BAIIBAHEA RIVULARIS IN BRITAIN 305 



fruited form regarded by Rouy and Foueaud as the type, for which 

 they cite Billot exs. 3011 (as B. stricta). These exsiccata have the 

 siliques in some cases not more than 12 mm. long, so it will be seen 

 that in plants with siliques *' double as long " these need not be 

 longer than those of typical B. vulgaris. The British plants of 

 var. silvestris have the fruits of normal length, and Mr. Miller's. 

 plant from Cossington which Mr. Marshall has kindly sent us is in 

 no way different. We cannot find, however, that Martrin-Donos 

 states that his B. rivularis was the short-fruited form, while Carion 

 says of his B. vulgaris var. longisiUquosa that it differs yrow typ^t 

 with which it grows intermingled, by its *' siliques tres longues, tres. 

 nombreuses et rapprochees de I'axe." The specimen Mr. Marshall 

 refers to the var. longisiUquosa has siliques scarcely, if at all, longer 

 than in those which he considers may be type, and all are of the- 

 same length as in typical B. vulgaris, neither shorter nor longer. 

 We have not yet seen in this coun.try any specimens of the var, 

 silvestris with siliques of other than typical length, and it was- 

 because of this fact that the matter was not more fully detailed in 

 the paper mentioned above. We regard it as unsafe to accept without* 

 verification the accounts given in Kouy and Foueaud, although they 

 are often very valuable. 



The strict-fruited form of B. vulgaris has often been confused 

 with the true B. stricta, even by such well-known authorities as 

 Babington and Newbould, but no one who has seen true B. stricta in 

 the living state would be likely to confuse the two. Mr, Marshall,, 

 misled no doubt by the inadequate description of B. stricta given by 

 Jlouy and Foueaud, contends that our British B. stricta is only 

 B. rivularis. Unfortunately he has overlooked the important con- 

 tribution to our knowledge of B. stricta by Messrs. Sprague and 

 Hutchinson (Journ. Bot. 1908, 106), where the diagnosis of the two 

 plants are so clearly set out as to leave no doubt as to their distinction.. 

 Mr. Marshall says that the Cossington specimens show a complete- 

 agreement with examples of B. stricta in his herbarium from Clifton 

 Ings (not Thirsk) and Upton on Sevei'n, both of which had beera 

 confirmed by us and one by Murbeck. A careful comparison of these 

 specimens shows the resemblance to be merely superficial. The colour 

 and shape of the petals do not in the least suggest B. stricta : they 

 are obovate, bright yellow with a whitish claw, while those of the- 

 specimens of B. stricta have the petals much narrower in outline and 

 of a different shade of yellovv (more lemon yellow) all over. What 

 is more important still, the fiower-buds are distinctly hairy in tlie 

 same specimens, while those of the Cossington plant are ^j^i^/^'e glabrous.. 

 We have measured the length of the styles in ten fruits of each of 

 the gatherings in question, and the averages are : — Cossington 2"4 mm.,. 

 Upton 1-8 mm., Clifton Ings I'G mm. Moreover, those of the first 

 taper and are less than '3 mm. broad, whije those of the B. stricta 

 are stout and truncate, '4 and -5 mm. broad respectively. Thes(^' 

 measurements correspond to a \ery real difference in appearance. 



We consider the question of the shape and size of the lateral 

 lobes of the leaves to be of subsidiary importance and unreliable for 



