The Roman fort at Gosbecks has been known from a published aerial photograph taken more than twenty years ago. The feature has not been photographed since that time. In 1996 Peter Cott carried out a resistivity survey of the area, and a second resistivity survey was made by Richard Archer in December 1998/January 1999. A magnetometer survey was carried out by Peter Cott in February 1999. In 1996 a site grid did not exist but by 1999 a permanent grid had been established by the Colchester Archaeological Trust. Comparing the two resistivity surveys of the fort, both sets of plotted results show the lines of the plough running in a north-east to south-west direction. Both surveys showed the same results; a series of wide ditches forming a rough square and the possible north ditch of the Roman camp. The wide ditch or watercourse also appears in both plotted results. These features also show up in aerial photographs of the area. Neither of the surveys show any other feature which might be related to the fort. The magnetometer survey results were very disappointing. The two parallel linear features on the southern edge of the plot represent the southern boundary of the fort. The single linear feature at the northern edge of the plot may be part of the northern defences of the fort, and the position agrees with the resistivity results for this feature. The north-south dimension of the fort derived from these results is 150m, whilst in the resistivity surveys the distance between the northern linear feature and the most southerly of the two parallel features at the bottom of the plot is 143m. Bearing in mind the difficulty in producing a plan of the fort from an aerial photograph, these two measurements are in reasonable agreement. There are small linear features representing ditches in the places where large ditch-like features are present in the resistivity plots. This discrepancy is not understood, since the depth of detection is roughly the same for both methods of survey, and all three plots were taken at the same time of the year. This lack of response to known features has also been noted at sites at Stanway, Birch and Abbotstone, all of which are on very gravelly soil. A tentative conclusion is therefore that this kind of soil does not retain the magnetic evidence upon which the magnetometer method depends. However, the Romano- Celtic temple site at Gosbecks is less than a mile from the fort and good results have been obtained there. Conclusion The two resistivity surveys show the same features. The position of the northern defence of the fort is therefore tentatively identified. The wide ditches are possibly old watercourses. The magnetometer survey shows the tentative position of the southern defences of the fort, and possibly the northern defence line. No other features within the fort have been detected. It is recommended that, if possible, a Caesium magnetometer should be hired when the crop has been harvested, as a further attempt to define the monument. <1>