Graphically each unit of combining capacity of an atom is represented by a dash added to its symbol. The valencies of different elements are thus denoted as follows: When two atoms combine, at least one valeney of each is employed, and in compounds like the following, the atoms are said to be linked to gether by single 'bonds,' each 'bond' evidently representing two `valencies' or 'affinities' (i.e. one unit combining capacity of each of the com bining atoms): The graphic representation of valency suggests an important question, viz.: Are the valencies of an atom forces acting only in certain directions, or do they act, like gravity, in all directions? A further question naturally suggests itself in the case of atoms having more than unit valency, viz.: Are the several affinities equal to one an other in power? To answer these questions is a matter not of idle speculation, but of necessity in the case—again—of the compounds of carbon. The study of these compounds has led chemists to make the following assumptions: ( 1) the four valencies of carbon are in all respects equal ; (2) they act in four different directions, which are perfectly symmetrical with respect to the carbon atom. The carbon atom is, namely, imagined to he placed at the centre of a regular tetrahedron, and four equal forces are assumed to act in the directions of the four vertices of the tetrahedron. A further assumption that thrusts itself upon the organic chemist is that in every compound capable of independent existence all the valencies of the constituent atoms are satis fied by combination, and that no valency is 'tree.' \\ it bout these assumptions organic chemistry can make no progress. These assumptions made, there is hardly a general tact that remains un accounted for. The assumptions, though hypo thetical in character, are therefore incorporated as principles of science, and thus in connection with the CoMponwk of earl on chemistry answers in a sense the question stated at the beginning of this article. viz.: In what manner does affinity aet in holding together the atoms of compounds? In the ease of other elements than carbon, the application of the idea of valeney has been much less useful and much less successful. In fact, the obstacles in the way of consistently applying the idea to the several elements are so great that the idea would probably been abandoned long ago, were it not for its great usefulness in tile ease of carbon. The chief obstaeles are as follows: Firstly, the valencies of most elements are found to be variable and hence unreliable as a basis for predicting the constitution of un known substances. Thins, while in ammonia the atom of nitrogen is trivalent (be cause combined with three univalent atoms of hydrogen), in nitric oxide (No) it is di-valcnt (because combined with one di-valent atom of oxygen), and in ammonium chloride (NII,Cl) it is lante-valent (because combined with five uni valent atoms, viz. four hydrogens and one chlorine). In other compound:: nitrogen seems to have still other valeneies. Turning to iron,
we find it di-valent in ferrous chloride (FeCI,) and tri-valent in ferric chloride ( Chlorine is univalent when combined with hydro gen, and quinqui-valent when combined with oxygen. Sulphur is di-valent when combined with hydrogen, and hexa-valent when combined with oxygen. Phosphorus is trivalent when combined with hydrogen, and quinqui-valent when combined with oxygen. Oxygen is di-valent in nearly all of its compounds; yet in di-methyl ether hydrochloride oxygen must be assumed to be quadrivalent. Even in the case of en rbon an exception is known: in ordinary carbonic oxide (C0) the carbon atom is apparently di-va lent (because combined with one di-valent atom of oxygen)—unless we assume that the oxygen atom in this compound is quadri-talent. and hence that the compound is an exception to the rule according to which oxygen is di-valent. Further, it has been stated above that the atoms of hydrogen, chlorine, iodine, and sodium were primarily assumed to be uni-valent. One might therefore expect that in all combinations of any two or three elements one atom of one would combine with one, and only one atom of the other. Yet the compound called tri-eldoride of iodine has the formula ICI„ Is iodine trivalent in this compound? Another compound. a hydride of sodium, has the formula. is hydrogen in this compound divalent? And is. therefore, the valeney even of hydrogen variable? Again when we find the molecule of hydrogen gas to be made up of two hydrogen atoms, we conclude that the affinity of each of these atoms is satis fied by that of the other atom. But the molecules of certain univalent elements (the vapors of sodium, potassium. iodine, at high temperatures, etc.) are known to be made up each of a single atom. Are the affinities of these single atoms `free?' Or shall we accept the verdict of organic chemistry, according to which the molecule of a substance capable of independent existence can contain no 'free' affinities? But then how can a single atom form a molecule? While we thus search in vain for an explanation as to what be comes of affinities iu certain compounds, we find that other compounds seem to involve the use of more valencies than those possessed by the con stituent atoms. Examples of such compounds are presented by crystallo-hydrates, like NaCl. 21-1,0, made up of several molecules within each of which all the available valeneies should be expected to be satisfied. Other examples of this kind are presented by many of the minerals found in nature. Do atoms, then, possess ad ditional valeneies which sometimes do and some times do not come into play? All of these questions remain, as yet, un answered. And hence while the conception of valency serves as an excellent working principle for the correlation of the compounds of carbon, many important farts remain for the present be yond its reach. See CHEMISTRY; CARBON COM _POUNDS ; STEREO-CHEMISTRY.