When investigation of this point was begun, a hopeful fact came to light, as it was pointed out that certain bodies of similar chemical composition had all the same kind of smell. These were the compounds of arsenic, bismuth, and phosphorus, all of which smell of garlic. But it was soon realised that this fact was of little or no significance, as the oxides of many of the metals, although quite different from the former group, also smell of garlic. To these we may add the instance of water and sulphuretted hydrogen, two substances which are related chemically, as their formulæ show (H_{2}O and H_{2}S), and yet one of them is odourless, While the other has a strong, unpleasant smell. Finally, according to Deite, natural and artificial musk have nothing in common but their smell. Chemically they are quite different. The property of odour, then, does not depend upon the Chemical constitution of bodies. The next question that arises is: Do bodies exhaling the same kind of odour resemble each other in the structure of their molecules? In other words, can odour be related to molecular structure? To the chemist all matter is made up of atoms and molecules. The elements, bodies which cannot be broken up by chemical action into any simpler form, are composed of atoms. On the other hand, when elements combine to form a compound, the unit of the new body, composed as it is of two or more atoms of different elements linked together, is known as a molecule. (Probably the elements also exist in the molecular state, the atoms of which they are composed being linked together in groups.) Both atoms and molecules are, of course, very minute in size. For reasons we need not enter into here, the molecule is held to have a certain structural form, which form is indicated by what is known as a graphic formula. The graphic formula of water, one of the simplest, may be written as H—O—H, and we may regard it as having a linear form. (Modern views indicate that it is not a simple line, but in two planes.) Many molecules, however, particularly those of the organic compounds, are highly complex, and their structural form must be very different from that of water. The question, then, now before us is: Does odour bear any relationship to the molecular structure of bodies? And again it has been maintained that a clue to the problem of the real nature of odour lies here. There is a well-known series of chemical bodies known as the “aromatics,” by reason of the fact that they possess strong smells more or less similar in quality. With regard to this series, which is made up of groups of what are known as radicles which occupy definite positions on a molecule shaped like a ring—the benzene ring, as it is called—Henning, a German observer, has expressed the opinion that the odour depends, not upon the radicles as such, but upon the position they occupy on the ring.