Page:English Historical Review Volume 35.djvu/387

 1920 ITINERARY OF ANTONINUS 379 letters of Caesaromagus. There are considerable variations of spelling between the Tabula and the Itinerarium ; and it is only for convenience' sake that I adopt the spelling of the latter in the following table : Roman name. Mileage. Modern name. Mileage. Itin. Tab. Venta Icinorum Sitomago Combretonio ad Ansam Camuloduno Canonio Caesaromago Durolito Londinio xxxii xxii XV vi viiii xii xvi XV xxii XV XV V viii xii missing missing Caister St. Edmund's • River Gipping J Stratford St. Mary- Colchester Kelvedon Chelmsford Romford London 33 15 6 1^ 17 13 Throughout Britain the difference between the Roman miles of the Itinerarium and our modern English miles is imperceptible. Durolitum must be placed a little to the north-east of Romford, but elsewhere the modern names given above seem to correspond with the sites of the places mentioned in the Iter. The difficulty lies in the two stages between Venta and Combretonium. Instead of 33 miles, the Itinerarium gives 54 miles and the Tabula 37. We have at p. 474, in the first part of the fifth Iter, a different version of the same road, as follows : Londinium, xxviii (cf . xxxi, the aggregate of xv and xvi above) Caesaromagus, xxiiii (cf . xxi, the aggregate of xii and viiii above) Colonia, xxxv Villa Faustini, xviiilcinos. Icinos means 'Icenians', and. therefore, refers to the same place as Venta Icinorum ; since there is abundant evidence in the Itinerarium of the practice of using the gentilic name in the plural form either instead of, or as an addition to, the old place- name of the capital of a canton. The Itinerarium uses freely either Augusta Suessonum, or Suessonas ; Ambianis, or Samara- brivas ; Luticia, or Luticia Parisiorum ; Tricasis, or Augustobona ; Bagacum, or Bagacum Nerviorum ; Durocortorum, or Durocortorum Remorum. In the fifth Iter, the compiler is abbreviating by the omission of intermediate stages and the substitution of the aggre- gate of two stages for the number of the second stage. In this process, his number for the double stage Caesaromagus — Colonia becomes three in excess of the numbers given in the ninth Iter, and his number for the double stage Londinium — Caesaromagus becomes three less than the numbers of the ninth Iter. The ninth Iter, however, is partially confirmed by the Tabula and, consequently, it is the numbers of the fifth Iter which are corrupt. Throughout the Itinerarium no confusion of numbers is so common as the confusion of u and ii. Here the compiler's error was to read