INTRODUCTION, LOCATION & PURPOSE

Kidwell’s Park is adjacent to the centre of Maidenhead, see location plan "Drg 1". The park is bounded to the south by Bad Godesberg Way and is between the Marlow Road to the west and Cookham Road to the east. Centre NG SU 8869:8138 The current park occupies an area of approximately 9.34 acres (3.78 hectares). Some third of the area to the east contains hard courts, skate board area and pavilion. The remaining area is laid to grass with planters and paved paths. Roughly in the centre is a mound running north to south across the full width. At the time of the survey contractors fencing and works occupied approximately all of the eastern third of the site.

The Royal Borough of Windsor & Maidenhead proposed to carryout extensive landscaping works to the park. A watching brief has been commissioned from Thames Valley Archaeological Services of the work. It was suggested that in addition a geophysical survey should be carried out to seek information regarding the feature shown on the Ordinance Survey and Alderman Silver’s map of 1809. The intention being to identify the area of archaeology and thereby enable the landscape design and works to avoid impacting on the archaeology.

ANTIQUARIANS

The suggestion of a Roman Road seemed to arise from a map produced by Alderman Richard Silver in 1908 which was presented to the Borough in 1909. The original of this map was eventually found by Maidenhead Library and a copy of the section showing Kidwells Park is included, see "Drg 2" A redrawn version by Tony Pickford in 1975 is included in Tom Middleton’s book on Maidenhead, See "Drg 3". It was decided to research sources contemporary with Alderman Silver to seek to understand what the basis was for the then suggestion of a Roman Road. Locations mentioned are identified as (R!) and shown on "Drg 4". Those in proximity to Maidenhead are shown on a yellow square, those remote from the town shown on a black edged yellow square with an arrow pointing towards the location.

See Appendix 1

In the reminiscence by Alderman Silver [iii] he states "It was all done by erasures from the Ordnance Map, and then I filled in the names of the fields, closes, lanes etc. as known when I was a boy". It is possible and perhaps likely that the feature was already on the map used, not put on by Alderman Silver. The Ordinance Survey Office can throw no light on what map was used but advised that the feature shown would arise from the survey not local opinion. Alderman Silver states that his map took a year to complete, there is some confusion as to the year he completed it but the statements indicate that the earliest he would have commenced is about 1907.

Close inspection of the original Alderman Silver map reveals from the drafting and penmanship style, that the details of the road were by the OS and already on the edition of the OS used, not put on by Alderman Silver. The map covers the area from Braywick to near Cookham and contains a number of plotting noted as "Roman Road". On the edition of the OS used the position is marked by a single solid line as are other boundaries with the words "ROMAN ROAD" alongside.

Charles Kerry in 1861 [iv] comments on the line of the road from Braywick to Cookham. He does not directly refer to it passing across Kidwell’s only that "it reappears in a direct line with the former track a little beyond North Town" This is a location some distance north of Kidwell’s and it seems strange that he made no mention of such an apparently obvious feature closer to the town.

After the production of Alderman Silver’s map, Stephen Darby, regarding a Mr Rutland 1891 mentions the same location as Charles Kerry. J W Walker 1931 comments on the feature crossing Kidwell’s.

The implications of all of this is that it seems that the Ordinance Survey recorded the feature accurately some time prior to the Alderman Silvers map, possible is that Alderman Silver in 1907 used a version of the Ordinance Survey of 1875 published in 1881. However the road is shown in greater detail on the 1875 OS than on the Alderman Silver version. Further investigation is being carried out to seek to establish which edition of the OS was used by Alderman Silver. Important as will be seen is that in 1875 the feature was accurately surveyed.

SITES & MONUMENTS RECORD

The record shows mainly evidence of Medieval occupation in the vicinity of Kidwells Park. Records of relevance to Roman activity are:

INVESTIGATION METHOD

Roman or other early roads can best be revealed by resistivity as this will often pick up the different densities (moisture content) of materials used in construction compared with natural or fill when analysed by Surfer software. The objective was to run three grids of resistivity, 20m wide, across the feature using a TR/CIA Resistance Meter. Unfortunately site inspection showed that various obstructions, trees, contractors fences and works prevented this objective being achieved. The RBWM supplied a survey which was transferred to our CAD system in a simplified format for clarity, see "Drg 5". The resistivity data analysis and plotting system required no less than 7 lines of readings running on the long Y axis of the resistivity survey. Considerable difficulty was encountered in achieving this but by positioning the grids between trees a grid of mainly 7m wide was achieved giving 8 lines of readings but at no point was this completely over all of the feature to the east, see "Drg 6".

Two survey points (ground control points, GCP) (steel survey pins) were established at points from which most of the feature could be seen, GCP A & B. Using a GTS (EDM) the footpaths and boundary were surveyed and the angle to grid north ascertained. The pegs for the resistivity grids were set out manually so as to avoid obstruction and then recorded using EDM. Levels were taken by EDM at close intervals for the contour plot.

The EDM survey was plotted using CAD and overlaid onto the RBWM survey to check for compatibility, there were no significant variations. The contour data was downloaded onto a spreadsheet for calculation and then downloaded to a Surfer 8 terrain mapping program for plotting. This plot was then linked to the CAD survey plot.

Resistivity data was downloaded to a spreadsheet for calculation and analysis. With the amount of obstruction, tree roots, footpaths etc. data glitches are normally encountered. The analysis on spreadsheet identifies such glitches but in this case there were none that were significant. Some of the analysis can reverse the data and by using the spreadsheet to calculate reciprocal data this problem is avoided. The data was then merged for each run of grids and loaded into a modified Surfer 8 system for plotting and overlaid onto the CAD survey.

PLOTTING & ANALYSIS

The ground contour plot over the survey "Drg 7" is indicating something of possible significance. The position of the feature shown on the Ordinance Survey (1875) was plotted onto this drawing with care (red) and checked against other editions of the OS. As may be seen the position of the main part of the current bank does not seem to coincide with the early position of the feature recorded by the OS. Attention is drawn to the features shown on the 3d plot "Drg 8" identified as "Step" and "Mound".

The resistivity data was plotted in several formats, those included are for clarity are the 3d surface plot which in effect gives a virtual model of the data. These for the full grids being grids 3,2,1 "Drg 9" and grids 4,5,6 "Drg 10". The north edge of 3,2,1 is picking up signal clutter from the site boundary. Sections along grids 4,5,6 were plotted by taking data slices along a wire frame plot, "Drg 11". Both plots are showing a wide range of readings, with the trough being 61.38% of average and peak 151.78%, indicating significant differences in moisture content and therefore density. The readings on the largely undisturbed ground to the west were some 75%, emphasising the density of material producing the peaks at twice this. As may be seen the resistivity 4,5,6 appears to peak in the vicinity of the OS position of the feature. The right hand side of both plots show peaks over the current mound.

It was decided to extract the data from grids 4,5,6 for an area immediately over the OS position for the feature. See "Drg 12". This plot shows the position of the data extracted and as a 3d plot, contour and vector. The vector plot is the rate of change and direction of the resistivity data. Attention is drawn to the linear nature of the vectors generally up and down at the centre with more random difference to the left and right. Linear vectors usually indicate something artificial, natural or mixed material produces a more random pattern.

On the extract plot of 4,5,6, although the area covered is not large enough to be certain, the dimensions shown of 7m between depression and 3m of more dense material, would be consistent with road construction.

The depth of penetration of the resistivity under the conditions prevailing is some 0.75m < 1m. The difference in signal representing moisture or density must therefore come from material within this depth.

The best guess as to what is producing the peak signals, based on experience in similar material, is densely packed gravel type material.

ROAD LINE

The alignment of apparent Roman or similar roads was checked on various OS maps, one example is included. See "Drg 4" Roads marked with red broken line, alignment as blue. Interestingly all the alignments seem to come together just to the north of Kidwell’s Park but this may not be significant. All known sightings between near Braywick and the Thames at Cookham were plotted, see "Drg 13" Closer inspection of the more numerous plotting of the road on early OS between Braywick and Maidenhead revealed that the line kinked several times. In a 3k length the line sharply kinked some 4 or 5 times between 3 and 4 degrees from a straight line.