MONTIA FONTANA L. 



3 



intermediates between geographical subspecies if they were only very local or rare; but 

 the occurrence of this type over a considerable area where, locally, the other subspecies 

 may be almost absent (as seems to be the case, for example, in the Isle of Man), suggests 

 that, whether the plant is of hybrid origin or not, it requires recognition as a widespread 

 population type of similar status to the others. It is moreover fairly clear that many 

 European records for ' M. rivularis,' based on seed type, refer to such plants with seeds 

 with small low tubercles near to ssp. fontana (cf, remarks by Hylander, 1945, 143-5), The 

 most reasonable taxonomic treatment would here seem to be to treat such types as con- 

 stituting a fourth subspecies (see below for description). It is interesting that Mansfeld 

 (1940), following Decker, gives /our species under the aggregate M. fontana in Germany, 

 which seem to correspond with the four subspecies here proposed. 



It is impossible to determine what Gmelin's M. rivularis was - the Gmelin Herbarium 

 was destroyed in Karlsruhe during the last war - and a great deal of confusion surrounds 

 the use of the name. This is partly due to the fact that there is a general habitat and 

 correlated habit difference between ssp. chondrosperma and the other subspecies, by 

 virtue of which it is often possible to distinguish roughly in the field between two types 

 of plant. For example, in areas such as the Channel Isles, N.W. France and N. Italy 

 where subsp. chondrosperma and subsp. intermedia are almost the only subspecies present, a 

 classification on habit into * M. minor ' and ' M. rivularis ' will very largely separate 

 these two subspecies. Material distributed as M. rivularis from N. Italy in Fiori, Beguinot 

 & Pampanini, Flora Italiana Exsiccata (no. 789) - which is typical subsp. intermedia - has 

 the following note on the label :- ' species rather similar to the preceding (i.e. M. minor) 

 . . . characterized by its submerged, not erect, habit, by the position of the flowering 

 branches, and by the structure of the seed surface . . . Grows only in water poor in 

 minerals . . .'. Decker (1927) makes similar remarks on his M. limosa. In areas where 

 all four subspecies may be found, as in England and parts of Germany, * M. minor ' will 

 generally be subsp. chondrosperma, whilst ' M. rivularis ' will include aquatic or sub- 

 aquatic forms of all four subspecies - though such forms of subsp. chondrosperma are 

 not common. Lastly, in areas where there is apparently only one subspecies, as in Portu- 

 gal, the attempt to apply the habit characters will give * M. minor ' and * M. rivularis ' 

 which are, as Sampaio rightly recognized, merely modifications of the one type - here 

 subsp. intermedia - to terrestrial or aquatic habitats. 



It is not possible to decide how valid is the ecological separation of the subspecies 

 (particularly of ssp. chondrosperma from the others) and to what extent the correlated 

 habit characters are purely phenotypic, without extensive study of the plants in the 

 field and in cultivation; but there is little doubt that differences in ecological preference 

 do exist, and that it is therefore possible to use a variety of habit and habitat characters 

 with some degree of success. If we consider the characters, other than those of 

 the seed, which have been used in attempts to define the different types, this will be 

 clearer. 



1 . The habitat of subsp. chondrosperma is generally described as wet, or at least season- 

 ally wet, sandy places. In such a habitat the habit of the plant is more or less tufted, 

 with short erect flowering branches. The aquatic or semi-aquatic habitats favoured by 

 the other subspecies would naturally call for a loose submerged or floating habit. 



2. Annual or perennial habit may be expected to show a similar correlation. The 

 typical subsp. chondrosperma of sandy soil is, as Necker's name M. verna suggests, a spring- 

 flowering annual, whilst aquatic plants would naturally flower later and may normally 

 remain green throughout the winter. This is presumably the basis of the colour differences 

 referred to by some authors - ' M. minor ' is described as yellow-green, whilst ' M. 

 rivularis ' is pure green, 



