DI A NTH US 
9i 
3. DIANTHUS CAESIUS. Cheddar Pink. Plate 88 
Tunica rupestris folio caesio molli flore cameo Dillenius Hort . Eltham. 400, t. 298, fig. 385 (1732). 
Dianthus caesius Smith Eng. Bot. no. 62 (1792)!; in Trans. Linn. Soc. ii, 302 (1794); FI. Brit. 463 
(1800); Syme Eng. Bot. ii, 48 (1864); Rouy et Foucaud FI. France iii, 174 (1896) excl. syn. D. gratianopolitanus 
Villars ; Hist. PI. Dauph. iii, 598 (1789); D. virgineus var. /3 L. Sp. PI. 412 (1753); D. glaucus Hudson FI. 
Angl. 161 [bis] (1762) excl. diag. ; ed. 2, 185 (1778) partim, pro min. part., non L. 
leones : — Smith Eng. Bot. t. 62 ; Reichenbach Icon, vi, t. 265, fig. 5044. 
Camb. Brit. FI. iii. Plate 88 . (a) Barren and flowering shoots. ( b , c ) Fertile shoots. ( d ) Portion of 
leaf (enlarged). ( e ) Petal. ( f ) Ovary, (g) Mature fruit. Somerset. 
Exsiccata : — Billot, 935 et 935 bis; Michalet {FI. Seq. Exsicc.), 26; {PI. Jura), 60; Reichenbach, 798; 
2293 ; Thielens et Devos, ii, 104. (All as D. caesius.) 
Perennial. Shoot strongly glaucous, up to 1 — 3 dm. high. Leaves linear, connate at the base, 
margin scabriusculous, apex acute, up to 4 — 5 cm. long and 3 — 4 mm. broad. Inflorescence with 
1 — 3 flowers. Flowers odorous, protandrous, about 2^5 cm. in diameter; late May to early July. 
Epicalyx appressed, with 4 segments, about 5 mm. high, the outer pair obovate and longer and over- 
lapping the inner pair, about one-third to one-quarter as long as the calyx-tube, all mucronate. Calyx 
faintly veined ; tube about 1 '8 cm. long and 4 mm. broad ; teeth nearly oblong, about 2 — 4 mm. 
long. Petals rose-pink or pink, with a distinct whitish claw, nearly contiguous, either shortly or 
deeply and rather irregularly toothed. Stamens of different lengths, ultimately about as long as the 
calyx, filaments white, anthers subversatile and yellowish. Capsule cylindrical. Styles white, ulti- 
mately as long as the ovary and protruding. Seeds large, punctate. 
If the Cheddar pink is really indigenous in Somerset — and we know of no British botanist who has denied this — the 
plant is one of a number of interesting central European species whose British distribution is restricted to the vicinity of the 
Bristol Channel. The following are examples of the species in question : — Dianthus caesius, Paeo7iia mascula {= P. corallina), 
Arabis stricta, Draba aizo'ides, Euphorbia pilosa, Stachys alpina, Hieracium lima, H. stenolepis, Koehleria vallesiana, Allium sphaero- 
cephalum. Several of these species are open to a more or less strong suspicion of having been introduced, whilst others are 
unquestionably indigenous ; and it is by no means easy to decide in which of these two categories the Cheddar pink should 
be placed. The continental distribution of the plant is rather against its being considered native in Somerset ; but, on the 
other hand, the plant has the appearance of an aboriginal in three localities on the limestone cliffs of the Mendip Hills. 
Various binominals have in recent years been brought into conflict with the name of Dianthus caesius of Smith. Of 
these, perhaps the most serious competitor is “ D. glaucus Hudson.” The binominal D. glaucus appears first in the Species 
Plantarum of Linnaeus (p. 41 1 (1753)): here D. glaucus refers to a plant closely allied to D. deltoides L., and with regard 
to which (see p. 90) we follow de Candolle and Syme in reducing it to a variety (var. glaucus) of D. deltoides. In this first 
edition of the Species Plantarum (412 (1753)), Linnaeus names the Cheddar pink D. virgineus var. (3. Linnaeus had never 
seen the plant he so named, but merely used a description and figure of Dillenius {loc. cit.). Linnaeus makes no change 
in the second edition of the Spec. Plantarum , apart from adding the word “brevibus” (“ squamis...brevibus ”) to his 
original diagnosis of D. glaucus. It is, we believe, admitted by all that the D. glaucus of both editions of Linnaeus’s Spec. 
Plantarum refers to one and the same plant (namely, to D. deltoides var. glaucus DC.). Now Hudson, in the first edition 
of his FI. Anglica (p. 161 [bis] (1762)) has a D. glaucus-, and this is unquestionably based on D. glaucus Linn., for Hudson 
repeats — ipsissima verba — the diagnosis and synonyms of Linnaeus, even to the extent of copying the erroneous citation 
(t. 348 for t. 384) from Dillenius. It is true that Hudson adds the locality “on Cheddar rocks,” and so perhaps adds 
D. caesius Smith to D. glaucus Linn. ; but this inclusion is not certain, as two or three species of Dianthus have been 
recorded from Cheddar (vide Hudson loc. cit., sub nominibus D. glauco et D. arenario, et vide Withering Bot. Arr. ed. 2, 
444 (1787) sub nomine D. glauco). However in the second edition of his FI. Anglica (p. 185 (1785)), Hudson substitutes 
another Dillenian name for the one utilised by Linnaeus ; and the substituted name undoubtedly refers to D. caesius. It 
has to be admitted therefore that the D. caesius of Smith is the D. glaucus of Hudson in part; but as Hudson’s diag- 
nosis 1 remains unchanged (i.e., as it remains the diagnosis of the original D. glaucus Linn.), we cannot allow the name 
“ D. glaucus Hudson ” to supersede Smith’s name of D. caesius. The fact is that Hudson blundered in his allocation of 
the name D. glaucus Linn. ; and his later substitution of another Dillenian synonym was merely a cloak apparently manu- 
factured with the intention of hiding his original mistake. Again, the name D. glaucus was formerly kept up as a species 
by many botanists, and it may well be that some future botanists will revert to this view — a view which is not at all un- 
reasonable ; and should they do so, they will rightly claim the name D. glaucus for their plant. 
Another name has also come into conflict with D. caesius, viz., D. gratianopolitanus Villars {loc. cit.). This name 
was taken up by the late Rev. R. P. Murray in the text (p. 44) of his Flora of Somerset, but not on the frontispiece of 
that work where the name D. caesius is used. On this matter we agree with N. E. Brown who states {Eng. Bot. ed. 3, 
suppl. additions and corrections, p. v (1892)) that “the description of D. gratianopolitanus does not agree at all with D. caesius.” 
It may also be (since some authorities regard D. virgineus L. excl. var. (3 as undeterminable) that the name D. virgmeus 
will enter the arena in conflict with D. caesius, though we ourselves would have no sympathy with .the view that D. virgineus 
should displace D. caesius. 
It only remains to add that Smith’s name D. caesius is in general use, and is perfectly definite and clear, and that the 
names proposed to supersede it are shrouded in ambiguity or error. 
1 Cf. the footnote on p. 83. 
12 — 2 
