STUDIES ON ULMUS : I 



151 



diflferentiation in the elms themselves. Thus, parish names based on elm are conspicuous 

 in Kent, Essex and Suffolk, but no names based on wice are reported (cf. Ekwall, 1947); 

 in Cambridgeshire, Norfolk and Northamptonshire, the reverse situation occurs. It will 

 therefore be assumed that place names involving (1) elm or wice, or (2) the Celtic lem 

 indicate the presence of elms in post-Roman and pre-Roman times, respectively, without 

 any restriction as to species. ^ 



The next step in our argument, namely, the postulation that the species or clones, 

 frequent in the ancient parochial hedges in localities whose names indicate the presence 

 of elms in earlier times, are the descendants, sexual or vegetative, of the earlier elms, is 

 obviously more hazardous. The alternative, however, would require the extinction of 

 one type of elm and its replacement by another, which is obviously a more elaborate 

 and less likely hypothesis. 



Other lines of evidence, too, provide grounds for supposing that the present elms 

 have persisted in situ, at least since Roman times. The elm played an important role 

 in Roman agriculture as a forage plant, in addition to its use as a timber and shade tree 

 and as a support for vines. Cato gives instructions for planting elms around fields and 

 along roads; the somewhat discrepant accounts of Pliny and Columella indicate clearly 

 that the Romans distinguished between different types of elm in respect of fodder value 

 and introduced clonal material from one region into another. The value of the elm as 

 a source of fodder is reiterated in the Middle Ages by de Crescentiis, and the fact that it 

 was so employed in England is stated in the writings of such early English agriculturalists 

 as Fitzherbarde (1534) and Markham (1631). It is quite feasible that a number of Roman 

 farming practices were perpetuated after the Anglosaxon invasions. Gray (1915) has 

 shown, with reference to field systems, that this may be particularly the case in Kent 

 and East Anglia. It seems reasonable, therefore, to suppose that elms were extensively 

 distributed in eastern England in Roman times and have persisted since then without 

 major distributional change. 



There is no reason to suppose, however, that the Romans are responsible for the 

 present pattern of distribution. Elm has been a conspicuous component in the British 

 flora since mesolithic times (cf. Godwin, 1940). The earliest type was presumably 

 U. glabra. This is not a frequent species in old hedges in East Anglia, and the present 

 distribution suggests that it may be a relict species. 



U. carpinifolia is abundant in East Anglia and occurs also in East Kent ; it is probably 

 the species that gave rise to the Celtic place names based on lem in Kent and also to the 

 frequent East Anglian place names based on the Anglosaxon elm or wice. The fact that 

 this species is seldom found in woods makes it rather doubtful whether it occurred in 

 England before the opening up of the country by neolithic agriculturists, unless, possibly, 

 it was present in some open communities no longer extant. The failure of the species 

 to set seed except in exceptionally favourable seasons suggests that it has been introduced, 

 directly or indirectly, from warmer regions to the south. 



The present evidence, therefore, seems to suggest that U. carpinifolia was introduced 

 by, or in the wake of, settlers arriving in England from the east and settling both in Kent 

 and East Anglia, some time in the neolithic. Bronze or Iron ages. Intensive study of the 

 distribution of the English clones composing U. carpinifolia and a comparison between 

 these and those of the European seaboard might well throw further light on the source 

 and date of this introduction. 



U. procera presents a much more difficult problem. It only occurs along the western 

 edge of our area, and, although it is generally recognised as a Midland species, no detailed 

 knowledge has been obtained yet as to its distribution in the Midland counties. It is 

 known to extend eastward along both sides of the Thames estuary, severing the East 



