Scientists have different theories on how our noses interpret smells.

Smell is the most mysterious of the five senses - scientists are still not exactly sure how the nose decodes odors.

The sense of smell often seems like the forgotten sense, perhaps because scent cannot be transmitted as obviously as images or sound. But watch a dog - with a sense of smell about a million times more sensitive than ours - identify a person by their smell or sniff out traces of drugs and it is obvious what a powerful means of communication it can be. For humans, scent plays a big role in attraction and is strongly tied to memory.

But how is smell written into molecules? And how do our noses interact with scent molecules? Since classical times, scientists have been trying to pin down solid olfactory rules but they still don't know exactly how the nose works.

Decoding the shape of smell

What we do know is the world is made of atoms and those atoms connect to make molecules. Molecules are what we smell, from wherever they are evaporating, and they reach our nose through the air. Though we know almost everything possible about molecules, we don't know how our nose reads them. Chemists make hundreds of new molecules every week but what each molecule is going to smell like is always a mystery.

The prevailing theory, first refined in 1952 by John Amoore at Oxford University, is the shape or steric theory of odor. The theory, simply stated, proposes that the shape of a molecule determines its smell. In other words, a rose molecule smells like a rose molecule because its shape is coded precisely for the nose to interpret this way. It does this by a lock and key method within the olfactory nervous system: the shape of an airborne molecule (the key) fits into complementary odorant receptor proteins on the outside of the nasal cell (the lock). Amoore also proposed that there are seven primary odours (ethereal, camphoraceous, musky, floral, minty, pungent and putrid).

But the shape theory is not without its pitfalls. "Shapists" are plagued by the indisputable evidence that not all similarly shaped molecules smell alike, while sometimes differently shaped molecules do. Also unexplained is the fact that humans can detect many more smells than there are odorant receptors in the nose. Aware the shape theory doesn't hold up to watertight scientific scrutiny, scientists have long been pursuing other explanations, with limited success.

Scent vibrations

In 1996, Luca Turin, a biophysicist at University College London, thought he may have come up with the answer to how we smell. In his new book The Secret of Scent (Faber and Faber 2006), he outlines his hotly contested vibrational theory of smell, and explains how "…like the origin of life, the mechanism of general anaesthesia, the extinction of dinosaurs, the kinship of the Basque language, [smell] is a scientific Sword in the Stone."