198 SALIX FRAGILIS, S. RUSSELUANA, AND S. VIRIDIS. 



nectariuin bis terve superante ;" and for viridis Fr., under which he 

 quotes as a synonym S. Busselliana Sm., " capsulis breve conicis 

 obtusiusculis glaberrimis, pedicellatis, pedicello nectarium subsupe- 

 rante," a statement which he afterwards modifies as " nectarium 



Wimmer, for fragiUs, " gei 

 jonico-subulata," and for fi 



mina pedicello brevissimo conico-cylindracea." British authors 

 have followed Smith more or less closely, the exception being 

 Bos well Syme, who says of fragiUs, M capsule conical-subulate, 

 glabrous, on a stalk twice or thrice as long as the nectary ;" and 

 for Busselliana (as a synonym of viridis Fr.), " capsule conical-subu- 

 late, glabrous, on a stalk slightly longer than the nectary." In 



except that it is pedicelled. 



ailed f\ 



by Koch, Grenier, Andersson, Wimmer, and Boswell Syme is 

 almost certainly identical with Smith's Busselliana, and distinctly 

 different from his fragiUs. Which of the two plants is Linne's 

 fragiUs must remain a little uncertain. For my part I retain the 

 name fro g His L. for the plant of the Continental botanists. That 

 these writers failed to notice that their fragiUs and Smith's were not 

 the same can, I assume, only be accounted for thus, that, trusting 

 in Smith's reputation and in his possession of the Linnsean Her- 

 barium, they neglected to compare his description and drawings (or 

 perhaps had not the opportunity), and taking for granted that his 

 and their fray U is was the same, attempted to discover some difference 

 between plants which are identical — their fragiUs and Smith's 

 Busselliana. 



What then is Smith 's fragiUs t I think there can be no doubt 

 but that it is an hybrid between S. fragiUs and S. alba ; in other 

 words, one of the series of which S. viridis Fr. is the central form. 



Putting aside for the moment other points of distinction, the 

 essential characteristics of what, following Andersson, &c, is the 

 female of the real S. fragiUs L., lie in the long, ovate, subulate cap- 

 sule, gradually tapering into the style, with a pedicel 2 to 3 times 

 as long as the nectary. Alba has a sessile or almost sessile capsule, 

 ovate-conic in shape, and distinctly obtuse. Viridis, like other 

 hybrids, varies according as it approaches or recedes from one 

 or other of its parents, and has a capsule, sometimes more and some- 

 times less, but always more obtuse than in fragiUs, i.e. not tapering 

 so gradually into the style, and with a pedicel sometimes scarcely 

 as long as the nectary, sometimes twice (or even a little more) as 

 long. Moreover, though I think none of the descriptions point 

 this out, the length of the capsule in viridis is intermediate 

 between fragile and alba. The scale of size in Andersson's plate 

 shows this, and my measurements corroborate it. In alba the 



nnnonln \a oK/-kTif <f*is\«-v* C\\ r* i ^ * 4%-» f-f/titil to oKrmfc 7 111111. I 3*11^* 



in viridis, about 5 to 6 mm. 



/ 





* Wimmer'a " brevis " will probably have a wrong value placed on it unless 

 his si)eehnens are examined.— F. B. W. 



