32 THE GENUS FUMARIA IN BRITAIN 



the Cornish plant, as I saw it in 1907, was almost equally good, 

 although in 1904 and 1905 the largest corollas did not exceed 

 12| mm. Seiior Vicioso's specimens are likewise variable in the 

 size of their flowers, which, on an average, are smaller than in 

 the British examples and rarely exceed 12 mm. in length. His 

 plants were collected early in the season, and are mostly quite 

 small — apparently the product of open fallows on poor, dry soil — 

 and the flowers, though uniformly well developed, are only about 

 11 mm. long in the weaker examples, and when, as in some cases, 

 the characteristic recurving of the pedicels is lacking, they bear a 

 great resemblance to F. Gussonei. The fruits of all these Calatayud 

 plants have unfortunately been crushed in the press, and con- 

 sequently they cannot be accurately diagnosed with certainty ; but 

 the remains show that they are subacute, of medium size and with 

 little rugosity, and hence practically identical with those of the 

 Navalmoral specimen. These latter are uniform in size and shape, 

 but while the majority appear quite smooth, the most mature are 

 clearly rugulose. Similar fruit-characters are also obvious in the 

 Cornish specimens, but in Britain the fruits, though never clearly 

 rugulose, evidently vary in size and form, for while in Bourgeau's 

 plant and Mr. Davey's first example they are scarcely separable, 

 in the plants that I collected in 1907 they are appreciably larger 

 and less acute. 



It will be seen that the name under which Bouigeau's plant 

 was sent out — F. Bastardi Bor. — is untenable, inasmuch as this, 

 in the type, is a species with rugose fruit, which, as will be shown, 

 is synonymous with F. confusa Jord., while its variety ^ major 

 is demonstrably F. Borcei. Haussknecht, judging from the spe- 

 cimen in the Vienna Herbarium, and seemingly guided by the 

 form of the fruit, refers the plant to F. muralis, but the floral 

 characters of the two are widely different in several respects, 

 and I think they cannot possibly be united as one species. In 

 Willkomm and Lange's Flora Hispanica, iii. p. 881, all the 

 Spanish forms of this subsection are placed under F. media Lois., 

 which stands as a. Gussonei, b. affinis {= F. affinis Hamm. = F. 

 vagans Jord. ex spec, authen.), c. Borcei, d. apiculata, and e. mti- 

 ralis ; and the Navalmoral plant, though not placed under any of 

 these varieties, is mentioned as having been wrongly referred by 

 Haussknecht to F. muralis. The variety to which it seems most 

 nearly related, and to which the Calatayud plant has been referred 

 is b. affinis, an identification which, however, appears inadmis- 

 sible, inasmuch as Hammar describes and figures his F. affinis 

 (of which I have seen no authentic specimen) as having tuber- 

 culate-rugose fruit and serrately toothed sepals. F. affinis is 

 regarded by Haussknecht as a large-flowered form of F. Gussonei. 

 It is equally certain, notwithstanding the terms of Jordan's 

 description, that Bourgeau's plant cannot be named F. vagans, 

 for of this the British Museum possesses two specimens named 

 by Jordan in fair preservation which are quite different, and not 

 easily separable from that author's F. confusa. 



From the other named forms of the subsection medice the plant 



