OPTICAL PROPERTIES Refractivity.—The change in the direction of a homogeneous ray of light when it passes from air into a denser medium is termed refraction. This property of the medium is suitably measured by the index of refraction Ba=sin i/sin r, in that the fluence of the angle of incidence is thereby eliminated. The re f ractive index is determined by the relative velocities of light in air and in the medium, and this depends on the wave-length. The choice of wave-length in actual experiment is largely determined by the ease with which homogeneous beams of certain wave lengths can be obtained. Apart from the arbitrariness which is involved in the selection of a particular wave-length for the com parative study of the refractivity of different substances, it is to be noted that the refractive index (say for the sodium D line) depends on the temperature, pressure and other factors. This variability of the refractive index is responsible for the fact that other quantities have been suggested as a more suitable measure of refractivity. The expression (n— i )/d was sug gested by J. H. Gladstone and T. P. Dale, and the relation r2= / d by H. A. Lorentz and by L. Lorenz. The former is purely empirical, but the latter has some theoretical justification. Both these r quantities are insensitive to changes of temperature, and the second has the very important additional qualification that its magnitude for a given substance is nearly the same whether the substance is in the liquid or gaseous state. This is shown by H. H. Landolt's data for the specific refraction (r2) of water, alcohol and chloroform.
In order to provide an appropriate comparative basis for these measures of refractivity, it has been customary to multiply them by the molecular weights of the compounds concerned, whereby the so-called molecular refractivity R=mr is obtained. The sig nificance to be attributed to the quantity R is a matter which does not appear to have received much consideration. It may, however, be pointed out that i) / d, which was derived from a consideration of the dielectric action of insulating sub stances in terms of the Clausius-Mossotti theory, would seem to be a measure of the specific volume if n refers to infinitely long waves. It would therefore seem that the corresponding value of R may be regarded as providing an approximate measure of the molecular volume. In support of this view it may be noted further that the variations of from compound to com pound are relatively small and that the differences in the R values are mainly determined by the changes in m or m/d. The use of molecular refraction in the comparison of the refractive powers of substances would therefore seem to have less justification than is usually assumed and the procedure followed in investigations which make use of this quantity may be regarded as largely empirical.
The relations disclosed by the results show that the molecular refractivity is additive in character and that its value for a given compound can be calculated from the molecular formula by the summation of constants (atomic refractions) which are peculiar to the constituent elements. As in the case of other physical proper ties, structural factors are clearly evident, and the correlation of the data for different groups of substances has made it neces sary to attribute refraction effects to many different kinds of structural peculiarities such as double bonds, treble bonds, ring formation, conjugation, etc. When two double bonds are present in a molecule, the refractivity would seem to be dependent on their relative positions. When the double bonds are associated with a common carbon atom (C = C = C) or when they are separated by more than two carbon atoms, their influence is normal, but when they are in juxtaposition (C=CH—CH=C) or "conjugated," their influence is abnormally large. This effect is termed optical anomaly or exaltation. The constants assigned to certain atoms and structural elements 'from observations on the sodium D line are as follows:—C 2.42, H i•oo, 0 (doubly linked) 2.21, O(hydroxyl) 1•52, O(esters) 1-64, Cl 5.96, Br 8.86, I 13.9o, double bond 1-73, treble bond 2.4o.
Rotatory Power.—When a beam of plane polarized light is passed through certain substances, the plane of polarization is changed in direction ; it is said to be rotated. Whilst for solid substances this may be due to an asymmetric disposition of the molecules in the crystalline aggregate, it can only be due to asymmetry within the molecules when this natural rotatory power is shown by isotropic fluids. The relations concerned are dealt with in the article on STEREOCHEMISTRY, and attention will be here confined to similar rotation effects which are shown quite generally by liquids when these are placed between the poles of an electromagnet and the beam of light is allowed to pass along the lines of force of the magnetic field. Our knowledge of mag netic rotatory power is based very largely on the work of W. H. Perkin, Sr., who measured the rotation produced by a column of liquid of fixed length and derived therefrom the specific rotatory power r=cs/d. Comparable values were sought by the introduction of the molecular weight in the usual way leading to the molecular magnetic rotatory power defined by R = ma/d. The comparison of such R values (expressed in relative measure, with water as the standard substance) has shown the existence of relations which are on all fours with those which have been de scribed under refractivity.
Viscosity and Fluidity are reciprocal measures of internal friction. The precise definition of the coefficient of viscosity need not be considered here, for actual measurements are almost exclu sively confined to the determination of relative viscosities. Such values are readily obtained from the times which are required for a given volume of liquid to flow through a capillary tube under pre cisely similar conditions. At one and the same temperature the vis cosities for different liquids vary between very wide limits. The coefficients for ethyl ether, water and glycerol at o° C are, for instance, 0.00286, 0.01793 and 46.o, respectively. With rise of temperature, the viscosity diminishes rapidly and in any com parative study of viscosities the question of temperature is very important. The temperatures selected by T. E. Thorpe and J. W. Rodger are those at which the temperature–viscosity curves have the same slope. At such temperatures the rate at which the vis cosity changes with the temperature is the same for all sub stances, and consequently the influence of temperature may be said to be eliminated. On the supposition that the internal friction of liquids depends on the magnitude of the molecular surface, the product of the coefficient of viscosity (n) and the molecular surface has been used in the comparison of liquids with one another. Such molecular viscosities exhibit additive and con . stitutive relations of the usual kind.
Relations of a very different type are shown by certain other quantities which have been systematically examined in the liquid state. Notable examples are afforded by the Trouton ratio, i.e., the ratio of the molecular heat of vaporization ( X ) to the ab solute boiling temperature (T), and by the Ramsay-Shields co efficient, which expresses the rate of change with the temperature of the molecular surface energy y (m 4) when y is the surface tension, m the molecular weight and the specific volume). These coefficients have very nearly a constant value for large numbers of chemically different substances, and it has been supposed that deviations from the normal values afford evidence of the com bination of the simple molecules to form more complex molecules (polymerization, q.v. ; association, q.v.) . The normal value of Trouton's function X/T is 21, whilst for water the value is only 14.7. If water is a mixture of double and simple molecules in equilibrium with one another, as represented by there can be no doubt that the boiling-point will be appreciably higher than it would be if the liquid consisted entirely of simple molecules. Since the vapour consists almost entirely of simple molecules, it is also clear that the observed heat of vaporization is a complex quantity which represents the sum of the heat quan tities involved in the depolymerization of the liquid molecules and in the volatilization of the simple molecules. The assumption of polymerization thus affords a possible interpretation of the abnormality of the Trouton ratio. In the case of the Ramsay– Shields coefficient, abnormally small values can be readily ex plained in terms of polymerization, but values greater than the normal present difficulties. Recent observations moreover on the properties of surface layers lead to the conclusion that the molecular condition in the surface is very different from that in the interior of a liquid. This fact would appear to have an important bearing on the interpretation which can legitimately be given to the observed value of the coefficient in question. (See SURFACE TENSION.)