perfume Eau d’Orange Verte have to do with Hermès? “Obviously, Hermès is leather,” Yves de Chiris, an independent perfume consultant, told me. “But how do you make leather friendly to the twenty-first century? It’s a problem. Maybe there should be a very subtle leather note in each perfume that says, ‘I am Hermès,’ that underlies but doesn’t lead. That might lend coherence.”

A signature is a difficult thing to create. Giorgio Armani perfumes have one of the industry’s cleanest signatures—a matte smoothness, the olfactory equivalent of brushed platinum. Ellena had been thinking about the signature problem. Many perfumers believed that a signature did not have to be something as literal as adding a specific note to each new scent. Indeed, a signature could be more subtle—an expression of temperament. One perfumer described the Hermès brand to me as elegant and respectful. Good taste. Jean-Michel Duriez, the in-house perfumer for Jean Patou, described Hermès to me as “restraint, delicacy, refinement.” Another Parisian perfume executive suggested that Hermès’s 24, Faubourg, with its subdued floral scent, best represented the house’s conservative ethic. “Verrouilléà tous les étages,” she said of the scent—every floor locked tight. The perfumes in the Jardin collection, which were delicate rather than lurid, fit the house’s aesthetic well, Ellena believed.

Another way to link Nil to other Hermès fragrances, Gautier and Dubrule had decided, was through bottle design. As with other perfumes in the Jardins collection, the bottle for Nil would be identical in shape to that of Calèche, one of Hermès’s most popular perfumes: a slim sheath of glass, with a rectangular profile, gently rounded edges, and a thick, heavy bottom. To distinguish Nil, the bottle would be tinted green—a nod to the Aswan mangoes.

In 2003, Hermès, a quintessentially Right Bank company, had placed Jean-Paul Gaultier, the French designer known for his avant-garde sensibility, in charge of its women’s collections. Choosing Gaultier was, in some ways, a savvy move for the staid company, but it was a risk. Hermès had been the fashion house preferred by Madame de Guermantes, whereas Gaultier was more the type to appeal to J. Lo. (Then again, in 2002 J. Lo had launched her own perfume, Glow, sales of which have exceeded eighty million dollars.) Ellena was not interested in competing with J. Lo. Indeed, he was pointedly dismissive of commercial concerns. “I’m certain that AJ2 would be the most commercial,” he said when I visited him in mid-June. He then added quickly, “But I never take a position on the commercialness.” He smelled his essais again. “I’m pretty happy with the sillage in these. The problem is still the persistance.” He paused. “But I’ll find the answer.” He paused again. “I hope I’ll find the answer.”

Ellena interrogated his four essais again and again, and finally devised a single scent, AJ, that combined elements of all of them. When he presented AJ to Gautier and Dubrule, their response was tentative. The perfume was lovely, they said, and you could tell that it had been carefully constructed. Yet they were concerned that it might be too citrusy. More important, they weren’t sure that AJ was “perfumey” enough—which is typically shorthand for saying that it lacked aldehydes. Gautier and Dubrule told Ellena to keep AJ’s enlivening freshness but asked him to give it more body.

So Ellena created AS, which was more flowery and fruity, with a bit more ripe mango. Gautier didn’t really like it, finding it too easy. As she put it, it was too “sixteen-year-old girl.” Dubrule found it pleasing at first, then tried it on her skin and pronounced it too sweet, lacking a certain elegance. And so Ellena returned to his lab, balancing millilitres of molecules against millilitres of other molecules. He began creating A*.

For Ellena, A* marked a gentle turn toward the woods. Among the natural materials, he lowered the citrus and excised a third of the bitter orange, then welded on a few Monique Rémy materials, including an incense redolent of pine trees, which costs nearly a hundred dollars per pound, and an absolute of honey, which costs a thousand dollars per pound. When, in 1976, Ellena had created the perfume First—his first perfume—he had used a hundred and sixty ingredients. There were thirty in A*. He smelled it. It was green mango, the very instant before being picked, though it was not the fruit itself. It was an idea he had created of green mango, a dreamy abstraction. And its sillage suggested the unblemished peel of the freshest, greenest sweet fruit.

In August, he returned to Paris and presented A* to Gautier and Dubrule. It was almost a month after the initial deadline. The European press launch for Nil had been set for January, and the product—which would cost a hundred and twenty-five dollars for a hundred-millilitre bottle—needed to be in U.S. stores in February. For weeks, Ellena heard nothing. Finally, Gautier telephoned him. A* smelled fantastic, she said, and it had more body and lasted longer than previous iterations. “It’s always ‘Longer, longer, longer,’” Ellena said, chuckling. “She’s terrible about that.”

By late summer, Ellena was installed in his new lab. With Gautier’s blessing, he had chosen to set up his workspace in an opulent glass-walled house outside the medieval French town of Cabris. The modernist house, which resembles a James Bond hideaway, is in a pine forest. In the living room, where he placed his simple wooden desk, there is a view of the distant Mediterranean. The chairs are upholstered with ostrich skin.

“Today, as far as I’m concerned, Un Jardin sur le Nil is finished,” Ellena said, though he was still tweaking it, “just for the pleasure.”

In his new lab, which he placed in one of the bedrooms, he had installed around two hundred and twenty ingredients. He had never even used a hundred of them, he said. He shrugged. “They are there because . . . peut-être. One day. Like a word in a dictionary. You have it. I have a tiny dictionary. I use few words.” He continued, “I want to master what I’m doing. Mastering means that for each word, every material in the formula, I know why it’s there.”

Ellena was now finishing work on a luxurious new collection of scents that would be called the Hermèssences. In Paris, Dubrule had told me that wearing an Hermèssence would be like dining with Pierre Gagnaire or Guy Savoy—“great French chefs who are going to search out unexpected contrasts. We will be able to use some very Hermès materials.” By this, she meant expensive. Her culinary description was metaphorical, but, in fact, Ellena was creating a scent called Ambre Narguilé—a narguilé is a water pipe—which smells of sliced apples wrapped in leaves of blond tobacco and drizzled with caramel, cinnamon, banana, and rum. And on his desk was a vial that contained the beginning of the next Hermèssence. It smelled, he said, like a leather bathing suit emerging from a swimming pool. He was working on a scent that smelled like leather sprinkled with sugar. His goal at Hermès, he said, was “to show that the perfume is not the result of chance but a reflection of a reasoned process.” He made a series of stepping motions with his hand, squinting at a target ahead. “When you start out, it’s more about your passions. At the end, it’s intellectual.”