Electrons are produced by sending an electric discharge through a glass bulb containing a minute quantity of air or other gas, using metallic plates or wires as kathode and anode. Experiments have shown that the electrons are identical in all their properties, whatever metal is used to form the kathode and anode, and of whatever gas there is a minute quantity in the bulb. The conclusion must be drawn that identical electrons are constituents of, or are produced from, very different kinds of chemical elements. As the facts about kathode rays, and the facts of radio-activity are (at present) inexplicable except on the supposition that these phenomena are exhibited by particles of extraordinary minuteness, and as the smallest particles with which chemists are concerned in their everyday work are the atoms of the elements, we seem obliged to think of many kinds of atoms as structures, not as homogeneous bodies. We seem obliged to think of atoms as very minute material particles, which either normally are, or under definite conditions may be, associated with electrically charged particles very much lighter than themselves, all of which are identical, whatever be the atoms with which they are associated or from which they are produced. In their study of different kinds of matter, chemists have found it very helpful to place in one class those substances which they have not been able to separate into unlike parts. They have distinguished this class of substances from other substances, and have named them _elements_. The expression _chemical elements_ is merely a summary of certain observed facts. For many centuries chemists have worked with a conceptual machinery based on the notion that matter has a grained structure. For more than a hundred years they have been accustomed to think of atoms as the ultimate particles with which they have had to deal. Working with this order-producing instrument, they have regarded the properties of elements as properties of the atoms, or of groups of a few of the atoms, of these substances. That they might think clearly and suggestively about the properties of elements, and connect these with other chemical facts, they have translated the language of sense-perceptions into the language of thought, and, for _properties of those substances which have not been decomposed_, have used the more fertile expression _atomic properties_. When a chemist thinks of an atom, he thinks of the minutest particle of one of the substances which have the class-mark _have-not-been-decomposed_, and the class-name _element_. The chemist does not call these substances elements because he has been forced to regard the minute particles of them as undivided, much less because he thinks of these particles as indivisible; his mental picture of their structure as an atomic structure formed itself from the fact that they had not been decomposed. The formation of the class _element_ followed necessarily