767 



leaves, when any, minute, and little apparent amongst the flowers. 

 The latter are, in the American P. scandens, somewhat superior, per- 

 haps, in size to those of P. dumetorum, and the wings of the enlarged 

 perianth project a little more beyond the apex of the seed, and are 

 more rounded at their anterior margins, which gives the appearance 

 of a deep notch or sinus, causing the fructiferous perianth to assume 

 more the form and aspect of the samara of an elm than in the Euro- 

 pean species. In the size, shape, and smoothness of the finely po- 

 lished fruit, I perceive no difference betwixt our own and the transat- 

 lantic plant. The American P. scandens I have always found growing 

 exclusively in wet, or at least very moist thickets, most commonly on 

 willows or other low bushes in swamps or osier beds, whilst our P. 

 dumetorum affects dry as well as damp spots, and, as we have just 

 seen, will even flourish in sandy hedgerows. 



I suspect P. dumetorum and the American form or species P. scan- 

 dens to be occasionally biennial. Wahlenberg says (Fl. Ups. p. 132) 

 "Radix subbiennis," and in Fl. Suecica, i. p. 251, 2nd ed., "Radix 

 circiter biennis et tota planta magis longaeva et extensa." The ap- 

 pearance of these plants favours such an idea, but P. scandens from 

 seed I collected at Philadelphia has shown itself an annnal hitherto 

 in St. John's garden, near Ryde, where in dry soil it has preserved all 

 its original character through two generations, evincing no disposition 

 to become P. dumetorum, which last from Hampshire seeds I shall 

 hope to have under my eye next year, to compare with the American 

 variety, if such it should prove really to be. Mr. Borrer tells me he 

 has a plant labelled P. scandens from the Cambridge Botanic Garden 

 which has proved more than annual. It is worthy of remark that al- 

 though the larger plant of the two, P. dumetorum has smaller flowers 

 at first than P. Convolvulus, it being only in fructification that they 

 so notably surpass those of the latter in size. The seeds of P. dume- 

 torum and scandens are smaller in all their stages than in P. Convol- 

 vulus. Wahlenberg, in his ' Flora of Upsal,' inclined to believe P. 

 dumetorum a wood variety of P. Convolvulus,* but in his later work, 

 the 'Flora Suecica,' he has tacitly renounced an opinion so manifestly 

 erroneous. 



Obs. — Polygonum Fagopyrum (Buckwheat) is excluded from the 

 floral census of the county, being only found as an occasional and 

 very fugacious intruder on waste ground, either strayed from cultiva- 



* Nihilo tamen minus vix nisi forma nemorosa precedents (P. Convolvuli) est. 

 Fl. Upsal. p. 132. 



