THE GENUS FUMARIA IN BRITAIN 19 



that the flowers of var. verna are larger than those of the type, it 

 is evident that this type is not the plant of Jordan, and I suspect 

 that the varieties muraliforinis and verna are really both forms of 

 one plant, which is probably prevalent in the Bordeaux district 

 and was correctly placed by Haussknecht under F. muralis. It 

 is not possible from Clavaud's specimens to fix its exact position 

 with certainty, but I think it is not very different from the 

 Madeira variety Loioei. 



In these circumstances it is obvious that Clavaud's name 

 verna cannot stand for the large and dark-flowered form of 

 F. Borcei found in Guernsey, and as muraliformis, too, is at least 

 doubtful, it seems advisable to abandon both names, and also 

 serotina, about which nothing can apparently be proved. This 

 necessitates a revision of the naming of the British varieties of 

 F. Borcei. 



Of these, the Guernsey plant formerly referred to var. verna 

 is one whose constancy from seed may, I think, be doubted ; and 

 so I rank it as a form of the type, rather than a real variety, and 

 propose to distinguish it as f. ruhens. 



The plant that I have referred to Clavaud's muraliformis 

 appears, on the contrary, to be more permanently distinct. Since 

 1902 I have seen it from localities where it assumes a more robust 

 habit than I had previously observed, but its foliage, as much 

 as its irregular pedicels, seems to be constantly characteristic, 

 and I now label it var. gracilis. 



But the most widely distributed of these varieties is that 

 which I have been naming var. serotina. This occurs in many 

 localities as a well-marked and uniform plant almost equidistant 

 in characters between the types of F. Borcei and F. muralis ; and 

 were it not that it is often associated with more or less typical 

 F. Borcei, and connected by intermediates, one might well hesitate 

 where to place it. I propose to adopt for this variety the name 

 hritannica. 



The remaining variety, amhigua, of my former paper, of which 

 I have seen but few specimens from fresh habitats, I think may 

 stand unaltered. 



In addition to these variations, a few other interesting forms 

 of F. Borcei have been noticed, which perhaps should be dis- 

 tinguished. Among them is a robust but very lax form of the 

 type, with glaucous foliage and few-flowered racemes, which I 

 have collected in West Somerset. It also occurs in Ireland, and 

 probably in other districts, and I am labelling it f. elongata. 



Near the variety hritannica, but difl'ering in its longer sepals 

 and bracts and its shorter and thicker pedicels, is a further form 

 very general in North Wales. This may be regarded as a sub- 

 variety longihracteata. 



Another plant, which I first received through Mr. C. E. P. 

 Andrews in 1900, from fields at Graie, in Guernsey, and have since 

 seen from Vazon, in the same island, resembles the variety hritan- 

 nica, but has almost globose fruits and seems to habitually assume 

 a suberect, much-branched habit, very different from what usually 



r 9. 



