NEW SPECIES OF POLYGALA FROM SOUTHERN CHINA. 119 
having the margin very much wider at the rounded apical angles than 
below. If any Chinese species is to be reduced, it will probably be P. 
elegans, Wall., which is likely to prove a form of P. Japonica, Houtt. 
Of the latter plant I possess good specimens from two Japanese locali- 
ties, from Formosa, and from Australia Felix (for I entirely agree with 
Dr. Mueller and Mr, Bentham in considering Mueller’s former P. 
veronicea as not distinct; itis a stunted form, with smaller flowers, but 
otherwise quite like the Asiatic plant). There is little to distinguish 
the Chinese from the Japanese species, except that the former has 
usually rather acuter calyx-wings: in the lateral or terminal position of 
the racemes, and the number of flowers in each, I find no constant dif- 
ference, and both plants are conspicuous for the gradually-diminished 
size of their leaves from above downwards. A few years back I re- 
ceived specimens from Foochow,—not, to my regret, now accessible for 
re-examination—which I was quite unable to refer to the one or the 
other species with any certainty. By ¿rue P. glomerata I mean the 
typical Chinese plant, which is by no means rare here and in Hongkong, 
though never, I believe, growing gregariously like P. elegans, but 
always found as isolated specimens. Dr. Thwaites has referred to P. 
glomerata two Ceylon plants (C. P. n. 592 t and 1079 !), the one his 
variety a. pedunculosa, the other P. hirsutula of Arnott. But, with all 
respect for my acute friend's opinion, it appears to me that the habit, 
weaker stems, scattered leaves, slender elongated peduncles, larger 
purplish flowers, and narrow-winged fruit of both these plants, whether 
distinct inter se or forms of one species, decidedly negative such a com- 
bination. The most discordant and confused views prevail amongst 
all writers as to the species of this genus. To prevent misapprehen- 
sion as to my own moderate conservative opinions, I may state that I 
do not consider many of the European species proposed of late years, 
as, for instance, P. Desangelisii, Ten.!, and P. Lebelii, Bor. !, to have 
. anything like a well-established claim to specific rank; and, more- 
over, I believe all the reputed modern species grouped round the old 
Linnean P. vulgaris need a careful and prolonged comparative study 
before any decided opinion ean be formed respecting them. But still 
there should be a method in inquiries, and to adopt the extreme views 
of some botanists is to my mind impossible. MM. Grenier and 
Godron, while admitting several of the most. doubtful species split off 
from P. vulgaris, nevertheless combine such extremely dissimilar plants — 
L4 
