Fundamental Units of Vegetation. 43 
appear to be necessary. It seems sufficient to denote the formation 
by a single word, e.g., Oxodion, whose termination would mean 
formation and whose root, and, if necessary, also a prefix, would indi¬ 
cate the habitat. The advantages of such a plan are that the technical 
term used to denote the formation would remain brief or compara¬ 
tively so, that it would emphasize the habitat, and that it could be 
used in connection with an associational name framed on the plan of 
Cajander. Thus, an association of Eriophorum vagiuatuiu belonging 
to a formation characterized by a sour peaty soil could be indicated 
by the term Oxodion Eriophoreti-vaginati. 
The question of universal names for vegetation units is bound 
up with that of the universality or otherwise of particular formations. 
Of course, whenever climatic differences are sufficiently strong to 
be related to a radically different vegetation, then it is obvious that 
a different formation has been produced ; for climatic factors no 
less than edaphic factors assist to form the habitat. Similarly, 
differences of flora may he related to differences in geographical 
position ; and the latter too must be regarded as an element, or 
rather as a complex of elements, of the habitat. Whilst the 
geographical position of the Pennines is sufficiently different from 
that of Connemara in the west of Ireland to bring about certain 
differences in floristic composition, these differences are scarcely 
sufficient to justify the placing of the peat moors of the Pennines 
in a different formation from the peat moors of Connemara. 
Following Cajander (1903: 24) and Warming (1909: 145, 
et. seq.), differences of this character may he said to bring about 
varieties of associations, which are the “facies” of some authors. 
With regard to peat moors of America and those of Europe, 
however, the geographical position is doubtless sufficiently different 
in the two cases to justify a formational separation ; and this would 
seem to be demanded whether the formations be determined by 
habitat or by floristic composition. Such formations should certainly 
be correlated into groups of allied formations ; and the resemblances 
of the edaphic factors of the habitats in question would necessitate 
such a grouping in any cosmopolitan treatment of formations. 
In a system of universal nomenclature, such affinity could be 
indicated by some prefix. For example, an Oxodion in one climatic 
or geographical region might perhaps be termed an a-Oxodion, and 
another Oxodion in a different region a /3-Oxodion, and so on. 
Warming, however, apparently places in the same “formation ” 
many communities which differ widely in geographical position and 
