Near Piddington, on the border of the same county, rises to a height of nearly 400 ft. above the village the eminence known as Muswell Hill. Close to the summit of the hill is an earthwork having the form of a square turfed level surrounded by a low bank and bearing the traditional name of "The Wilderness." It is often spoken of locally as a Roman camp, but there is nothing in its structure to suggest such an origin: no satisfactory explanation of its origin or purpose has, in fact, been hitherto forthcoming. Now the word "wilderness," as we shall see in a subsequent chapter, was employed during a certain period to denote a maze of the horticultural type, and it is not improbable that it was used in connection with mazes in general. May we not, therefore, allow it to be within the realms of legitimate surmise that this mysterious work constitutes the remains of a square turf maze, perhaps of a design similar to that of the Clifton maze mentioned above? The situation, the enclosing bank, and the regular outline of the latter accord well with this supposition, though the dimensions—250 ft. square—are rather excessive unless we allow for a considerable margin between the bank and the maze itself.
Another Oxfordshire locality, Tadmarton Heath, is mentioned as a turf maze site in a manuscript by the Rev. T. Leman, quoted in a county history published in 1861, but if such a work ever existed there all traces of it have now disappeared; the same is true with regard to another reputed maze site to which reference is made in the manuscript, namely, the Herefordshire Beacon, in the Malvern Hills. In both cases the situations are such as might well have been selected for the purpose, judging by analogy with other known turf maze sites. Sixty years is none too long a period to allow of the complete obliteration of the turf figures, if such existed, in the absence of care and attention, so that it is not surprising if we now find ourselves unable to trace them, especially if they possessed no circumscribing bank or ditch.
In Surrey a "Troy-town" was formerly well known in the neighbourhood of Guildford. It was cut in the turf on Hillbury, between Guildford and Farnham. It may be that the earth-rings of which traces are yet visible on St. Martha's Hill, on the other side of Guildford, constitute the remains of a similar work. It is said that the youths and maidens of the town used to congregate here on Good Friday and indulge in boisterous celebrations, the origin of which is not known. Another Surrey spot formerly alluded to as having a turf maze is Putney Heath. Unfortunately, however, we cannot at present point to any authentic traces of a single specimen in the whole of the county.
At Chilham, near Godmersham, in Kent, is an earthwork on the downs known as Julaber's Barrow or Juliberry's Grave. It bears no traces of mazy paths, but the name carries strong suggestion of "Julian's Bower," and there is perhaps as much force in this suggestion as in the opposing view that the mound is the grave of Quintus Laberius Durus, one of Julius Caesar's tribunes (hence Julii Laberius), who was killed by the Britons. The latter theory was, however, maintained by a writer to The Times as recently as April 5, 1920. The fact that as late as 1893, according to a letter to Notes and Queries of that year, traces of a "bower" or "Troy-town" were still observable in the neighbourhood of Walmer, shows that the Chilham work, if a turf maze, would not have been unique in the county. It is also said that one formerly existed near Westerham; the name "Troy-town," moreover, survives in other parts of the county (see p. [211]). Additional support for the theory of a turf maze site at Chilham is found in the occurrence of the name Bowerland, applied to a district to the north of the village. We find, too, a hamlet of Bower, only a few miles to the south-west.
In Bedfordshire, not far from Dunstable, there is a circular earthwork on the downs called "Maiden Bower." Stukeley refers to it in his discussion on Julian's Bowers as being in his opinion the site of a former turf maze, and there is some force in this contention. He mentions in the same reference a similar work at Ashwell.
Dr. Trollope stated that specimens had been reported also from the county of Devonshire and in Scotland, but actual details are not at the moment available.
There is no doubt that the custom of cutting these devices in the turf was formerly very widespread throughout the land, although comparatively few examples now exist. Even during the past generation, as we have seen, some are known to have disappeared. Let us therefore hope that all possible care will be taken to preserve those that remain to us.