Rubiales. 
233 
Affinities of Caprifoliacece: Sambucus. As we have remarked 
already, the distinction between the flowers of Caprifoliacece and 
Rubiaceae is not critical. Here again we are driven to vegetative 
characters as criteria; and in general facies the constantly her¬ 
baceous Galieae differ much more widely from the rest of the 
Rubiaceae than do the woody Caprifoliaceae. Moreover the flowers 
in the latter family display a range of diversity in essential 
characters even greater than that exhibited among Rubiaceae. 
In this connection, as with Galieae, a great deal of stress has 
been laid on a single vegetative character—stipules. The presence 
or absence of these organs is the sole constant feature of distinction 
between Rubiaceae and Caprifoliaceae, and even this fails in the 
case of Sambucus ; but the latter differs from Rubiaceae, from the 
rest of Caprifoliaceae, and indeed from most sympetalous shrubs and 
trees, in having compound leaves. In virtue, apparently, of its 
stipules alone, Sambucus has been regarded by some botanists as a 
link between Rubiaceae and Caprifoliaceae. The need would seem 
to be to discover a gulf rather than a link between these two families, 
at least from tl>e aspect of progressive evolution. Sambucus has 
further been considered as relatively primitive, and as affording the 
way through which the connection between the epigynousSympetalae 
and the epigynous Polypetalae may be traced. The 3-5-locular ovary 
with a single pendulous ovule in each loculus, and the branching of 
the leaf-lamina are characters which suggest affinity with 
Araliaceae, apart from the general umbellifloral characters which 
Sambucus shares with other Caprifoliaceae ; and it may be that this 
genus is of all the Rubiales one of the most nearly allied to the 
umbellifloral stock. But the question is scarcely of primary 
importance for our purposes, and we may dismiss it with the 
suggestion that if Sambucus it to be separated from Caprifoliaceae 
as a distinct family, then Galieae should certainly receive similar 
treatment. 
The value of stipules in classification. The foregoing has 
suggested a question of somewhat serious importance in regard to 
the Rubiales, namely, the phyletic value of stipules. On the face 
of it, it is difficult to conceive that a minor vegetative character 
can have any material weight in broad general considerations of 
phylogeny. Indeed, the character must needs be disregarded in the 
recognition of the descent of Rubiales from an umbellifloral ancestry, 
for in Umbelliflorae stipules are conspicuous by their absence; 
similarly the arrangement of the leaves upon the axis must be set 
