1055 



quite at a loss for his character of the iomentosi, for the hairs amongst 

 them seem to be individually just as translucent as those of the Cory- 

 lifolii. The passage above cited forbids us to oppose it to shining, 

 and besides, R. leucostachys, than which no species has more shining 

 hairs, is placed among the opaque-haired species. While on this 

 head, let me add that Dr. Bell Salter puzzles me, by describing the 

 hairs of R. argenteus, as " divaricating and at the same time somewhat 

 patent," and in speaking (p. 147) of the figure of R. macrophyllus, in 

 the ' Rubi Germanici, he says, the hairs are drawn " not truly patent, 

 but patent divaricating." I have always understood divaricate to be 

 patent in a very high degree. With respect to R. discolor, it appears 

 to me that the closeness of the pubescence does not depend on its 

 being adpressed, but on its being stellate and very short. Another 

 incorrectness, and in which I confess I have been a participator, is in 

 describing the margin of the leafits as bent down. The curve is lon- 

 gitudinal, slightly modified by the somewhat wavy nature of the mar- 

 gin. Let me add, in making these remarks, that I am very sensible 

 of the great merit of this paper, which I think is a considerable step 

 forward in our knowledge of this intricate genus. 



R. rhamnifolius is also common about Pont-nedd-vechan, and as be- 

 fore, usually hairy, so that I have sometimes found it difficult to draw 

 a line between it and R. leucostachys. When the hairs are very nu- 

 merous it is difficult to determine whether the stem beneath is polished 

 or not. Sometimes, however, its red polished stems are very conspi- 

 cuous. At other times it shows a tendency to produce over-lappin g 

 leafits. The prickles on the old stem are sometimes few and small, 

 while those on the young shoots are large and numerous, showing 

 that in different years, or at different seasons, a different degree of 

 prickliness is produced. 



Dr. Bell Salter joins R. villicaulis,/3. of Babington's [first] edition, to 

 leucostachys, and Babington in his later edition has joined R. villi- 

 caulis altogether to R. leucostachys. I am not inclined to dispute this 

 decision, but I observe here two forms which apparently belong to this 

 species, the first has a lengthened panicle, generally furnished with one 

 or two simple cordate leaves, lilac flowers, and a stem with few or none 

 of the smaller scattered prickles. The other, with a shorter, leafless 

 panicle, white flowers, and a much greater proportion of smaller scat- 

 tered prickles. The young shoots also, in the specimens I have 

 brought home, exhibit few setae. The latter form is very common 

 among the hills of Glamorgan and Monmouthshire, the first is compa- 

 ratively scarce. 



