On the Excanations at Rotherley, Woodcuts, and Bokerly Dyke. 309 
or departure of a still earlier dyke branching off to the westward 
from this spot, or a shoulder, covering the termination of the en- 
trenchment at some previous time. A section was therefore cut, to 
the south-east of the epaulement, at which spot two ditches were 
again found, as in Section 1. But in the rampart no coins were 
found, and the pottery was of an earlier and coarser kind than in the 
other sections. No distinctly Roman pottery was found in the body 
of the rampart, and only a few doubtful fragments near the surface. 
A long strip was then cut along the gap, where the upper part of 
the rampart had been removed, and where the old surface-line conse- 
quently could be got at quicker, but with the same results. Nothing 
distinctly Roman was found. 
We then attacked the epaulement itself. The rampart of the 
dyke had, at some time, been thrown over the ditch of the epaulement 
continuously in the line of the Main Rampart ; but this must have 
been done subsequently to the time when it served as the northern 
termination of the entrenchment, if it ever did so serve. The part 
of the rampart which runs across the ditch of the epaulement I call 
the “Traverse.” Was the old ditch to be found beneath the 
Traverse? If so, it would prove that it once formed the termination 
of the dyke before it was extended further to the north. . I cut a 
section along the length of the Traverse into the rampart at the 
shoulder of the epaulement, and found the solid chalk sides of the 
old ditch beneath the Traverse. ‘The section showed that the ditch 
had silted up to a great extent by denudation from the rampart 
before the Traverse was thrown over it. In the Traverse nine 
fragments of Samian pottery were found, and at 2°4ft. from the 
summit of it a well-preserved coin of Magnentius. This proves 
that the Traverse was erected in Roman times, but on digging 
further into the old rampart beneath, and at the end of the Traverse 
where it abuts upon the epaulement, no Samian or other Roman 
remains were found. The difference in the contents of these two 
deposits is made more striking by their juxtaposition. Similar 
differences in parts of the entrenchment that were remote from one 
another would prove only a difference in the previous occupation of 
the ground, but in this case it is evident that, at the time when the 
