THE GENUS FUMARIA IN BRITAIN 43 



in the other subsections of the genus, coarsely rugose when dry 

 and often strongly keeled. 



The species of this subsection are generally well marked by 

 the form of the lower petal and the large rugose fruits. 



In F. agraria Lag., F. major Badarro, F. atlantica Coss. et 

 Dur., and my F. occiclentalis the largest and handsomest flowers 

 of the whole genus, sometimes attaining a length of 15 mm., are to 

 be found. A notable feature of the corolla in this subsection is the 

 gradation that occurs in the development of the purple colouring 

 of the tip of the petals. In F. major this dark colour marks the 

 inner petals and the entire wings of the upper one ; in F. occiclen- 

 talis these wings are purple with white margins ; in F. agraria, F. 

 atlantica, and F. amarysia Boiss. et Held, the dark tips are con- 

 fined to the inner petals ; and, finally, in F. macrocarpa Pari, the 

 purple colour disappears entirely. 



F. amarysia, as represented by the specimen " Heldreich, Herb. 

 Grsec. Norm. 817. Attica, prope Hagios Sabbas, Apl. 1884," shows 

 a resemblance to F. Bastardi Bor. of the previous subsection, and 

 seems intermediate between that species and F. judaica Boiss. ; 

 while Heldreich's plant "F. amarysia Boiss. forma scandens, In 

 olivetis prope Podoniphti, 1889," of which there are good examples 

 both at Kew and at South Kensington, is apparently, judging from 

 its small, pointed fruits, not a variety of that species at all, but 

 rather a form of F. Bastardi and very near to Jordan's F. vagans. 



F. macrocarpa is remarkable not only for its pale flowers but 

 for the size of its fruits, w^hich are the largest in the genus, being 

 fully 3| mm. in diameter and almost perfectly globular. This size 

 is approached also by the fruits of F. atlantica. 



[F. agraria Lagasca. This species was collected in 1903 by 

 Mr. Druce at Iver, Bucks, where it is supposed to have been 

 introduced with London street refuse. As it has presumably not 

 established itself there, and has not occurred regularly to my 

 knowledge in any other British station, it can hardly be given a 

 place among the British Fumitories.] 



6. FuMARIA OCCIDENTALIS Pugsley. 



In the case of this species, which I discovered at Penzance in 

 1902 and described as a new plant two years later, there is but 

 little to add to my paper in Journ. Bot. vol. xlii. pp. 217 sqq. 



As was implied in that article by the comparison with F. 

 flabellata Gasp., F. occidentalis shows some resemblance to the 

 Caineolatce, in which respect it is unique in its subsection. The 

 likeness is seen in the relatively long peduncles, the frequently 

 arcuate fruiting pedicels and the large sepals ; but it is altogether 

 outweighed by the form of the corolla and the fruit, which are 

 both of markedly agraria type. 



The figure accompanying my paper of 1904, which was drawn 

 from one of my original Penzance specimens after pressing, de- 

 picts the fruit a little broader than in life, owing to an insuflicient 

 allowance having been made for shrinkage of the fleshy neck in 

 the process of drying. The fruit is indeed similar in form to that 



