The crystal and molecular structure of benzoic acid (space group P21/c, 4 molecules per unit cell) has been accurately determined from a study of projections of Qo and (0o--Oc) along the a and b axes. The molecules occur as nearly planar, centrosymmetrical dimers, with hydrogen bonds (2.64 A) between the adjacent carboxyl groups. The final difference synthesis reveals the hydrogen atoms clearly. Corrections have been made for temperature-factor variations and thermal anisotropy. The C-C bonds in the benzene ring are found to vary by 0.05 A (total estimated standard deviation 0.017 A), and the C-O bonds are 1.29 and 1.24 A (total standard deviation, 0.021 A). The benzene ring is accurately planar, but the carboxyl carbon atom and one of the oxygens are found to deviate significantly from this plane.

The first singlet-singlet n*-n transition of thiocarbonyl chloride CSC12 gives rise to a system of bands whose origin occurs near 5340A. Vibrational structure associated with the 35C1,CS and 35CWClCS isotopes has been analyzed in detail. As predicted by m.0. theory, the pure electronic jump is forbidden as an electric dipole transition, eA24wA1, and the bands observed are mainly vibvunic &*A1and A2-231 combinations. The organization of the spectrum is similar to that of the corresponding band system of formaldehyde, the resemblance being accentuated by the fact that the 1A2 electronic state of thiocarbonyl chloride is non-planar with sinlilar geometry to the 1Az state of CH20.Five (out of six) excited state fundamental frequencies, one of which is strongly anharmonjc, are identified by the isotope effect and other means. Two possible excited state structures are obtained from Franck-Condon considerations.* In 37Cl2CS, a and b are probably reversed ; but this isotope contributes only about 6 % to the total intensity. * It is always possible to construct a reference-plane (yz) through or parallel to the three atoms C1, S , C1, so that any configuration of the molecule can be represented in terms of orthogonal internal displacement-coordinate y,z, in-plane and x out-of-plane based on an origin at, say, the centre of gravity. Whether the x-coordinates of the carbon atom at any instant are positive (above the y,z plane) or negative (below the y,z plane) depends on our arbitrary choice of the positive direction. Any function of the x-coordinate representing a physical observable, e.g., a vibrational potentialfunction, must therefore be independent of our arbitrary choice of positive direction, i.e., a totallysymmetric even function of x : reflection (T in the y,z plane is one of the symmetry-operations of the system. This is independent of whether the planar configuration (x = 0) happens to coincide with a potential minimum or maximum. If the former, the molecule is planar ; if the latter, pyramidal.Similarly, any non-symmetrical in-plane configuration can always be described in terms of orthogonal (y,z) symmetry displacement-coordinates S,, S, based on a configuration in which a two-fold rotation about the C-S bond exchanges indistinguishable chlorine nuclei, and the potential-function must again be an even function of S,, Sz, so that C ~( Z ) is also one of the symmetry-operations. In three-dimensional space, C ~ ( Z ) and ~( y z ) = ov generate the point-group CzV.

By HENRY BASSETT and (in part) THOMAS HENRY GOODWIN.The present investigation has followed more or less conventional phase-rule methods so far as the less basic compounds are concerned. The more basic compounds are so insoluble that their study had t o depend much more upon finding satisfactory methods by which they could be prepared. Having found such methods it has been possible to examine their mutual relations and to determine, at least qualitatively, their regions of existence in the phase-rule diagram of the system Al,O,-SO,-H,O.So far as the three-component system is concerned, the stable fields of existence of the more basic compounds are crowded into a very small space in the neighbourhood of the water point. I t is clear that complex ionic, molecular, and micellar equilibria exist involving the numerous basic sulphates. The positions of these equilibria can be altered by change of temperature and are then readjusted only slowly, so that solutions can be obtained which have the same composition but yield different compounds. The system contains a great variety of phases, There are, besides Al,(SO,),, 1 6H,O, a t least eight well-defined crystalline basic sulphates, most of which occur in more than one state of hydration.In addition to these crystalline compounds, there are two quite different sets of two-liquid systems. Both of these form rough ellipses when plotted in an isothermal of the Al,O,-SO,-H,O system. Two-liquid system I is found in a relatively weakly basic region of the three-component system, and the second liquid phase occurs in the form of glassy discs or spheres. Its second liquid phase constitutes the so called hydroxide " precipitates obtained from sulphate solutions with ammonia and the gels thrown down from hydroxide sols with sulphate ions. The recognition of the liquid nature of " hydroxide precipitates" appears to be of some importance. These are (NH,),SO,,[l 1A1,O3,6SO3,xH,O] and 6(NH,),SO,,[llAl,O3,6SO3,xH,O]. The " alunites,"another group of basic double salts, are also important in connection with the system AI,O,-SO,-H,O. These compounds are usually given the formula MI[Al,(OH) a] (SO,),, but such varied types of substitution are possible in the crystal lattice that the composition can differ greatly from that required by the conventional formula. Alunites can exist which contain only Al,?,. SO,, and H,O.The X-ray diagrams of all the compounds obtained during the investigation have been examined, as have those of a number of minerals reputed t o be basic aluminium sulphates. The only basic aluminium sulphate minerals (other than the alunites) which are really welldefined compounds are aluminite '' A1,O,,SO3,9H,O, and a new mineral 2A1,0,,S03,xH,0, which occurs in two different stages of hydration (Bannister and 'Hollingworth, Future, 1948, 162,565).basaluminite ; ! f i e other, with x of the order of 30, has been distinguished by the name These three minerals have perfectly distinct and characteristic X-ray spectra but, for some reason, all our efforts to prepare them synt...

No abstract

No abstract