and comose appearance of the bracts, is not unfrequently much reduced 
in P. niceensis, when it could scarcely be said to differ from P. vulgaris. 
The supposed distinctive points of P. niceensis and P. vulgaris may be 
compared as follows :-— 
P. niceensis.—Intermediate bract P. vulgaris.—Intermediate bract 
longer than young buds; lateral | shorter than young buds; lateral 
bracts equalling or exceeding the | bracts only half as long as pedicel; 
pedicel, subelliptic ; wings larger. suborbicular ; wings smaller. 
But on fixing my attention on any one of these details, and applying 
the test of absolute and relative measurement, it became evident that 
though in the majority of cases these characters are really distinctive, in 
a large minority there is variation which tends to bridge over the inter- 
val between the two species. For example, in ten plants of P. niceensis 
examined, the intermediate bract varied in length from 3 to 53 milli- 
metres; though the length of these bracts in each individual plant was 
constant with but slight exceptions. 
Similarly the lateral bracts varied in shape and length; and the wings 
in proportion and absolute length and breadth as represented in the 
figures D, HE, F, G, where the wings of the plant D measure 9 mm. by 7 
broad ; of E, 8 by 54; of F, 102 by 63; of G, 94 by 53. 
In the monograph of the Italian species of Polygala by Prof. Caruel, 
mentioned above, P. nicwensis is described as having pink flowers, but it 
is a curious fact that at Nice, Mentone, and Cannes the flowers of this 
plant are almost always deep blue, and very rarely white or pink, while 
from Oneglia to Genoa its flowers are, as Prof. Caruel describes them, of 
a fine deep pink. This pink-flowered form, which appears to me to con- 
stitute a distinct subvariety, characterized by its straight and more dense 
racemes and upright habit, is found as far westward as the banks of the 
Nervia river near Ventimiglia, about nine miles east of Mentone. It is, I 
believe, the plant which has been taken by some botanists for Polygala 
anatolica, Boiss.,* with which it has no connexion, P. anatolica having a 
carpophore but slightly shorter than the ripe capsule itself, and the tube of 
the corolla projecting beyond the wings, while the carpophore of P. nice2n- 
sis is extremely short, and the corolla tube quite included. P. rosea, 
Dsf., which has frequently been confused with P. newensis, is much more 
closely related to P. anatolica, Boiss., and P. major, a very handsome plant, 
and has the carpophore nearly half as long as the capsule, and petals 
which project considerably. PP. Preslit, Spreng., a Sicilian species, has 
narrow ovate lanceolate wings, almost recalling those of P. monspeliaca, 
L., anda more exsert corolla than P. nicwensis, which it otherwise closely 
resembles, and of which it may indeed be a variety, being labelled by 
M. Boissier in his*herbarium P. niewensis, var. Preslit. 
* T am greatly indebted to MM. Boissier and Reuter for a series of specimens of 
several species of Polygala, and among them P. anatolica, P. rosea, Dsf., and P. Preslit, 
Spreng. 
—_e ee 
