HAZZLEDINE WARREN : STUDY OF PRE-HISTORY IN ESSEX. 1 83 
these Troy Towns, the most probable suggestion which I 
have seen (but I cannot at the moment say who first 
introduced it) is that they were not for the purpose 
of confusion, like the modern garden maze (of which 
there is also an example at Saffron Walden), but for the 
purpose of guiding the feet of the performers in sacred 
ceremonial dances. In this paper, the suggestion is referred 
to that they are supposed to represent the siege of Troy. 
To my mind the converse supposition, that the ancient 
City of Troy took its name from a Troy Town, is more 
probable. In any case, they are traceable over a wide 
territory, and are known everywhere by practically the 
same name. 326 
The Maze, Tilty Abbey (G. E. Pritchett, E.N., x., 1897, p. 
184). There is a “ Troy Town ” near this place. In 
Mediaeval days, they were used for penance—the offender 
crawling round them on hands and knees. 327 
Sea Walls, Thames Estuary, date of (W. Crouch, E.N., 
vi., 1892, p. 156). These may be Roman, although this is 
doubted by Spurred. Plumstead Marsh was first reclaimed 
and walled in A.D. 1279. But many sea-walls are certainly 
. earlier, although there is no record of their erection. In 
1259, there is a record relating to the management of the 
sluice gates in the walls, which let the water out from the 
marshes at low tide. ( E.N ., vii., 1893, p. 99). A violent 
hurricane in 1090 caused the river to overflow the Essex 
Marshes ; so the banks must have existed at that date. Thus, 
if the sea-walls can be traced back to 1090 without finding 
their beginning, it seems not improbable that the Romans 
may have been their constructors. Evidence which I 
gathered in Lincolnshire suggested that subsidence occurred 
during the Roman occupation, and that the sea-banks 
there were constructed by the Romans in order to save as 
much as possible of the land which had previously been 
occupied without difficulty. We must not forget that 
considerable local subsidence may take place, by the 
vertical shrinkage of the underlying alluvium, and without 
tectonic movement, particularly when a marsh is drained 
for agriculture [118, 122, 128]. 328 
Tree-Trunk Water-Pipes (T. V. Holmes, E.N., xiii., 1903, pp. 
60-75 ; 2 figs. ; xiii., 1904, pp. 229-240). They were made 
from elm trees, as the most suitable wood, and were super¬ 
seded by iron pipes about 1808 or 1809. 
(A. M. Davies, E.N., xiii.. 1903, pp. 117-118 ; p. 303). 
Leaden water pipes were used by the Romans and during the 
Middle Ages. Wooden pipes were of later introduction. 
