442 
47. COMPOSITACEJE. 
slightly tomentose on both sides with the midrib w. but pre- 
sently becoming more naked or glabrescent than in G. saxatile 
(L.), 1^-2 in. long, ^-1^ line wide, the upper crowded stiffly 
erect thickish and rigid very straight and entire with strongly 
revolute margins ; the lower broader and flatter erecto-patent 
not revolute or waved at the margins ; on the whole all are 
stiller longer and narrower than in G. rupestre above, in which 
they rarely exceed an inch in length and are of a softer thinner 
substance with at least the lower flaccidly or loosely spreading 
or recurved. Ped. slightly cottony gradually thickened up- 
wards and distinctly swollen beneath the heads which are de- 
pressedly globose in bud, semiglobose in fl., 4-5 lines broad and 
high, swelling out abruptly, like that of a Sonchus, at the base 
and of a thick short squab instead of pyramidal or slender 
ovate-oblong more or less elongated shape ; disk broad with 
numerous florets. Scales of inv. wholly scarious or membranous 
(not coriaceous or rigid) light brown or pale chestnut or fawn- 
colour with a darker stripe up the middle, all altogether loose 
but erect and not even their tips recurved or spreading; the 
outer of a short abrupt broadly spathulate or rounded shape, 
the innermost ligulate and lacerato-truncate. Female florets 
of the margin numerous in a broad band or ring sulphur or 
lemon-y. with a very long slender tube from which are ex- 
serted the 2 divaricate branches of the style ; perfect fi. of disk 
or rather their anthers golden-y. ; the 5 teeth of all the fl. ob- 
tuse and either tipped with a tuft of glandular short hairs or 
all over glandular-pubescent. Pappus sessile nearly smooth. 
Ach. hairy stipitate. Ilecept. quite naked papillose. When 
the seeds are fully ripe, the scales of the inv. become altogether 
reflexed back upon the ped. I 11 G. saxatile (L.) and G. ru- 
pestre (L.) they are merely horizontally spreading. 
P. Tenorii Gren. et Godr., by the scales of the inv. “ toutes ar- 
rondies au sommet,” agrees somewhat better with G. calycinum 
than with G. rupestre : but the rest of the description rather, 
and the synonyms entirely, belong to the latter. 
Seeing the great variableness of G. rupestre on the one hand, 
and the rare and casual occurrence of G. calycinum on the other, 
I have sometimes doubted whether the latter were more than an 
extreme form of var. /3 of the former. The difference however 
above noted in the teeth of the florets, if corroborated and 
found constant, would be quite decisive. 
The following is an extract from a letter dated “ Geneve 15 
F6vr. 1834” from the late Chev. I)e Candollo about this pi. 
“ Quant au no. 231 ” (P. Bennettii Lowe Mad. MS. dim) 
