520 
47. COMPOSITACEJS. 
leaf-like br. Scales of inv. membranous feebly spinescent or 
cuspidate gr. with pale or whitish margins. FI. not conspicu- 
ous with few small florets contracted into a tuft or pencil, all 
narrowly and shortly ligulate with their tube clothed upwards 
for the greater part of its length with dark brown or black tine 
short woolly hairs. Ach. with its coat small obcompressed or 
flattened, scale-like or resembling a pod or silicle of Lepidium 
or Thlaspi, oblong or obovate-oblong rounded or subtruncate at 
top, 5 millim. long, 3-3^ broad, close-packed in a compact 
shortly conoidal mass, the inner only deciduous, the outer per- 
sistent and concrete on the alveolate shortly pyramidal obtuse 
recept. by then’ chaffy coat or pseudo-pericarp ; stripped of this 
they are pale grey, smooth and even, narrow-pearshaped ob- 
compressedly flattened subtriquetrous dorsallv convex, their 
ventral face broad and flat, coated and broadly winged all 
round and at top with the closely adhering w. or pale straw- 
col. pericarp-like chaff completely enclosing it except a small 
Y-like opening on the shining inner face at top, in which ap- 
pears its scar-like apex or epigynous disk. Pappus 0 except 
a minute entire rim or border round the exposed knob -like 
epig. disk. 
Brotero’s description of this pi. is most full and accurate, 
especially as to the ach. or fr. 
Sc. hispamcus L. is common in Tenerife, especially between 
Cruz and Laguna, but it has not occurred in Mad., though by 
some mistake DC. has referred on my authority to it instead 
of to S. maculatus, the only Mad. Scolymus ever found in 
Mad. (See Proceed. Linn. Soc. v. no. 17, p. 40.) It is at 
once distinguished by its much less strongly or hard-spined 
and nerved 1., larger orange or darker y. fl. sessile or on short 
axillary side branchlets all the way up the often very tall 
(3-6 ft.) mostly simple virgate not corymbosely branched st., 
shortly membranaceo-coroniform 2-setose pappus, &c. 
Scolymus is certainly allied by its psoudoperiearpiate ach. to Xan- 
thium , but not more than Lappa is by habit, and, indeed, 
allied by its ach. but a degree or two more than Xerotium , 
Micropus &c., or even Hedypnois, Hyoseris , Bhayadiolus, 
Thrincia &c. The former relation however explains perhaps 
its being placed by Grenier and Godron at the end of Compo- 
sitacece immediately before Xanthium. But Lappa has equal 
and perhaps better claims to be so placed. And certainly in 
point of habit Scolymus ought not to be far separated from 
Carlina , Car/ ham us See., and comes in very badly between Crc- 
pidccB ( Crepis , Hieracium, Andryala &c.) and Amhrosiacca 
{X ant hi u m, Ambrosia ) . 
