56 THE GKNUS FUMARIA IN BRITAIN 



graph, and has been followed by Hammar, Syme, Boissier, Eouy 

 & Foucaud and Nicotra, all of whom adopt Lagasca's name for 

 the species under discussion. 



In view of the wideness of its distribution, extending, as it 

 does, from Spain to Persia, F. micrantha seems to exhibit com- 

 paratively little range of variation. The first variety to be 

 established was ^ Parlatoriana Boissier, published in Flora 

 Orientalis, i. p. 137 (1867), and founded on the F. Parlatoriana of 

 Kralik, which was discovered at Alexandria. Its diagnosis runs : 

 " Laciniae tenuiores setaceae, sepala subangustiora. Flores parvi 

 rosgei " ; and it is maintained as a variety in Haussknecht's 

 Monograph, where additional characters are given. The small 

 flowers are there stated to exceed but slightly 3 mm. in length, 

 with sepals little more than 1 mm. long and ^ mm. broad ; and 

 the pedicels are said to be shorter and thicker than in the type, 

 and the fruits more clearly keeled. Oran, in Algeria, is quoted as 

 a second habitat. 



In Eouy & Foucaud's Flore de France this variety is inserted 

 as an inhabitant of Marseilles and Perpignan, but the only essential 

 varietal character mentioned is the smallness of the sepals, and it 

 may be doubted whether the plant intended is identical with the 

 African form. 



There is good Egyptian material of this variety at Kew. 



I possess specimens from Oran, sent under the name of 

 F. densiflora DC. /S bracteosa Batt. {F. bracteosa Pomel, Nouv. 

 Mat. Fl. Atlant. 239), which are not readily separable from the 

 variety Parlatoriana. 



The only remaining variety of F. micrantha that I can trace 

 is /3 littoralis Eouy & Foucaud, which was described as a distinct 

 species, F. littoralis, by Du Mortier in Bull. Bot. Soc. Eoy. Belg. 

 vii. p. 359 (1868), from specimens found at Ostend and elsewhere 

 on the Belgian coast. Of this plant Du Mortier says : — " Ses 

 tiges . . . sont flasques, decombantes et tres-rameuses ; ses feuilles 

 . . . sont glauques et a pinnules aplaties ; son epi est comme cache 

 dans les feuilles ; ses sepales fortement dentes . . . ne depassent 

 pas la moitie du tube de la corolle (corolla angustioribus) ; enfin 

 ses fleurs blanchatres sont en epi serre compose d'un petit nombre 

 de fleurs . . ." No authentic examples of this plant have come 

 under my notice, but judging from its reputed characters it seems 

 at most only varietally distinct from ordinary F. micrantha. 

 Indeed, the lax habit and pale flowers in poor racemes may be 

 simply due to growth in very light soil, and the flattened leaf- 

 segments, which I have noticed in seaside Scotch examples of 

 F. micrantha, can well be attributed to maritime influence, and 

 regarded as analogous with the thick leaves often found on such 

 plants as Lotus corniculatus in similar situations. 



As this variety is recorded not only from Belgium but from 

 Dunkirk and other places on the adjacent French coast it may be 

 expected to occur in the south-east districts of Great Britain. 



Among the British material of F. micrantha that I have 

 examined no specimen of either of the above-mentioned varieties 



