We describe the performance of the Boolardy Engineering Test Array, the prototype for the Australian Square Kilometre Array Pathfinder telescope. Boolardy Engineering Test Array is the first aperture synthesis radio telescope to use phased array feed technology, giving it the ability to electronically form up to nine dual-polarisation beams. We report the methods developed for forming and measuring the beams, and the adaptations that have been made to the traditional calibration and imaging procedures in order to allow BETA to function as a multi-beam aperture synthesis telescope. We describe the commissioning of the instrument and present details of Boolardy Engineering Test Array’s performance: sensitivity, beam characteristics, polarimetric properties, and image quality. We summarise the astronomical science that it has produced and draw lessons from operating Boolardy Engineering Test Array that will be relevant to the commissioning and operation of the final Australian Square Kilometre Array Path telescope.

A proposal for the siting of additional 14 metre antennas for the Fleurs Synthesis Telescope (FST) was previously presented by Christiansen (1976). Since that time, difficulties have arisen which preclude the use of two of the locations. In addition, solutions to the problems of data reduction for irregular antenna spacings have been found (Frater and Skellern, 1978; Frater, 1978; Frater, 1979). It has now been possible to choose the locations of the final antennas to optimise the performance of the array within broader geographic constraints — subject now to the availability of sites on ‘friendly’ land.

This paper describes the system architecture of a newly constructed radio telescope – the Boolardy engineering test array, which is a prototype of the Australian square kilometre array pathfinder telescope. Phased array feed technology is used to form multiple simultaneous beams per antenna, providing astronomers with unprecedented survey speed. The test array described here is a six-antenna interferometer, fitted with prototype signal processing hardware capable of forming at least nine dual-polarisation beams simultaneously, allowing several square degrees to be imaged in a single pointed observation. The main purpose of the test array is to develop beamforming and wide-field calibration methods for use with the full telescope, but it will also be capable of limited early science demonstrations.

Digital signal processing is one of many valuable tools for suppressing unwanted signals or inter-ference. Building hardware processing engines seems to be the way to best implement some classes of interference suppression but is, unfortunately, expensive and time-consuming, especially if several miti-gation techniques need to be compared. Simulations can be useful, but are not a substitute for real data. CSIRO’s Australia Telescope National Facility has recently commenced a ‘software radio telescope’ project designed to fill the gap between dedicated hardware processors and pure simulation. In this approach, real telescope data are recorded coherently, then processed offline. This paper summarises the current contents of a freely available database of base band recorded data that can be used to experiment with signal processing solutions. It includes data from the following systems: single dish, multi-feed receiver; single dish with reference antenna; and an array of six 22 m antennas with and without a reference antenna. Astronomical sources such as OH masers, pulsars and continuum sources subject to interfering signals were recorded. The interfering signals include signals from the US Global Positioning System (GPS) and its Russian equivalent (GLONASS), television, microwave links, a low-Earth-orbit satellite, various other transmitters, and signals leaking from local telescope systems with fast clocks. The data are available on compact disk, allowing use in general purpose computers or as input to laboratory hardware prototypes.

The dependence of oxygen isotope fractionation on ice growth rate during the freezing of sea water is investigated based on laboratory experiments and field observations in McMurdo Sound, Antarctica. The laboratory experiments were performed in a tank filled with sea water, with sea ice grown under calm conditions at various room temperatures ranging from −5°C to −20°C. In McMurdo Sound, the ice growth rate was monitored using thermistor probes for first-year landfast ice that grew to ∼2 m in thickness. Combining these datasets allows, for the first time, examination of fractionation at a wide range of growth rates from 0.8 × 10−7 to 9.3 × 10−7 m s−1. In the analysis a stagnant boundary-layer model is parameterized using these two independent datasets. As a result, the optimum values of equilibrium pure-ice fractionation factor and boundary-layer thickness are estimated. It is suggested that a regime shift may occur at a growth rate of ∼2.0 × 10−7 m s−1. A case study on sea ice in the Sea of Okhotsk, where the growth rate is modeled by coupling the thermodynamic properties of the sea ice with meteorological data, demonstrates the utility of the fitted models.

Near ice shelves around Antarctica the ocean becomes supercooled and has been observed to carry small suspended ice crystals. Our measurements demonstrate that these small crystals are persistently present in the water column beneath the winter fast ice, and when incorporated in sea ice they reduce the mean grain size of the sea-ice cover. By midwinter, larger ice crystals below the ice/water interface are observed to form a porous sub-ice platelet layer with an ice volume fraction of 0.25 ± 0.06. The magnitude and direction of the oceanic heat flux varied between (5 ± 6) Wm-2 (upwards) and (-15 ± 10) Wm-2 (downwards) in May, but by September it settled between (-6 ± 2) and (-11 ± 2) W m-2. The negative values imply that the ocean acts as a heat sink which is responsible for the growth of 12% of the ice thickness between June and September. This oceanic contribution should not be ignored in models of Antarctic sea-ice thickness close to an ice shelf.

We assessed the architecture of the medial gastrocnemius in nine children (five males, four females; age range 6 to 15 years; mean 10 years 10 months, SD 3 years 6 months) with spastic diplegia by ultrasound imaging before and after a gastrocnemius recession. The children were ambulant (seven independent, one with a posterior walker, one using crutches) before and after surgical intervention. We compared values for fascicle lengths and deep fascicular-aponeurosis angles with those from a group of normally developing children (five males, five females; age range 6 to 11 years; mean 8 years 4 months, SD 1 year 4 months). Despite a variable interval between assessments (from 56 to 610 days), fascicles were shorter (p=0.00226) and the deep fascicular-aponeurosis angle increased (p=0.0152) after intervention. Fascicle lengths of patients were similar to those in the group of normally developing children before surgery. After surgery, fascicles in the group of children with spastic diplegia were shorter than in their normally developing peers (p=0.00109). The gastrocnemius recession procedure alters muscle architecture, though the degree of fascicular shortening varied, with four of the participants in our study losing less than 10% of their original fascicular length at maximum dorsiflexion. Increases in ankle-joint power in walking, observed after surgical intervention in children with spastic diplegia, may be due to a more normal ankle position rather than to improvements in the active mechanical performance of the gastrocnemius.

PKS 1830–211 is the strongest known radio gravitational lens by almost an order of magnitude and has the potential to provide a measurement of H0, provided the lensing system can be parameterized. Attempts to identify optical counterparts, to measure redshifts, have so far proved unsuccessful and this has lead to radio and millimetre spectral line observations. We present our discovery of an absorption system at z = 0.19. A brief description is also made of our ATCA observations to measure the lensing time delay for this source.

We report the initial results of one year of continuous observations of the Sun‘s internal structure from the Michelson Doppler Imager (MDI) on board SOHO. The results have been obtained by inverting frequencies of p and f modes determined with two different methods of averaging over split multiplets. Small systematic differences between the two frequency sets depend primarily on mode frequencies, and, thus, did not significantly affect the inversions. A preliminary study of the systematic effects resulting from asymmetry of oscillation power peaks has also shown no significant influence on the inversion results. The inferred sound-speed profile is in general agreement with the previous data from MDI and ground-based networks. In the energy-generating core, the resolution is substantially improved, and the inversion results indicate a sharp negative perturbation of the sound speed in the core, tending to a positive value near the center. High-precision measurements of the f-mode frequencies have been used to determine the seismic radius of the Sun. The global asphericity estimated from frequency variation across the split mode multiplets has been found to be small, and is consistent with the asphericity during the previous activity minimum. Variations of the solar frequencies during the first year of MDI observations have also been detected.

PKS 1934–638 is an archetypal GPS source, peaking at 1.4 GHz and exhibits almost no flux density variability. VLBI images at frequencies of .843, 2.3, 4.8, & 8.4 were made with the southern hemisphere VLBI array and they reveal that the source is a 42 mas compact double. There is no detectable change in separation over the last 20 years, yielding an upper limit of ~ 0.03c ± 0.2c on any expansion velocity. The spectral shapes of the two components are remarkably similar, despite indications of finer structure on longer baselines. Magnetic field calculations indicate fields of a few mGauss and the results are consistent with equipartition.

We present high-resolution radio observations of the second Galactic superluminal radio source GRO1655-40, which was detected as an X-ray transient on 1994 July 27. Our radio radio images reveal two components moving away from each other at an angular speed of 65±5 mas/day, corresponding to superluminal motion (υ/c = 1.4 ± 0.4) at the estimated distance of 3–5 kpc. The 12-day delay between the X-ray and radio outbursts suggests that the ejection of material at relativistic speeds occurs during a stable phase of accretion onto a black hole, which follows an unstable phase with a high accretion rate. A complete description and discussion of these observations can be found in Tingay et al 1995 (Nature, 374, pp 141–143).

The Medium-l Program of the Michelson Doppler Imager (MDI) instrument on board SOHO provides continuous observations of oscillation modes of angular degree, l, from 0 to ∼ 300. The initial results show that the noise in the Medium-l oscillation power spectrum is substantially lower than in ground-based measurements. This enables us to detect lower amplitude modes and, thus, to extend the range of measured mode frequencies. The MDI observations also reveal the asymmetry of oscillation spectral lines. The line asymmetries agree with the theory of mode excitation by acoustic sources localized in the upper convective boundary layer. The sound-speed profile inferred from the mean frequencies gives evidence for a sharp variation at the edge of the energy-generating core. In a thin layer just beneath the convection zone, helium appears to be less abundant than predicted by theory. Inverting the multiplet frequency splittings from MDI, we detect significant rotational shear in this thin layer.

Two important factors for understanding the physical nature of compact steep spectrum (CSS) radio sources are determining the correct radio morphological classification of these objects together with their characteristics in wavebands different from the radio (Fanti et al. 1995, A&A, 302, 317). Seven CSS sources (linear dimensions < 30kpc for Ho = 50 kms–1Mpc–1 and α > 0.5, S ≃ v–α) have been found in a complete sample of strong southern radio sources. This group of CSS sources is particularly interesting because some optical and X-ray information is already available as part of a more general study of southern radio sources (Morganti et al. & Siebert et al. these Proceedings). The spectra of all the sources were presented in Tadhunter et al. (1993, MNRAS, 263, 999.) Here we present VLBI observations for three of these sources (0252-71, 1306-09 and 1814-63). The remaining four have already been imaged with VLBI (King et al. these Proceedings).

The Southern Hemisphere VLBI Experiment (SHEVE) program is aimed at producing high-resolution images of southern radio sources. The radio telescopes of the present SHEVE array are described below and some recent results presented.

In Zimbabwe, catches of Glossina morsitans morsitans Westwood and G. pallidipes Austen in traps baited with acetone and 1-octen-3-ol were increased by the addition of the synthetic mixture of eight phenols found in cattle urine to a level equal to or greater than those with natural urine. The addition of natural urine to the synthetic mixture did not increase catches further, indicating that the phenols account for essentially all the attractiveness of cattle urine. 4-Methylphenol and 3-n-propylphenol were shown to be the naturally-occurring components essential for attractiveness, and 2-methoxyphenol was found to reduce attractiveness. 4-Methylphenol alone was slightly attractive to both species, but only for males, increasing catches by approximately 30%. Catches of both species were increased by approximately 50% by 3-n-propylphenol. The addition of 4-methylphenol increased catches of G. pallidipes by up to a further four times, but catches of G. m. morsitans were decreased. Of 14 other phenols tested, phenol, 3-methylphenol and 4-ethylphenol increased the attractiveness of 3-n-propylphenol to G. pallidipes without decreasing the attractiveness to G. m. morsitans; (E)- and (Z)-3-(1-propenyl)phenol, potential contaminants in 3-n- propylphenol, did not reduce the attractiveness of mixtures of 3-n-propylphenol and 4-methylphenol, and the E and, to a lesser extent, the Z isomer could substitute for 3-n-pro-pylphenol in these mixtures. Mixtures of phenols which increased the attractiveness of traps to tsetse showed similar effects with targets but at a slightly reduced level.

The attractiveness of cattle urine to Glossina morsitans morsitans Westwood and G. pallidipes Austen was shown to be entirely attributable to the phenolic components which it contains. Four of the eight naturally occurring phenol derivatives (3- and 4-methylphenol, 3-ethylphenol and 3-n-propylphenol) were electroantennographically active, induced upwind flight in wind-tunnel bioassays and increased trap catches in field tests in Zimbabwe. One of the minor components, 2-methoxyphenol, had little antennographic activity but induced upwind flight in the wind-tunnel and appeared to be repellent in field tests.

Is it always necessary to cull large populations of wild animals such as elephants and hippos, when they appear to be destroying their habitat? In ORYX May 1969, C. A. R. Savory argued the case for doing so in Rhodesia, and the question of whether to crop elephants in parks such as the Tsavo and the Kruger has caused heated controversy. In this article the authors, drawing on their experience of the Luangwa Valley in Zambia, where a cropping scheme was started in 1966, suggest that what appears to be destruction there may not necessarily be so, and that the real vegetation killer is fire in the dry season. R. M. Lawton is an ecologist with the Land Resources Division of the British Directorate of Overseas Surveys, and Mrs Gough is a skilled observer of animal behaviour with considerable experience in Zambia.

In May and June 1952, as a member of the Anglo-Turkish expedition working at Sultantepe, between Harran and Urfa, I excavated a bath building in the Roman settlement at the foot of the mound. The main hall of the bath was paved with a fine mosaic, decorated with geometric patterns, some of which had affinities with motifs current at Antioch on the Orontes during the first half of the fifth century A.D. At the northern end of the pavement, incorporated in the mosaic, was a five-line hexameter inscription in a tabula ansata (Fig. 1). The inscription was almost undamaged, except for a diagonal break at the left-hand end (which resulted in the loss of about four letters in ll. 1/2, and of about three or two in ll. 3/5), and a small hole which obliterates three letters not far from the beginning of l. 1.