1054 



is not easy to determine with certainty whether the old flowering 

 stem and the new leafy shoot proceed from the same root. Add to 

 this, that where the root spends itself in making a great number of 

 flowering branches, it is very apt to neglect forming new shoots; and 

 vice versa, when the new shoots are strong and numerous, the flower- 

 ing branches are apt to be deficient, so that it is no uncommon thing 

 to find the shoots of one species and the flowering branches of another, 

 almost exclusively in the same bush. 



My first station was at Sketty in the neighbourhood of Swansea. 

 The species most abundant there are R. discolor, R. cordifolius, and 

 R. Koehleri ; the scarce one R. macrophyllus. These are all tolerably 

 distinct in their general appearance. R. macrophyllus is rather more 

 prickly than it usually is in Sussex; but even with us the character is 

 variable. R. cordifolius is almost exclusively the variety with nar- 

 rower leafits, and a stem more or less hairy, i. e., it is the variety consi- 

 dered by Dr. Bell Salter to be at once the R. sylvaticus and the R. 

 villicaulis of the ' Rubi Germanici.' In the 'Tourist's Flora,' I have 

 followed Babington, in making R. cordifolius the typical form. I 

 am now almost inclined to repent of this, since, though R. cordifolius 

 be the common variety in the south and east of England, the var. R. 

 rhamnifolius seems to be the most widely diffused. The leaves are 

 elliptic or somewhat rhombic, or perhaps obovate and acuminate, with 

 or without a slight indentation at the base, but not truly cordate. 



On the borders of the Crwmlyn Bog, I observed a still more woolly 

 plant, which I at first thought might belong to R. cordifolius, but the 

 leaves have underneath something of the glittering appearance of R. 

 leucostachys, with which I am now disposed to join it. It has the 

 mixture of a few smaller prickles. R. radula occurs in the same 

 neighbourhood. 



At Horton, on the borders «of Gower, and, I believe, on the lime- 

 stone, I noticed what is perhaps R. fuscus (according to Dr. Bell 

 Salter, a variety of R. Koehleri), but which, but for the presence of 

 setae, might pass for a variety of R. corylifolius. R. caesius, is abund- 

 ant on the limestone of Gower. 



My next station was Pont-nedd-vechan. R. discolor and R. 

 Koehleri are here abundant, and easily distinguishable. The former, 

 however, does not ascend into the ravines, which form the character 

 of this part of the country. Dr. Bell Salter, in his notes on R. argen- 

 teus, Bot. Gaz., ii. p. 151, says, that the hairs have the " opaque, lus- 

 trous tint common to the tomentosi." I do not understand this, since 

 if I am to consider the word opaque as opposed to translucent, I am 



