]6 Nature Vol. 284 6 March 1980 logical and resource geology map to be Indian links persisted well into Late vibrations with zero momentum parallel to published at 1: 10,000,000 in 198 I and de Cretaceous times (E.H. Colbert, Flagstaff) the surface, as in infrared studies. Wit requested data to assist in this. Such a although it is not clear if such links were The short range interactions are much compilation will allow further testing of the provided by the Laccadive-Mascarene weaker but their effects can be separated reconstruction but it is unlikely that Ridge or whether access was along the because they can transfer large amounts of location of the major blocks will be in error Iranian-Afghanistan coastlines. Palaeo­ momentum to the electron (Ho, Willis & by more than a few 100 km. The biogeographic provinces formed a Plummer Phys. Rev. Lett. 40, 1463; 1978). importance of this map is thus to highlight significant part of the meeting, with the key Tuning the spectrometer to the non­ areas for future research. glossopterid flora now being divisible into specular electrons immediately separates A major consideration was the extent of six stratigraphic units that have, so far, these processes, but at the price of Gondwanaland. Nepal (C.l. Klootwyk, been correlated between India and requiring much greater sensitivity. We Canberra), Iran (H. Wensink, Utrecht) Australia (J.F. Rigby, Geological Survey, immediately gain two advantages: the need and Thailand (S. Bunopas & P. Vella, Queensland; S.C. Shah, Geological for a dipole moment is removed; and Wellington) were formerly parts of Survey, Calcutta). The presentation of vibrations with non-zero parallel Gondwanaland and there is evidence that maj or advances in palynological momentum can be seen. There is a further much of China (1.B. Waterhouse, correlations were predictable, but nonethe­ less obvious gain in information. The Queensland; P. Tasch, Wichita) and less of critical value for the Carbo-Permian matrix element for the scattering involves possibly some central Asian blocks, with (E.M. Truswell, BMR, Canberra) and an incident-wave momentum k j an exit parts of southern Europe (D.H. Tarling, Permo-Triassic (J. Anderson, wave momentum kr and the displacement Newcastle upon Tyne), also formed parts Witwatersrand). Although this work solves ofthejth atom, Cj , caused by the vibration: of this supercontinent at various times, the many problems it also raises others; it degree of floral and faunal interchange destroys the previous, elegantly simple 3 exp{(kj.r-kr·r)}C • VJ.j(r-R }d r (I) between the Gondwanan and Laurasian concept of a Gondwanaland drifting ~ J j j continents having clearly altered as both through polar regions and hence being J continents evolved. The distribution of glaciated at successive periods in different dinosaurs seems to indicate that Afro- areas. The age of these ice sheet deposits where Vj is the potential of the jth atom, are almost entirely Sakmarian (Lower assumed localised near Rj. If we were Permian) and require much more complex investigating the modes of a molecule, it D.H. Tarling is a Senior Lecturer in the Department of Geophysics and Planetary explanations (D.H. Tarling in Climatic would be useful to know how Cj varied Physics at the University of Newcastle-upon­ Change, (Ed. Gribbin) Cambridge within the molecule. Equation (1) shows Tyne. University Press, 1978). 0 that the contribution of each atom to the matrix element has a phase Vibrations of atoms at surfaces therefore by measuring at different values of kj - kr it will be possible to identify the from John Pendry individual components of the mode, provided only that we already know the THE vibrational properties of atoms and spectrometry can approach 1 meV mean positions Rj of the atoms. That is, we molecules adsorbed on surfaces are resolution and so the technique is mainly know the phonon 'wavefunctions' as well determined by the detailed interaction with concerned with studying strong bonds to as the energy levels. the surface: whether a molecule relatively light atoms. Despite these There is an analogy with angle-resolved dissociates, to what degree an atom is considerable restrictions EELS is proving a photoemission experiments in which the coordinated by surface atoms, whether useful technique which is continually energy and momentum of an ejected new species are formed. Knowledge of the developing both experimentally and photoelectron implies by conservation the way atoms vibrate is an essential theoretically. energy and momentum of the original state complement to our static picture of atomic There are two ways in which electrons of the electron before excitation. (Li, Tong arrangements. The traditional tool for interact with vibrating atoms: through long & Mills Phys. Rev. in the press). studying vibrations is infrared absorption, range and short range forces. If the Alternatively knowing both the original which has yielded much information about vibration has an associated dipole and final electron E and k we can infer the q vibrations but is limited by the need for the moment, an electric field extends far from and w of the photon. If we read phonon for bond to have a strong polar element to be the surface and operates on an photon we have the EELS experiment infrared active, and in the case of metal approaching or receding electron for a (whether it involves the loss or gain of a surfaces by the electromagnetic boundary relatively long period. If a dipole moment phonon). The analogy extends even condition which allows only a normal is present, this is the dominant mode for further: equation (1) gives the matrix component of the electric field. Electron inelastic scattering. (Evans & Mills Phys. element for interaction with a phonon energy loss spectroscopy (EELS) offers an Rev. B7, 853; 1973). Because the forces are defining atomic displacements Cj • The alternative means of studying vibrational long range the electron loses energy but analogous matrix element for photo­ frequencies which to some extent escapes remains undeviated in direction. A emission is from these restrictions (Ibach Phys. Rev. specular reflection at the surface turns the Lett. 24, 1416; 1970). electron around into an analyser which Experiments make use of 1-10 eV separates the inelastic component. Until electrons, which can be monochromated to recently this has been the predominant a few meVs. This represents much worse mode of operation of EELS experiments. resolution than infrared can provide and The requirement of a dipole moment where Aj is the 'A vector' of the photon severely restricts the power of the imposes some of the restrictions of infrared field at the jth atom. This close analogy technique. It is unlikely that electron spectroscopy and the lack of any should lead to rapid development of the substantial momentum transfer to the associated theory of EELS which will be John Pendry is at the SRC's Daresbury electron in the inelastic collision limits the needed to realise the full potential Laboratory. information that can be obtained to information in the data 0 0028-<>836/80/10016·01 SOI.00 © 1980 Macmillan Journals Ltd .