How do the mechanics of smell work?


Dear Straight Dope:

When an aroma reaches whatever internal sinus patch is responsible for the sense of smell, what is actually making contact with my body? Are scents particles or waves or is there some kind of bizarre dual nature like light? I am concerned that when I walk into a public restroom, poopie particles may actually be adhering to the insides of my nose and mouth. So what's the dope on the mechanical nature of scents?

SDSTAFF Fierra replies:

Cecil has addressed the fundamentals of this topic before, Rich, if only tersely. I refer to his column on how scratch-n-sniff cards work, wherein the Perfect Master writes that smell is “is a matter of molecules dislodging themselves from a substance’s surface and finding their way into your nose.” Here, however, we’ll go into the subject in somewhat more detail. These airborne molecules are called odorants, and the place where they get detected isn’t in your sinuses but rather at the top of the nasal passage.

From the beginning: Every time you inhale, you draw air into your nose. The air passes over structures called turbinates – seashell-shaped, mucus-covered bones that serve to (among other things) filter it, heat it to an appropriate temperature, and direct it, together with any odorants it carries, in an orderly fashion to the olfactory epithelium. In humans (a group that still makes up the bulk of our readership), this is a region of about one inch by two inches situated on the roof of the nasal cavity. It contains something like five million olfactory receptor neurons, and these connect directly into the olfactory bulb, a part of the brain that rests right behind the bone at the top of the nasal apparatus.

Projecting from these neurons into the layer of mucus covering the epithelium are cilia, tiny hairs containing the receptors themselves – the specialized proteins that bind to odorants. The chemical interplay between receptor and odorant is what’s known as a lock-and-key reaction: each receptor has a particular shape, allowing only molecules of a particular corresponding shape to bind with it. When an odorant binds to a receptor, the neuron transmits a certain electrical signal to the brain via the olfactory bulb and olfactory nerve.

Although humans can distinguish over 10,000 odorants, the sense of smell is very like the sense of taste – just as we experience complex flavors via combinations of five basic taste-receptor types (sweet, salty, bitter, sour, and umami), we have seven primary odorant groupings. These are camphoric (think mothballs), musky (like certain animal smells), rose, peppermint, ethereal (like various solvents, e.g., dry-cleaning fluid), pungent (e.g., vinegar), and putrid (like rotten eggs or other sulfurous stuff). Primary odorants bind to our receptors in only one way, producing a unique olfactory signature. What are known as secondary odorants can bind, albeit more weakly, in two different ways to a combination of receptors and generate a composite signal. (As a certain combination of blue and red is perceived by the brain as magenta, setting off two receptors in this way produces a chemical reaction distinct from the one associated with either receptor on its own.) Some chemicals with the same formula but a different molecular structure – like orange and lemon scents, or spearmint and caraway – smell different, because they bind differently to different receptors. Rats are much better at detecting this sort of difference than we are; some pairs of mirror-image odorant molecules smell identical to us, but rats can tell them apart just fine.

It’s those electrical signals sent by the receptor neurons that the brain translates into smells. Some smells, like that of ammonia, are associated with single molecules; others, like the smell of a garden, are composed of a huge number of molecules in combination – various floral odorants, those emitted by greenery when it transpires, geosmin (the smell of wet earth), and so on. Your brain identifies scents via the relative strengths of the various components.

There’s plenty we don’t yet understand about the way smell works, but there’s been a lot of progress in the last 15 years or so. At this point we do know it’s strongly linked with memory (see Proust, Marcel) – the brain learns to recognize scents by recalling what was going on in the vicinity the first time it smelled them. During one’s formative experiences with, say, roses, the brain seems to tag the odorant signature with other sensory information retrieved at the same time: maybe what the flower looks like, maybe how the petals feel on the skin.

So yes, the smell you associate with an overused restroom really is caused by particles of feces, urine, etc, arriving inside your nose and binding with the receptors there. The good news, such as it is, is that some of the poop molecules (as well as much of the dust, germs, etc you breathe in) get stuck in the turbinates and never make it to the deepest recesses.

And remember, Rich, you could have been born a snake. The smelling procedure for a snake involves flicking out the tongue, catching airborne molecules on its damp surface, and transferring these to a sensory apparatus on the roof of the mouth called Jacobson’s organ. Think about that during your next bathroom break.

Now to Steven’s question. In general, yes, most things that emit a smell are doing so by losing infinitesimal amounts of their mass to the air that flows around them. As it happens, though, in the case of keys or other metallic-smelling objects, that’s not really what’s going on.

Again, there have been some significant advances in our understanding of smell in recent years. In 2006, Virginia Tech researchers Andrea Dietrich and Dietmar Glindemann demonstrated that when we smell the metallic odor we typically associate with iron, we’re not actually smelling the metal itself – there are no iron atoms in the odorant molecules. What’s happening is that the metal causes the oxidation of lipids (small fatty molecules) produced by the skin, and it’s the resulting chemical compounds – aldehydes and ketones – that we think of as smelling metallic.

The chemical interaction between blood and skin produces the same compounds, so it’s no surprise that blood is typically described as having a metallic tang to it. Humans are very sensitive to some of these substances – we can smell one such compound, called octeneone, at concentrations as low as 5 parts per trillion – leading the researchers to theorize that way back in our predatory past, this sensitivity may have helped us to track wounded prey.

It makes sense that there’d be an evolutionary advantage to being able to perceive certain other smells at very low concentrations – the smell of spoiled food, for instance, or those of naturally occurring toxins. We’re quite sensitive, for instance, to the smell of hydrogen sulfide, which has an odor threshold of 0.2 ppt. The catch, though, is that in concentrations above 150 parts per million hydrogen sulfide deadens the sense of smell very quickly, meaning it’s possible to get a lethal dose (800 ppm over five minutes will do it) without realizing it.

And sometimes evolution just gets it wrong: while some folks can smell cyanide at concentrations of 2 ppm – well below the 100-500 ppm level at which it becomes dangerous – about 20 percent of the populace is genetically unable to smell it at all.

Send questions to Cecil via


Comment on this Column