4 S. M| WALl'ERS 



3. The difference between ' terminal ' and ' lateral ' inflorescences is often men- 

 tioned. The short, erect branches of typical ssp. chondrosperma terminate in c^tocs of 

 flowers. The so-called ' lateral ' cymes of ' M. rivularis ' arise by the continued vegetative 

 growth of a lateral bud below the inflorescence ; this is clearly a reflection of the general 

 vegetative vigour of an aquatic or semi-aquatic plant, and is in no way a fundamental j 

 difference. \ 



4. Connate leaf-bases, it is sometimes claimed (? following Gmelin's original des- < 

 cription of M. minor), distinguish subsp. chondrosperma. Again this character seems to be | 

 merely an expression of difference in vegetative vigour - the comparatively short, slender '. 

 stems normally show the opposite leaf-bases in contact, as opposed to the separated 

 leaf-bases of the more vigorous aquatic shoots of ' M. rivularis.' Quite apart from the 

 difficulty of using such a character, it is very doubtful whether it has any basis other than 

 the obvious modificatory one. 



Given this different habitat preference, these vegetative characters, even if shown 

 to be purely phenotypic, would normally serve to distinguish subsp. chondrosperma from i 

 the other subspecies. . Yet herbarium identification relying on habit characters alone ! 

 may be completely unreliable - aquatic forms of subsp. chondrosperma, for example, 

 undoubtedly occur, in Britain and elsewhere. 



The comparative descriptions of the subspecies which follow are based on personal j 

 observations, and on the descriptions of the seeds given by Lindberg (1901) and Beeby 

 (1909). The details of the cellular structure of the seed-coat, and the shape, size and 

 distribution of the tubercles on it, can be seen either directly under the binocular micro- i 

 scope or, as Lindberg stated, by removing a portion of the coat (easily detachable in 

 subsp. fontana, less easily in the others) and studying under low power of the microscope. 

 The vice-comital distribution is given for Britain and is based only on material personally 

 examined. It is clearly very far from complete, but is useful as indicating the difference 

 in British distribution. Records from Ireland are as yet too inadequate to justify ; 

 publication, and Irish material needs much further study. 



(1) subsp. FONTANA 



M. fontana L. (1753), Sp. PI, 87, sensu stricto. 



M. lamprosperma Chamisso (1831), Linnaea, 6, 564. 



M. rivularis auct., ? C. C. Gmehn. j 



Ripe seed 1-1 to 1-35 mm., smooth and shining seen under lens; under higher ' 

 magnification, individual cells of seed-coat easily seen as a reticulate pattern, rather 

 elongated, irregular in size and shape, with slightly convex outer walls; those in rows 

 by keel elongated along row; seed-coat thin, brittle, often, in dried material, broken j 

 open. Plant variable in habit, usually looser than subsp. chondrosperma, often submerged 

 (var. horeo-rivularis Lindb. f.), green or yellow-green. Habitat : trickles of water or 

 very wet places, on acid soil or rock only. 



Distribution : Circumpolar Arctic and N. Temperate; also in the S. Hemisphere 

 into the Antarctic. Common in Scotland and N. England. V.c.'s : 45, 46, 49, 50, 58, 

 61, 64, 65, 68, 69, 70, 72, 73, 77, 78, 80, 83, 85-90, 92, 93, 96-98, 100, 104, 105, 107-112. 



(2) subsp. chondrosperma (Fenzl) comb. nov. j 



M. fontana var. chondrosperma Fenzl (1843), in Ledebour, Fl. Ross., 2, 152. \ 

 M. verna Necker (1768), Delic. Fl. Gallo-Belg., 1, 78, nomen illeg., quoad j 

 descr. 



M. minor C. C. Gmehn (1805), Fl. Bad., 1, 301, nomen illeg., quoad descr. 

 Ripe seed 1-0 to 1-2 mm., dull and entirely covered with rather coarse tubercles as 



