The best artificial semi-permeable material thus far discovered. especially well adapted for separating water from dissolved sugar, is a mem brane of ferrocyanide of copper, formed by the action of potassium ferrocyanide upon copper sulphate. Pfeffer. who was the first to employ this substance for measuring the pressure of substances in solution, proceeded as follows: He filled a porous clay cylinder with a solution of copper sulphate and immersed it in a solution' of potassium ferrocyanide; the two solutions, pene trating into the clay from the opposite sides, yielded a precipitate of copper ferrocyanide where they met within the walls of the cylinder, the walls serving to impart to the precipitated membrane considerable mechanical resistance. The cylinder was now filled with a solution of sugar, its upper end was tightly closed with a lid bearing an ordinary mercury manometer, and the apparatus was placed in pure water so that the level of the latter was precisely the saute as that of the solution within. To understand the phenomenon that followed. imagine a cylin drical vessel ABCD in which, say, air has been compressed within the volume FWD, while the space ABFE is empty; it we relieve the piston EF, it will be driven up IT the expansive power of the air until it is stopped by AB or by sonic other resistance; if, instead, we hold up the cylinder in the air by the handle, the expansive power of the compressed air will cause the entire volume ABCD to move over, the result being, again, a larger space occupied by the air.
Precisely analogous phenomena, would be observed if EFCD were filled with a solution of sugar and ABFE were pure water, while EF were a semi permeable membrane: Either the piston would move upward or the entire liquid volume (pure water plus solution) would move in the direc tion of the dissolved sugar; in either case the cause would be the expansive power of the sugar and the result an increase of the volume occupied by it, i.e. an addition of pure water to the solu tion. As a matter of feet, in Pfeffer's apparatus, the semi-permeable walls being fixed, the expan sive power of the dissolved sugar caused pure water to enter the clay cylinder. The increasing amount of liquid naturally caused an increasing, compression of the air within the cylinder, and finally a point was reached when the expansive power of the sugar was no longer capable of over coming the resistance of the air, the latter having grown precisely equal to it. Then equilibrium ensued, the mercury manometer showing the pres sure of the air within the cylinder, and hence the equal of that pressure—the 'osmotic pressure' of the sugar in solution. Similar experiments showed: ( ) That the osmotic pressure of sugar and other substances in dilute solutions is pro portional to the concentration, i.e. inversely pro portional to the volume of the solution; (2) that the osmotic pressure of sugar and other substances in dilute solution is proportional to the absolute temperature ( i.e. the centigrade temperature plus
273 degrees) ; (3) that the osmotic pressure of substances in dilute solution is equal to the pressure that the solute would exert if removed from the solution, vaporized, and inclosed within an empty volume equal to that of the solution, at a temperature equal to that of the solution. In brief, the laws of gases, viz, the law of Boyle and Mariotte, the law of Charles and Gay Lussae. and Avogadro's rule, hold good in the ease of dilute solutions as they do in the ease of gases. Further experiments have shown, be sides, that the osmotic pressure in solutions is the same no matter what the solvent.
The importance of these results will be evident to those who realize that the science of chemistry is based on the laws of the gaseous state, Avo gadro's rule, which embodies those laws, being the only sure guide in finding those comparable units of compounds—their molecular weights. ( See CHEMISTRY; MOLECULES — .MOLECULAR WEIGHTS ; AVOGADRO'S RULE; ATOMIC WEIGHTS; GASES, GENERAL PROPERTIES OF; etc.) Yet a majority of compounds are non-volatile, and therefore our theoretical knowledge of them re mained uncertain, and in many cases vague, until the above results proved that what we can learn of a substance by studying it in the gaseous state we can learn with equal certainty by studying it in dilute solution. Very few indeed are the sub stances that are neither volatile nor soluble in any liquid. Direct methods for measuring osmotic pressure, like the one described above, have been of importance only in demonstrating the funda mental laws; the experimental difficulties in volved render their use for determinations of molecular weights practically impossible. But, on the other hand, it has been shown that the depression of the freezing-point or the elevation of the boiling point caused by dissolving a sub stance in a given liquid is proportional to the osmotic pressure in the solution; and so. molecu lar weights are now generally determined by ob serving the freezing-points or the boiling-points of solutions. (See MOLECULES—:BOLECULAR `EIGHTS; FREEZING-POINT: BOILING-POINT.) At first, experimental research seemed to show that compounds of three importent classes, viz. acids, bases, and salts, do not obey the laws of osmotic pressure ; their osmotic pressure was found to be much higher than it should be theoretically. But Arrhenius's theory of elec trolytic dissociation (see DISSOCIATION) soon en me to add itself to the theory of osmotic pres sure, and, instead of disproving it, only furnished further proof of its correctness, just as the phe nomena of chemical dissociation, when correctly understood. hail once furnished additional proof of the reliability of Avogadro's rule for gases. See AVOGADRO'S RULE.