THE GENUS FUMARIA IN BRITAIN 17 



clavis on p. 180, I.e., accounting for all the forms of which ade- 

 quate material had then come under my notice. 



Since 1902 I have been able to examine additional foreign 

 material at the British Museum and at Kew, which shows that 

 abroad also this species is unstable, and besides many varying 

 British forms, a more distinct plant has come to light in West 

 Cornwall, which can only be classed as a fresh subspecies. 



The original plant of Sonder, of which I have seen numerous 

 dried specimens, either cultivated or gathered at different dates at 

 Hamburg, seems perfectly uniform in its slender habit, small, 

 apiculate flowers, and very small, smooth fruits, subrotund-ovate 

 and subacute in outline and less than 2 mm. in length. Exactly 

 similar plants occur in the Atlantic Islands and elsewhere, but at 

 Madeira, where F. miiralis is very abundant, a somewhat different 

 form is prevalent which seems to deserve varietal distinction. In 

 this the fruit is completely rounded-obtuse — indeed, almost 

 globular — though no larger than in Sonder's type ; and it appears 

 to be the F. muralis a. vulgaris of Lowe's Fl. Mader. i. p. 13, of 



which the author remarks : " Achenes perfectly globose 



or equally orbicular all round .... not apiculate." Some of 

 Lowe's specimens at South Kensington show fruit of this form, 

 while others entirely resemble the Hamburg plant, from which it 

 may be concluded that Lowe attached little importance to the 

 pointed fruit. His specimens, however, in conjunction with 

 others more recently collected and kindly lent me by Mr. Druce, 

 show that this rounded type of fruit is usually associated with a 

 slight difference in the corolla, the upper petal in such cases being 

 quite blunt (instead of apiculate) owing to its wings extending to 

 the extreme apex. At times, this form also shows much longer 

 racemes, bearing as many as sixteen flowers, and in view of these 

 differences, it may, I think, be conveniently distinguished as var. 

 Loioei, Forms intermediate in character also occur in this and 

 other habitats. It is noteworthy that, especially at Madeira, 

 the fruit of F. muralis is subject to a disease, which I have 

 observed with no other species, that causes it to swell to a 

 monstrous size. 



Of the British plants referable to F. muralis, as restricted in 

 my former paper, I have seen none agreeing with var. Loiuei. Mr. 

 Reader's Staffordshire examples are identical with the type, 

 except perhaps that the apex of the fruit is apiculate rather than 

 subacute. Specimens from Hants, Kent, and other localities, 

 seem to differ only in the size of the fruit, which, though larger, 

 is still of true muralis form. Such plants provide a connecting 

 link with the slender varieties of F, Borcei, but as in habit and 

 flower they are essentially F. muralis, I consider them to be a 

 large-fruited variety of that segregate, and am distinguishing 

 them as var. decipiens. 



The chief difficulty with these plants, however, lies in the 

 arrangement of the slender forms whose fruits are obtuse, more 

 or less obovate and larger than those of the true plant of Sonder. 



Journal of Botany, March, 1912. [Supplement.] c 



