Mr. Woods oti the British Species of Rosa. 199 



some of the varieties ; and perhaps p and ^, in which they are re- 

 markably elongated, ini^ht be taken for the Rosa tomentosa of 

 the Fl. Br.; and the figure in English Botaiii) is not very dif- 

 ferent from those varieties. Kay, however, says nothing of the 

 curved aculei of his R. sylvcstris fructu majore hispido, the s}- 

 nonym quoted by Sir J. E. Smith ; while, on the contrary, he 

 describes the fruit of R. sylvcstris pomifera major " fructus pyri 

 parvi forma et magnitudine" — a description which appears exag- 

 gerated if applied to R. tomentosa rt of this essay, but which agrees 

 with that variety better than with any other ; but perhaps still 

 better with an appearance sometimes met with in R. scabriiiscula. 

 Ray adds " spinulis obsiti ;" a description which altogether does 

 not agree with any fruit I have seen ; but which we may easily 

 perceive cannot indicate the same thing as the " germen glo- 

 bosum" of Linnoeus; especially if we consider that in this fa- 

 mily the fruit is uniformly rounder than the immature recep- 

 tacle. Hudson has merely joined the synonym of Ray to his 

 R. villosa (B, without adding any remark of his own to either 

 variety. Lightfoot, Fl. Scot. i. 261, has added, that the fruit is 

 black when ripe ; a circumstance which renders his species very 

 doubtful. 



In such a labyrinth what is the course to be pursued ? I have 

 already mentioned in the account of R, villosa, that in the appro- 

 priation of that name I have followed the Linna^an Herbarium. 

 R. tomentosa is therefore left for this ; and as the name cannot 

 reasonably be objected to in a genus where it is so difficult to 

 find names at all characteristic, and as some of the varieties are 

 already well known under this name, I cannot hesitate to pre- 

 serve it. The synonyms above quoted do not appear to me at 

 all doubtful as to the species ; but I have not attempted the dif- 

 ficult, or rather impracticable, task of determining the correspon- 

 dence 



