C H E M ■other cases between the molecules of solvent and solute. The existence of compounds such as potassium triiodide renders it probable that the electrolysis of conducting chlorides, such as silver chloride, is conditioned by the attraction which they exercise for chlorine, and the behaviour of protochlorides in comparison with that of perchlorides is readily explained from this point of view. Among the objections to the ionic hypothesis in explanation of chemical change, that afforded by the behaviour of enzymes as hydrolytic catalysts is one of the strongest. The hydrolysis of ethereal salts and of carbohydrates such as cane sugar by acids is supposed to be conditioned by the hydrogen ions, but enzymes—which act even more powerfully than acids on the carbohydrates— are not electrolytes, and therefore cannot be supposed to furnish hydrogen ions. Moreover they act selectively, a particular carbohydrate as a rule being hydrolysed only with the aid of a particular enzyme ; in fact, there must be a correspondence somewhat like that of key to lock, and there is no doubt that the geometrical configuration of the enzyme relatively to that of the carbohydrate is the determining factor on which its action depends. Although it is unquestionable, however, that besides being in many respects incompatible with the facts and Dissocia= ^rra^ona^ the ionic dissociation hypothesis tion of makes demands upon the imagination which solvent it is practically impossible to grant, it is C
ec uall exes °iTxes i y as beyond question inthateffect the behave substances ' classed electrolytes as though they were dissociated in solution. What is the explanation of this remarkable circumstance? Can any rational process be imagined comparable with that of ionic dissociation and susceptible of similar treatment? Such an explanation is perhaps not far to seek, and may be found in the fact—too commonly overlooked—that in solvents generally there are always present two sets of molecules in equilibrium, viz., simple or fundamental molecules and complex aggregates formed by the association of several such simple molecules. The internal state of a mass of liquid must be pictured as one of intense activity and turmoil—as one in which dissociation is constantly going on, but inter-molecular, not intra-molecular or ionic dissociation. Liquid water, for example, is a mixture of fundamental molecules or monads such as are represented by the symbol OH2 with complexes formed by the union of several of these simple molecules, in proportions depending on the temperature, and these two kinds of molecules are ever changing backwards and forwards the one into the other. The monads are presumably the chemically active molecules, but the complexes have little if any chemical activity. If it be assumed that chemical change and electrolytic conduction in liquids are dominated by the process expressed in the equation of equilibrium Ms = irM, M being the symbol of the fundamental molecule or monad, any condition which increases the proportion of monads will render the solvent more “active.” The effects produced on dissolving non-electrolytes are wrongly attributed to the dissolved substance; they are in reality due to the solvent monads, and it is the solvent, not the dissolved substance, that alone undergoes alteration in the case of substances which enter into solution as monads. Thus, to take the case of the osmotic pressure developed on dissolving any non-electrolyte in water—if a vessel containing water be divided into compartments by a permeable diaphragm, the equilibrium will be the same on cither side, and therefore the pressure will be the same. But if a substance which cannot penetrate the diaphragm be dissolved in the water in the one compartment, its
I S T R Y
741
molecules will interpose themselves as screens between those of the solvent, and prevent the reassociation of the monads •, consequently these will be present in a greater proportion in the solution than in water. The monads will necessarily exercise an attractive influence over those in and beyond the walls of the diaphragm, and a flow will take place into the compartment containing the dissolved substance until equilibrium is restored. The osmotic pressure developed is therefore a measure of the extent to which the dissociation of solvent complexes into monads takes place. All substances which enter into solution as monads, if non - electrolytes, should produce the same effect when present in equimolecular proportions under like conditions of temperature and pressure — and Avogadro’s law should therefore hold for dilute solutions. This is the well-known classic conclusion formulated by van t Hoff. In the case of associated non-electrolytes a disturbing factor is introduced if the molecule only gradually undergo dissociation into monads as dilution proceeds, and the behaviour of such compounds must be irregular. The depression of the freezing-point and the rise in boilingpoint brought about by dissolving a substance in a liquid are necessary consequences of the presence of an increased proportion of monads. Electrolytes, besides acting as mere mechanical screens, also exercise a specific influence in excess of that exercised by non-electrolytes, by themselves combining with solvent molecules. In such cases an attractive influence is exerted not only between the solvent monads, but also between these and the solute monads. If it be granted that electrolysis is the outcome of an interaction of the fundamental molecules of solvent and solute under the influence of an electromotive force, there is no difficulty in understanding the apparent increase in conductivity of electrolytes as the solution is diluted. The apparent effect produced by a dissolved substance must be greatest in a highly dilute solution, because in such a solution the solute is resolved to the maximum possible extent into its fundamental molecules, and these come most fully under the influence of the solvent monads, which are then necessarily present in maximum proportion. From the point of view of the hypothesis here developed ions need have no separate existence, and the phenomena exhibited by solutions—whether physical or chemical— are mainly influenced by changes in the molecular composition of the solvent, the chief if not the only factor by which the changes are conditioned being the obscure function of the atoms termed residual affinity. It will be appropriate to introduce here a brief reference to the nature of the chemical interchanges involved in vital processes, whether in plants or animals. Owing to the great advance in our knowledge, Chemistry especially of synthetic processes, it is possible to ot Vltal form some conception of their general character, although such changes are necessarily to a great extent enshrouded in mystery. In the main they are of four kinds—reductive, synthetic, hydrolytic, and oxidative. It is generally believed that the first stage in assimilation involves the reduction of carbon dioxide under the influence of light, and that it proceeds pari passu with the separation of oxygen, somewhat in the manner formulated in the equation : C02 + H20 = COH2 + 02; it is difficult, however, at present to regard this as more than a mere statistical expression. To judge from its properties, chlorophyll belongs to the class of quinones. But a quinone cannot cause reduction; on the contrary, it acts as an “ oxidizing ” agent in consequence of the readiness with which it combines with hydrogen, forming a hydroquinone or quinol. It is, therefore, probable that chlorophyll first undergoes reduction at the expense of some
