146 THE JOURNAL OF BOTANY 
article as that commonly met with here must be referred to the 
subspecies horminoides. This is equally evident from its habit, 
its broad and — radical leaves, dark-coloured calyces, and 
inconspicuous corollas 
considerable degree of variation, however, may be found, 
par oa 3 in the leaf-cutting, which is certainly much influenced 
by local conditions, and sometimes shows remarkable differences, 
even on the same individual. The typical subspecies may be sai 
to be the plant growing in ordinary seasons on the sand-hills at 
Deal, in which the radical leaves are for the most part broadly 
ovate, and slightly sinuate with shallow, rounded lobes bluntly 
serrate along the margins. In dry, exposed situations, the foliage 
radical leaves uniformly pinnatifid, with contiguous, crenate lobes, 
idrib. 
and meee reaching fully ser ay to the m Between these 
extremes every gradation may be found here as on the Continent, 
the Siete features in the foliage of these horminoides forms 
that are visible in the herbarium being that the radical leaves are 
never very narrow, and the uppermost cauline pair invariably very 
pee 
ish examples also show noticeable gk get in the pilosity 
of the stom — the number of its glandular hairs, whic very 
scarce in specimens, and extremely bo eek in others ; but 
I cannot find that the variations i in a respects are in any way 
correlated with those of other chara 
The calyx in our plants is uate uniform, varying only 
in size and in the somewhat greater abundance of os in 
those erg where the glandular hairs of the stem are most 
humer 
Of yes corolla, which very frequently shrivels in dried speci- 
mens, and often away, 1 have been able to make a series of 
observations in the field which entirely confirm M. Briquet’s 
statements as to its polymorphism. The descriptions in our 
British ian’ seem to overlook this important characteristic 
(although they note the dimorphic flowers of S. pratensis) and 
rs applicable only to a corolla of semi- cleistogamous form which 
appens to be very prevalent. This polymorphism is not readily 
seen in most dried specimens, and was not apparent to me on first 
examining the collection in the National Herbarium; but in 
September, 1905, when at Newquay, Cornwall, I contrived to 
recognize, among a number of plants presenting, as I thought, 
the ordinary appearance of S. Verbenaca, one individual with 
the corollas conspicuously large and exserted, with the lips sub- 
equal but divergent, and the upper one rey eee bares thus 
according fairly well with the intermediate flowe by 
M. Briquet. 
