CELTIC AND ROMANO-BRITISH HERTFORDSHIRE 



In both types the houses were seldom, if ever, carried higher than the 

 ground floor, and were usually built on a foundation of masonry, above 

 which were timber and plaster, the roof being covered with tiles. They 

 were often fitted with hypocausts and bath-rooms, their floors were 

 decorated with mosaics and their walls with paintings. These types occur 

 in both town and country, though eminently unsuited to the former, as they 

 do not adapt themselves for arrangement into streets. Such were the 

 dwelUngs in Verulamium and the villas at Boxmoor and Purwell Mill, 

 The remains of the other villas found in the county are too fragmentary to 

 show their type. 



Very different, however, was the building discovered at Sarratt, which 

 was oblong in plan, and was, apparently, divided longitudinally into three 

 parts by two rows of timber posts. At one end there was an apse or room. 

 Similar buildings have been found at Spoonley Wood in Gloucestershire, 

 Ickleton in Essex, Clanville, Carisbroke, Castlefield near Andover, and 

 Petersfield in Hampshire and Mansfield Woodhouse in Nottinghamshire." 

 In certain instances, as in the case of the building at Sarratt, this type 

 of house apparently bears some relation to a larger dwelhng or villa 

 adjoining and may have been the residence of the bailiff. 



The hoards found in the county tell us little. An early hoard found at 

 Kimpton, so far as the coins have been identified, covers the period from 

 B.C. 43 to A.D, 15, and a number of coins discovered at Hemel Hempstead, 

 possibly not a hoard, date from b.c. 144 to a.d. y^. No reason for the 

 depositing of either of these can be assigned, but a hoard found at Ashwell 

 which ends about a.d. 180, a recognized hoard period, may have been 

 hidden on account of the Pictish raids of the time of Commodus and the 

 subsequent mutiny of the Roman army in Britain. The hoard discovered 

 at Brickendon, which ends about a.d. 250, and that at Aldbury, which ends 

 about a.d. 272, are probably connected with the troublous times of the 

 latter part of the 3rd century. The later of these dates is perhaps the 

 commonest for hoards in this country and refers to the period of disturbance 

 about A.D. 260—80.'° Another hoard from Cheshunt dates to about a.d. 365, 

 again a disturbed time in Britain by reason of the incursions of the Picts 

 and Saxons. 



The only vestige of anything like an industry of which evidence has 

 been discovered is the manufacture of pottery. Kilns have been discovered 

 at Radlett, where the name of the potter ' Castus ' can be assigned and the 

 type of pottery, principally mortaria, and the mode of firing could be 

 distinguished. Kilns are thought to have existed also at Aldenham, Great 

 Amwell and Hitchin. These, however, were merely local provisions for 

 local needs and cannot be dignified with the name of an industry. 



Verulamium 



The site of Verulamium," which lies a little to the south-west of 

 St. Albans, has not been systematically explored and so the history of the 



3* cf. F.C.H. Hants, i, 296, 316 ; F.C.H. Notts, ii, 30 ; Arch. lii, 651. '^ F.C.H. Leics. 1, 180. 



'' The variants in spelling are : (i) from the coins of Tasciovanus, Verlam or Virlam, Verlamium, 

 possibly Verolamium {y.C.H. Herts. {,239-42) ; (2) Tacitus gives Verulamium (Annals, xiv, 33) ; (3) Ptolemy 

 has Urolamium ; (4) The Antonine Itinerary, Verolamium, Verolamum. 



125 



