414 REPORTS ON THE STATE OF SCIENCE, ETC. 
during the centuries immediately before and subsequent to a.D. In the hope 
of finding something more datable, two stretches, thirty-five and forty feet long 
respectively, of the ditches belonging to this first hill-fort were completely 
emptied, but without better success. 
In a section cut immediately west of the south-east entrance, a new feature 
was revealed in a half-moon, bastion-like projection of the wall belonging to 
the hill-fort of the second period; this was built against the ruins of the 
wall of the first hill-fort and curved round to fcrm a previously undiscovered 
sidewall of an entrance also belonging to the second period. Four roadways 
had already been found in this main entrance of the hill-fort, three super- 
incumbent and one a few yards to the east. But definite ruined sidewalls 
belonging to three only of these roadways had been previously discovered. 
Now, the existence of four consecutive main entrances is proved by ruined 
sidewalls as well as by roadways. 
Further patches in the interior area of the hill-fort were also excavated. 
Here many relics belonging to the last occupation were unearthed, similar to 
those previously described; also, on the same horizon, for the first time upon 
the site, several pieces of native hand-made pottery, showing that this rude ware 
continued to be made here after the importation of the better Romano-British 
vessels. 
To sum up the results of the excavations made in this hill-fort during five 
seasons. They show that— 
1. There were dwellers upon this hilltop before its fortification, as proved 
by relics found. 
2. There were four successive hill-forts built by the natives upon the hill- 
top, the ruins of which have been brought to light. 
3. Each of these four hill-forts in turn suffered attack and forcible destruc- 
tion from the hands of an enemy. 
4, The first hill-fort was defended by a stone-faced rampart with a deep 
ditch cut in the rock some ten feet in front of it. Its destruction was shown 
by relics to be during the later Iron Age, and it was so complete that one 
can hardly doubt that it was the work of Roman legicnaries invading the 
district. ‘The remarkable Late-Celtic bronze horse-trappings found with quan- 
tities of human bones at the foot of the precipice at the west side of the hill-fort 
would seem to synchronise with this. 
5. The native builders of the second hill-fort constructed oblong guard 
houses at the main entrance. This apparently points to an acquired knowledge 
by this time of Roman methods. 
6. The second and also the third construction at Dinorben, like the 
neighbouring hill-fort at Pen-y-Corddyn, were apparently occupied for short 
periods only, as shown by the small amount of silting found at the bottom of 
the ditches and by the paucity of relics left behind by the occupiers. 
7. The destructions of both the second and third hill-forts occurred before 
peaceful trading relations between the natives of this district and the Romans 
were established, as shown by the absence of Roman objects in the layers of 
occupation of these periods, so far as they have been uncovered. 
8. The natives returned in numbers to the hilltop later; they refortified 
themselves there and occupied the stronghold continuously during the third 
and fourth centuries a.p. During this period they traded extensively with the 
Romans. 
9. Apparently about the close of the fourth century the stronghold met 
with its final destruction, probably at the hands of Irish and other raiders 
from overseas. 
ne 1 eee 
