150 ON THE UUMETOEUM GROUP OF RUBI IN BRITAIN. 



for all purposes of practical observation, absent. Bamelormi, then, 

 is essentially a glandulose group. Another character of some value in 

 helping us to keep clear of corylifolius is the tendency which the 

 sepals of all dumetorum Brambles show, some strongly, some faintly, to 

 become erect or at times adpressed. This statement, however, if too 

 rigidly accepted, wovdd make us miss the great lesson of this group, 

 namely, that in certain cases the direction of the sepals is of very little 

 value as a diagnostic character. This point, Mr. Baker observes, 

 dametorum illustrates better than any other Bramble ; and he is in- 

 clined, applying what may be learnt iiere, to conclude, that Balfonrianus 

 and althceifoUm are simply varieties of corylifolius, with adpressed 

 sepals. 



In drawing our line on the other side between dumetorum and 

 CfBsias, many immature and ill-nourished forms Avill occasion us some 

 perplexity. A curious point about c^slus is this, that although so 

 common and widely distributed in Britain, it is with us, under all its 

 variations, essentially eglandulose, but this is not the case either in 

 France or Germany. (See E. serpens of Grenier and Godr., and R. 

 ccBsius, var. hisjndus and ferox of Wcihe and Nees.) Generally, then, 

 the essential points* of ccesius, comparing it with dumetorum and 

 corylifolius, are its much weaker growth, slender terete always con- 

 spicuously bloomy stems, copious very weak subulate prickles, ternate 

 fully developed leaves, leaflets essentially more irregularly toothed and 

 less hairy beneath, sparse corymbose inflorescence, sepals more length- 

 ened out at the point, always adpressed to the fruit, which is of fewer 

 larger grains, and always decidedly pruinose. 



In Prof. Babiugton's 'British Rubi' the two partly or wholly equi- 

 valent factors of which onr dumetorum group is composed, are classed in 

 different sections to each other. They seem, however, so nearly, or even 

 at times inextricably, related, that it is most hard to disjoin them. A 

 fuller view of the group, by adding in such links as concinnus and 

 pilosus, renders this separation doubly difficult.. I am free, as I said 

 above, to allow strong Koehlerian affinities in divers folius, but this is 

 only another instance of those puzzling cross-resemblances in Rubi, 

 which make any mere Imear arrangement of their subspecies in- 

 evitably fall far short of expressing the whole truth concerning 

 their analogies. There is, indeed, a question whether some forms of 

 * I owe these remarks to Mr. Eaker. 



