20 THE GENUS FUMARIA IN BRITAIN 



obtains with other forms of the species. This I have distinguished 

 as a second subvariety sarniensis. It is, I beUeve, the F. iimralis 

 of Mr. Marquand's Flora of Guernsey. 



This subvariety resembles in some features another pecuHar 

 form occurring at Malvern Link, Worcestershire, whose relation- 

 ship is at present uncertain. Specimens were sent me by the 

 late Mr. Townsend as long ago as the autumn of 1904, and I have 

 subsequently seen others, collected at different dates by Mr. 

 Bickham, but, unfortunately, all of these have been gathered too 

 late to show characteristic flowers. The plant is slender, and, 

 like the Guernsey form, much branched ; but it is less erect, and 

 its racemes of subglobose fruits are denser, while the corollas are 

 small and appear quite blunt, like those of F. vmralis var. Loivei. 

 The sepals of the later flowers are large and nearly orbicular, and 

 in this, as in the dense racemes, the plant approximates to the 

 description of F. muralis var. Lehelii, of Eouy & Foucaud's Flore 

 de France. I have, however, seen no authentic examples of var. 

 Lehelii, and in view of the poor flowers of the Malvern specimens 

 that I have examined it seems undesirable at present to definitely 

 fix the position of this form, either under F. Borcei or F. muralis. 



It may be well to point out that these essentially slender 

 forms of F\ Borcei should not be confounded with starved or 

 shade-grown plants of the type, which at times may exhibit a 

 more or less similar aspect. 



The new suspecies that has been mentioned was discovered in 

 a cultivated field at Gilly Tresamble, to the north of Penryn, in 

 West Cornwall, in September, 1907, when I was visiting the 

 district for the purpose of seeing in situ the handsome new 

 species that Mr. F. H. Davey had met with three years pre- 

 viously. On reaching the locality for Mr. Davey' s plant I found 

 that it was associated, not only with F. Borcei and F. Bastardi, 

 but with another less conspicuous fumitory that had previously 

 been overlooked though obviously distinct from either of them. 

 The salient features of this plant were its short, robust habit, its 

 long racemes of small, blunt flowers, and its very obtuse, almost 

 truncate fruits ; and at first sight I suspected it might be a form 

 of F. Gussonei Boiss. A little examination, however, served to 

 show that, while the racemes and pedicels recalled F. Bastardi, 

 the foliage and the flowers rather resembled those of F. muralis, 

 and the fruits, when dried, exhibited no signs of rugosity or broad 

 apical pits. 



As the plant is evidently a member of the subsection Medice, it 

 must thus be located among the smooth-fruited species, and as it 

 differs widely from F. Beuteri and F. Munhyi in its subtruncate 

 and larger fruits, and from F. sepiicm in its robust habit, many- 

 flowered racemes, and broader sepals, the only described species 

 with which it remains to compare it are F. Borcei and F. vmralis. 

 Its resemblance to the former of these is but slight, owing to its 

 longer racemes and very much smaller and blunter flowers, with 

 shorter bracts and smaller sepals ; and to typical F. muralis it 

 presents an equally great contrast in its robust habit, long 



