2 ON THE "species QUESTION AS KEGARDS RUBUS. 



we talk of a bramble species, we merely mean a form or connected 

 series of forms, isolated in itself or in their aggregate at the present 

 day, and through the area of our personal observation. This is our 

 test. Take, for example, R. macrophyllits, Weihe, a bundle of forms 

 hugely different inter se at their extremes, still graduated into each 

 other, or as we conceive graduated, by a series of transitional links so 

 finely, that the very double of R. thyrsoideus, Wimm., finds itself in 

 one box with the near neighbour of R. rhamnifoUus, W. and N. An 

 arrangement which again ties up typical macrophyllas, W. and N., with 

 extreme Westmoreland R. umbrosus, Arn., which seems miles away." 

 Very well, let us accept species in Rnbus to mean forms between which 

 transitional links are not found. We have drawn in our horns, re- 

 member, in saying " are not found." We have said nothing about 

 links having existed in all time, and ceased to do so, links that have 

 altered through soil, climate, and the infinite factors of vegetable 

 variance. We pass in silence the fact how imperfectly known and in- 

 adequately described are still many native forms, even in this little in- 

 sular corner of our observation. But let us try species as applied to 

 Rubm even on this test, and return to the remark of high authorities. 

 " R. ccBsius is a good species, R. discolor is a better." So they are ; 

 excellent groups, for their forms are legion, for all practical purposes, 

 but distinctly failing as species ; if to run by imperceptible gradations 

 into other " species," constitutes a failure to the name. There is the 

 starved discolor of the north {R. rusticanus, Mercier), passing into the 

 enormous real discolor of W. and'N., as at Thames Ditton ; often with 

 its leaflets quite green beneath and broader than R. cordifolius, E. G. ; 

 and then this real R. discolor, W. and N., passes so gradually into a 

 Bramble, which you may call R. sylvaticus, and lump uniler the better 

 known R. villicauUs, AV. and N., or label R. thyrsoideus, Wimm., and 

 class with the very different typical form of that plant. And when we 

 get so far into the haven of the vague sylvaticus synonym, we stumble 

 unaware on the confines oimacrophjllus, and so much for the " species " 

 R. discolor. Now R. casius, L., repeats the same story in the glandular 

 brambles. Nothing can look more distinct than a nicely dried, fully 

 grown, well-selected panicle and stem of R ccesius in a little neat her- 

 barium ; here is surely firm ground at last. Why, we should know 

 it again in a hundred. So you would, some R. ccesius, but have you 

 ever examined that troublesome conterminous group — that mass of 



