150 



R. H. RICHENS 



It is not supposed that the sampUng procedure utiUsed in this study is at all adequate 

 for ascertaining the precise geographical limits of the various groups recognized. There 

 is every reason to suppose that the overall pattern of distribution is extremely complex, 

 rather in the nature of a jig-saw puzzle involving a large number of distinct clones. 



The topocline concept has been applied to East Anglian elms (cf. Melville, 1949); 

 the present evidence suggests that this concept is too coarse and involves far too great a 

 simplification of the data to warrant its use. None of the five quantitative characteristics 

 studied here show any simple relationship with geographical direction. 



Historical Analysis 



It is pertinent to inquire what light the above data throw on the history of the genus 

 and on the general question of the sequence of postglacial vegetation in this country. 



Many lines of evidence suggest that the present distribution of the three major 

 species, U. glabra, U. carpinifolia and U. procera, may be of considerable antiquit^^ The 

 earliest account that differentiates satisfactorily between the British species is in the 1633 

 edition of Gerard's Herball, where their relative distribution appears to be more or less 

 the same as now. Earlier than this there is little direct evidence, but it is possible to 

 draw some inferences from the distribution of place names and surnames derived from 

 elms, and from mention of elms in early documents, in particular charters. 



In East Anglia, the surnames Elmes and Elmy occur in the seventeenth century 

 Hearth Tax returns for Suffolk and in early wills pertaining to Essex, Huntingdonshire, 

 Norfolk, Northamptonshire and Suffolk ; the Latin rendering de Ulmis also occurs. Making 

 due allowance for migration, there seems no reason to doubt that elms were sufficiently 

 widespread in the Middle Ages to afford a basis for East Anglian surnames. 



Place names take us back much further. There are three etymological roots to 

 consider, the Anglosaxon elm and wice and Celtic lem. Parish names in our area thought 

 to be based on elm include Elmdon and Elmswell ; those based on wice include Witchford 

 and Witcham (cf. Ekwall, 1947). Other parish names based on these roots occur in 

 the adjacent regions. No examples based on lem have been detected in major square 52, 

 but the River Lymn in Lincolnshire and Lympne (Rother), from which the Kentish 

 parochial names Lympne and Lyminge originate, are believed by Ekwall (1928) to be 

 derived from this root. 



It is highly probable that all these etymological roots refer to elm in general, without 

 restriction to particular species. The suggestion of Hoops (1903) that elm is U. campestris 

 L. (i.e. U. carpinifolia or U. procera) and wice, U. montana Stokes (i.e. U. glabra), appears 

 to be based on very tenuous grounds ; the argument from modern English usage is parti- 

 cularly weak since this dates back no further than the eighteenth century. In Gerard's 

 Herball, U. carpinifolia is called wych; earlier still in Turner's list of plant names (1548) 

 and in the fifteenth century Promptorium parvulorum, elm and wich appear to be used 

 indifferently for elms in general. Cognates of elm occur in High German and in the 

 Scandinavian languages. Wych corresponds to the Low German Wiecke or Wietzer, 

 also with the Slav vyaz and the Lithuanian vinksna (cf. Pokorny, 1930). A whole series 

 of etymological roots for the elm occur on the continent ; in Germany, in addition to the 

 two already cited, there is a root ip, corresponding to the Dutch iep, and a root rust, in 

 addition to the borrowed Ulme. In the Slav languages, there is also a root brest and 

 words based on the borrowed Teutonic root elm. It may well be that, at certain times 

 and in certain places, these terms have been used differentially to discriminate between 

 different types of elm, but there is little satisfactory evidence of any consistent discriminant 

 usage. In England, the distribution of place names based on elm and on wice suggests 

 different linguistic traditions in different regions rather than a reflection of regional 



