Retreat of the Late Pliocene and Lower Pleistocene Crag sea Late in Norwich Crag times, the sea retreated to north Norfolk (Fig. 2b), depositing the Weybourne Crag. This situation continued with deposition of the Cromer Forest-bed. The cause of the retreat was tectonic, involving the relative uplift of western Britain. It was this uplift that initiated the early Thames, Bytham and Ancaster Rivers as important drainage lines with catchments reaching Wales and the Pennines. The history of the Thames from this time on is much better documented, so is that of the Norwich Crag and Cromer Forest-bed Formations. Unfortunately correlating the two has been fraught with difficulties and has not been achieved satisfactorily The fluvial 'Sudbury Formation' corresponds to the mainly marine deposits of the Weybourne Crag and earlier part of the Cromer Forest-bed Formation, but so far detailed correlation has not proved possible. The 'Colchester Formation' contains interglacial deposits of three different warm stages. However, all are designated Cromerian, which until recently in north Norfolk was attributed to single warm stage (West, 1980). Recent reassessment of the molluscan and vertebrate evidence (Preece & Parfitt 2000) now suggests that the Cromerian deposits of Norfolk and elsewhere may belong to five different warm stages (comparable to the case in the Netherlands). As yet this work has not been correlated with the 'Colchester Formation' interglacial deposits, but it does imply a way forward. The fundamental driving force for the formation of the Thames terraces is now recognised to be the continued relative uplift of western Britain, with the triggers for downcutting and terrace aggradation being climatically controlled (Bridgland 1994, 2000; Bridgland & Allen 1996; Maddy 1997). As yet the 'Colchester Formation' has not been linked in with the oxygen isotope climatic stages, as has been achieved for the Lower Thames terraces (Bridgland 1994, 1995; reviewed by Allen 1998), but it may be possible as there are five climatic cycles (OIS 13-22) (Funnell 1995) and four recognised terraces. However, the 'Sudbury Formation' has six recognised terraces, covering approximately 40 climatic cycles (OIS 23-64) (Funnell 1995). The nature of the earlier cycles was very different, with very much milder cold stages (Funnel 1995; see Fig. 1 in Allen 1999). The strength of the climatic oscillations may not have been sufficient to trigger terrace formation on a regular basis. It is also significant that none of the 'Sudbury Formation' terraces contain interglacial deposits. Acknowledgements I am very pleased to acknowledge the help of Jim Rose, David Bridgland, Colin Whiteman, John Sinclair, Richard Hey and Simon Lewis. This paper and its two predecessors in the Naturalist draw heavily on their research. Their friendship and assistance throughout my own research has been invaluable. This paper was reviewed by CW, JR and JS, whose comments greatly improved it. The views expressed here may not coincide with theirs, for which I take responsibility References ALLFN, P. (1998) The geological history of the Lower Thames in F.ssex -a review. Essex Naturalist 15 (New series): 5-22. ALLEN, P. (1999) The Anglian cold stage in Essex - a review. Essex Naturalist 16 (New series): 83-100. ARKELL,W.J. (1947a) The Geology of Oxford. Clarendon Press, Oxford. 269pp. ARKELL, W.J. (1947b) The geology of the Evenlode Gorge, Oxfordshire. Proceedings of the Geologists' Association 58: 87-114. DOULTON, G.S. (1992) Quaternary. In: Duff, P.McL.D & Smith, A.J. (eds), Geology ofEnglandand Wales. The Geological Society, London. Pp. 413-444. Essex Naturalist (New Series) 18 (2001) 25