This paper reviews past and current research into the dreams and dream ing of people over 65 years of age. Recent studies have shown, for example, that frequency of dream recall declines with age in both men and women, but that the main decrease occurs in early middle age and is thus unlikely to be connected with aging effects. The results of studies having to do with dream tone, active/passive stance and typical dream themes are less conclusive. This paper also surveys the attitudes toward dream s one encounters among the elderly, the temporal referen ces in their dreams and the use of dream s in the psychotherapy of aged and aging persons and concludes by considering the dreams of those at the end of their life.

Volume 16, number 3 PHYSICS LETTERS 1June 1965Distorted static field. Observed annihilation rates [4] are greater than the high energy limit so one must assume considerable atomic distortion. As a positron approaches an atom from a great distance, the first interaction to take effect is via the electric dipole moment induced in the atom. The interaction energy iswhere ~ is the polarizability of the atom. There is experimental evidence for the importance of this long range dipole interaction: over a wide range of atomic and molecular gases a correlation has been observed between annihilation rate and dipole polarizability [9].As the positron gets closer, higher multipoles become important. A stationary state treatment exaggerates the distortion because it takes no account of the ~nertia of the electron cloud, but is correct for very low energies.It is possible that attraction of the positron to the atom as a result, at first, of the long range force could end in the positron attaching to the atom. And there is some evidence that this effect has been observed [4] at very low positron energies. Annihilation of an attached positron would be nearly instantaneous, so that here it is the cross-section that should be calculated.Virtual attachment and virtual positronium formation are ways of describing the short range atomic distortions. The former may be suitable for very low energies, the latter near positroni-um production threshold. Again a stationary calculation of the annihilation rate would provide information for the experimentally interesting case of very low energy.For a complete interpretation of experimental data, which is in the form of lifetime spectra, one needs the cross-section and the annihilation rate as a function of energy. And whereas previously cross-sections alone have been computed, there would seem to be no great difficulty in obtaining the rates as well.The remarkable properties and the important advantages of several forms of 'Fourier-transform spectroscopy' have now been established [1-6], following initial work by Jacquinot [4], Fellgett [5], Strong [6] and others [1,3,6]. Among the principal advantages are a simultaneous recording of all spectral elements (recording-time independent of the spectral width) and high luminosity. The method requires very accurate (ruling-engine quality) moving-mirror motion (or scanning) and computation (by Fou-272 rier transformation) of the spectrum from the photoelectrically recorded interferogram. Recent advances in holographic (wavefront-reconstruction) imaging [7][8][9][10][11], and in its Fouriertransform formulation [7,10,11], make it reasonable to investigate possible simplifications which might result from extensions of holography to such applications as spectroscopy and astronomy.In this letter we demonstrate the theoretical principles and experimental verifications of a

No abstract

In this letter we show that the "spreading" effect due to the use of an extended source in the recording of a hologram does not result in an irretrievable resolution loss, but that the "loss" may be compensated and the resolution retrieved by illuminating the hologram, in the reconstruction, by the same source as that used in the recording (or by another suitable source), provided that the auto-correlation function of the function describing the "extended" source (or the crosscorrelation of the source functions) has itself a narrow central-peak width, of the order of the desired *resolution. This is the case for suitably "structured" sources with a broad spatial-frequency representation, for example. We have verified the principle by using an arbitrarily selected "structured" source ( fig. 1) in place of the usual point sources (recording and reconstruction) in a Fourier-transform holography arrangement [1][2][3][4]. With the use of (a) "lensless" Fourier-transform holography arrangements [2, 3] (having recordability gains, with spherical plates, exceeding 103 compared to Fresneltransform holography), and (b) "structured source" holography arrangements, the two major limitations which appeared to stand in the way of extensions of high-resolution holography to the X-ray domain now appear to have been largely cleared away.Our method may at first sight appear to have some similarity with the arrangement independently considered by Denisyuk [5 ], who suggests using essentially "the same source" in the reconstruction, is in the recording. However,

The separation of local and diffuse visual field loss is important for evaluating the nature and extent of glaucomatous visual field damage. Here, five automated methods for estimating diffuse loss in glaucomatous visual fields (as measured with the Octopus G1 program) are compared. Four are taken from the published literature, and one is introduced in this investigation. It is shown that the new index (here called diffuse loss) provides the best agreement with a value determined using a more empirical approach.