ON THE GENUS RUBUS. AAD 
is derived; but I have proposed to restrict the word cyme to the 
latter arrangement, as exhibiting a peculiar and characteristic 
mode of division for which we otherwise want a name. Where 
however this arrangement occurs once and is not repeated, it can 
hardly be called a cyme; it might perhaps be named a cymel or 
a cymet, or, as the word is Greek, a cymary, but I prefer the 
English word knot, as sufficiently expressive. A knot then is a 
group of three flowers or fruits, one in the middle on a short 
stalk which is a continuation of the common stalk, and one on 
each side on longer stalks, springing at one point or nearly at one 
point from the central stalk. These knots characterize the inflo- 
rescence of Rubus, and the panicle is rarely without some indica- 
tion of a tendency to form them. It is however often only a 
tendency, sometimes on one side a mere rudiment, or perhaps a 
bract indicating the place where the stalk should be; sometimes 
even this is wanting, and there is merely a branched stalk, always 
however with the one branch forming a continuation of the 
common stalk. Where there are two stalks they are perhaps not 
opposite, and the lateral stalk is not always longer than the central 
one. Again, we do not find even these indications throughout 
the panicle, and indeed in the first section of the Brambles it is 
hardly exhibited except with the terminal flower of the panicle. 
On the other hand, the arrangement is sometimes repeated in the 
lower branches of the panicle. Yet even here I should hardly 
call it a cyme, since it shows no tendency to repeat itself any 
further, and is mixed with the more simple arrangements, but 
would term it a compound knot. 
The prevailing panicle in what is usually the first segment 
among the Brambles has been called a corymb. This is hardly 
correct. The stalks are not successively longer so as to produce 
a conical outline, but the form is rather that of a short cylinder, 
each lateral stalk supporting, not a knot, but a single flower; 
where however the plant is luxuriant and the panicle large, one 
or two knots may be found on the lowest stalks. 
It is one of the difficulties of the genus Rubus that we often 
are at a loss to say whether the whole of the flowering-shoot is 
to be considered as a leafy panicle, or whether it ought to be de- 
scribed as branched and supporting several panicles. If we look 
through the ‘ Rubi Germanici’ for this purpose, we may find ex- 
amples to support either view of the subject, and still more if we 
