The survey comprised 37 squares, most of which had sides 20m in length. On the western side two squares did not extend to the full width in the east-west direction, in order to avoid treading on the growing crop in the adjacent field. The survey grid was referenced to the main site grid markers established by the Colchester Archaeological Trust. The survey was carried out using a Geoscan Research Fluxgate Gradiometer FM18. In each 20m square, readings were taken at 0.5m intervals in the north-south direction, and 1m intervals in the east-west direction, using a parallel method of survey, walking always in the north-south direction. There were therefore 800 readings per square. The data from each grid was dumped into a laptop computer on site, and finally processed on a Pentium 100 desktop machine off site. The computer programme used was the InSite programme from GeoQuest of Durham. Each 20m square in the raw data plot has been plotted without any upper or lower cut-off value, and by means of the programme, each square has been matched in contrast as nearly as possible to its neighbour. No contrast factor has been used, so the plot density in each square varies linearly from the minimum to the maximum value according to the magnetic response. Since the east and west walls of the portico run nearly north-south, the response of these features is disjointed in places due to the survey having been carried out in this direction. The north and south side show the true response. Summary of results Results of the magnetometer survey included the recording of all of the known features of the Temple, including the double portico, the cella and the great ditch. There were several anomalies recorded where instances of ferrous metal may have caused disturbances of the magnetic field. Two points recorded appear to be pits, one inside the great ditch and one outside the outer portico. There was no indication, at the north-west corner of the outer portico, of the presence of the terminus, which Atkinson and Hull recorded. A search for this wall will be the subject of a future survey. The portico foundation trenches are less well pictured at the north-west corner than at the south-east corner, possibly due to excavation in the last century. At approximately the centre of the outer portico foundation on the eastern side there is a suggestion that there may have been an entrance. The entrance to the area enclosed by the great ditch is clearly shown on the eastern side. Between the ditch and the inner portico foundation there appears to be two tracks, one leading north and one leading south from the entrance. On the north, south and west sides, outside the great ditch but within the inner portico foundation, there are line of magnetic responses, and one or two larger responses which may be pits. The north-west, south-west and south-east corners of the great ditch have slightly larger responses than other parts of the ditch. The results of the survey proved that the great ditch is not actually a square, and neither is the double portico. Some attempt has been made to measure the angle subtended at all four corners of the ditch and portico, although this cannot be highly accurate. The results show that the great ditch does not sit centrally within the portico foundations, and the distance between the great ditch and the inner portico foundation on the western side decreases as one travels from south to north. In conclusion, the survey achieved its objective, and the features of the complex were detected. However, further work is required to determine the course of the terminus. <1><2>