the cooler. The technique will have to be rethought. 



At this point, I realize that a good part of the resid- 

 ual porcupine smell is coming from quills scattered 

 on the bottom of the cooler, not from the cooler 

 walls. That makes my job a lot simpler. I won't need 

 to collect odor from a living, thrashing porcupine — 

 freshly pulled quills will do just fine. 



David assigns the problem of structure determina- 

 tion to a talented graduate student, Guang Li, who 

 sets competently to work. We eliminate the problem 

 of room-air contaminants by filtering the incoming 

 air. Guang improves the discrimination of the system 

 until eighty-nine compounds in the porcupine odor 

 can be chemically separated. The active principle of 

 porcupine warning odor must be lurking some- 

 where among the eighty-nine peaks of Guang's 

 chroniatogram. But which one? 



Guang sets out to trap the porcupine odor in a 

 different way. Solvents such as water or alco- 

 hol vary in polarity, or the amount by which neg- 

 ative electric charge is concentrated on one side of 

 the solvent molecule and positive charge on the oth- 

 er. Guang knows that solvents of differing polarity 



It's an eerie experience, smelling porcupine 

 in a bottle — one has a sense of imprisoned 



extract different components of a mixture, so he 

 washes a cartridge containing porcupine odor 

 through three solvents of increasing polarity. 



But Guang has never encountered a porcupine 

 and doesn't know what one smells like, so he asks 

 me to smell the three extracts. The first vial is odor- 

 less. The second vial has an odor, but it is not por- 

 cupine-like. But the third vial, collected with the 

 strongest solvent, strikes the nose with a strong por- 

 cupine smell. It is an eerie experience, smelling por- 

 cupine in a bottle — one has a sense of imprisoned 

 wildness, like the unfortunate genie of Arabian folk 

 tales. On chromatography, the third vial sorts into 

 just three principal components. Two of them are 

 common compounds that can be eliminated at once. 

 The third is an unusual compound called delta- 

 decalactone, a ring-shaped molecule with ten car- 

 bon atoms, two oxygens, and eighteen hydrogens. 



To confirm this molecule is the active element of 

 the warning odor, Guang sets up an elegant experi- 

 ment. He sends the contents of an odor cartridge im- 

 pregnated with porcupine odor through a gas chro- 

 matograph. Then he splits the instrument's output — 

 the usual eighty-nine components — into two parts. 

 One part goes to a strip-chart recorder, which gen- 



erates the familiar pattern of peaks and valleys of por- 

 cupine odor. The other part goes to a biological de- 

 tector — the human nose. Locke, Guang, and I all do- 

 nate our noses for detector duty. (By this time Guang 

 has accompanied me on a Catskill visit to catch and 

 smell a porcupine, and so recognizes its unique odor.) 

 We take turns going off nose duty so that one of us 

 can annotate the strip-chart recorder. 



Then it happens. "Porcupine!" I cry out, and Guang 

 marks the spot on the strip chart. The odor builds, 

 incredibly strong because the vapors are heated, with 

 the signature of pure porcupine. Then it ends, and 

 there is olfactory silence. The peak that Guang has 

 marked on the strip chart is delta-decalactone. 



Only one small step is left: to cross-check the 

 odor against commercial delta-decalactone. 

 The small, brown bottle arrives. I unscrew the cap 

 and sniff. Oh, no, there is a strong smell of coconut, 

 not the expected porcupine. Something is terribly 

 wrong with our hypothesis. 



In the excitement of assigning a name to the por- 

 cupine odor, we had forgotten that delta-decalactone 

 is the name for two closely related compounds. The 

 two are optical enantiomers — they 

 differ from each other in the way 

 two mirror images differ, thanks to 



wildneSS t ' ie as y mmetr i ca ^ arrangement of 

 other atoms around a carbon atom. 

 Chemists call one of the pair the R- 

 enantiomer, the other the S-enantiomer. 



The commercial sample I had sniffed was a fifty- 

 fifty mix of the two enantiomers. If there is to be 

 any hope for our hypothesis, only one of the two 

 compounds smells like coconut, the other like por- 

 cupine, and the coconut smell overwhelms the por- 

 cupine. At least there is a well-known chemical 

 precedent for different-smelling enantiomers. Car- 

 vone, for instance, another ten-carbon molecule, 

 also exists in two enantiomeric forms, each with its 

 own distinguishing odor: R-carvone is the pungent 

 fragrance of spearmint, whereas S-carvone gives car- 

 away its characteristic aroma. 



So which of the two delta-decalactones has the 

 smell of porcupine? To answer the question, we need 

 two things: a specialized gas-chromatography col- 

 umn, known as a chiral column, that can separate 

 optical enantiomers from each other, and a sample 

 of authentic R- or S-delta-decalactone. Locke 

 piques the interest of an arm of Sigma- Aldrich Co., 

 a technology firm in St. Louis, Missouri, in the proj- 

 ect, and the company donates three of its Supelco 

 chiral columns to the laboratory. Guang investigates 

 chemical databases to find a chemist who has worked 

 with delta-decalactones, and we finally obtain a tiny 



52 



NATURAl HISTORY March 2006 



