The following information from Pastscape site.
"A promontory fort some 5 acres in extent was established at Lydney in or just before the 1st. c. B.C., & was subsequently, during the 2nd. & 3rd. cents. A.D., occupied by a Roman-British population engaged to some extent in iron-mining. Soon after A.D. 364-7, a temple, dedicated to Nodens, was built, & with it a guest-house, baths & other structures. About the end of the 4th. c., the buildings were surrounded by a precinct-wall, but later fell into decay, & the final phases of occupation, probably in the 5th. & 6th. cents., is represented by a reinforcement of the prehistoric earthwork. (2)
Surveyed at 1/2500.
Of the Roman remains, the excavated temple has walls up to 1.0m. high and has been restored. The bath building has walling up to 1.5m. high, and the reservoir is a shallow depression with a flagstone floor. The top of the stone-lined drain which leads from the reservoir to the baths has been fully exposed. No other buildings can be identified.
See G.P.'s AO/63/94, 2,3, & 6. (3)
One of the best-known and most interesting Roman sacred sites in Britain - a prosperous temple complex dating from after AD 364 and consisting of, besides the temple itself, a courtyard house, a suite of baths and what was probably the "abaton" of a healing centre. To the north stood a water-tank and a guest-house. Very numerous small finds with some 8000 coins (including a hoard of minute barbarous copies) prove a rich occupation extending certainly into the 5th century, and such finds as representations of dogs in stone and bronze, 320-odd pins and nearly 300 bracelets point quite definitely to a healing cult. (4)
The position of the Roman name 'NEMETOBALA (?) in the Ravenna Cosmography (where it is shown as 'METABALA') indicates a location in the countries of Monmouth (Gwent) or Gloucester; a possibility is therefore the Roman temple in the hillfort at Lydney. (5)
Excavation and re-excavation of deposits at Lydney Park, see plans (5), indicate that a date of inception for the temple complex after 364 is not tenable. (6-7)
Exacavation of the context where the corridor of the bath building impinges on the end wall of the long building (Abaton) confirmed that the bath, which incorporates a hoard datable to c360 in a repair to an eroded mosaic post-dated the long building. It was demonstrated that the mosaics of the long building, which sealed a coin of Gratian, are secondary, the original floors being flags or opus signinum. On present evidence the complex appears to date to the late 3rd or early 4th century. The present temple may not be the earliest on the site, though no predecessor has been discovered as yet. Fragments of cement floors and tesserae, incorporated in the make-up floors of the long building, indicate that a building of some pretensions existed on the site at an earlier phase. (8)
Two decorative relief carvings were noted lying close to the temple site. These apparently dated from the later 2nd or 3rd Century and suggest the possible former presence of a major building which would pre-date any so far found. (9)
Two small horizontal adits or mines (for the extraction of iron ore), discovered during Wheeler's 1932 excavation noted in a review of industrial archaeology in Gloucestershire. (10)
The entrance of one of the iron mine adits excavated by Wheeler was consolidated by Dean Local History Society in 1985. (11)
The attempted consolidation of a hypocaust system in the bath house in 1958 revealed a fissure associated with early Roman or Iron Age iron workings. these were partly excavated by C Scott-Garrett. (12)
SO 6102 The room of the temple in which the coin hoard was discovered by Wheeler was re-excavated in 1981 disclosing a phase of flooring beneath that in which the hoard was found. The excavation cast doubt on Wheeler's chronology for the site suggesting that the majority of the buildings may date from the late 3rd or early 4th centuries. A section cut through the rampart to the S of the site revealed an Iron Age core (with a sling-stone hoard), which had been heightened in 3 phases during the Roman period. These developments were suggested as representing a temenos wall around the temple. The uppermost layer (seen by Wheeler as a post-Roman re-defence) was re-interpreted as a redeposit. (13-15)
SO 616 027. Lydney. Listed in gazetteer as a univallate hillfort covering 1.8ha. (16)
Scheduled, RSM Number 28870. (17)
Parts of the Roman bath house and temple are visible as exposed stonework on aerial photographs. Most of the hilltop is covered in trees so the Iron Age fort and ironstone mining were not visible. Possible Ironstone mining is visible immediately ot the north (see SO 60 SW 73). (18)."
http://www.pastscape.org.uk/hob.aspx?hob_id=111668
Something is not right. This message is just to keep things from messing up down the road