

i 

all its wide range of variation by its large oblong-spathulate dull green 
leaves, large decurrent end- -spine, large distant deltoid-cuspidate mar- 
ginal teeth, stout peduncle with crowded ascending imbricating lanceo- 
late bract- leaves, the lower a foot or a foot and a half long, and very stout 
comparatively short panicle branches. I saw it in flower at Hyères, La 
Mortola, and San Remo. The leaves reach a length of six or eight feet, 
end a breadth of 15-16 inches. On the Riviera it is usually called A. 
Salmiana, but I believe that quite a dozen plants named or maintained | 
as species by Jacobi must range here as forms. 














Group AMERICAN. 
A. seemanniana, Jacobi. This I saw at Hyéres just coming into. 
flower, not. iari very materially from the plant of English con- 
servatories 
A Koch. This species, n in the open air at La 
Mortola, - is developed much better than any I have seen at. home. 
It has slightly glaucous oblong-spathulate ares, 3 feet long, 9-10 inches 
broad at the middle, a large pungent non- decurrent end-spine, and very | 
large irregular deltvid-cuspidate marginal teeth, with the edge hollowed 
out between them. I did not see it in flower. 
ais Scolymus, Karw. Not grown commonly in the Riviera phage 
but I saw it at Hyéres, La Mortola, and Monte Carlo, in flower at the 
last locality. A. Verschaffeltii, Lemaire, and several other plants “which x. 
have been described as species, must clearly be placed here. 
A. potatorum, Zucc. What was called by this name at La Mort ola, 
and it may be named correctly, was clearly conspecific with A. atrovirens 
(Salmiana). 
A. coccinea, Roezl. Grown at La boo qe not differing materially from 
the plant of English eonservatories l 
A. mexicana, Lam. The plant grown under this name at Hyères i 
no doubt named correctly, but I do not think in any broad sense is is 
more than a irati of americana. i 
A. americana, L. Everywhere abundant along the Riviera, not in 
gardens dii, hut by roadsides, and along the sea margin, flow , 
very freely. Besides the type, two varieties, one of which is called 
letevirens and Milleri, with very glaucous yin f towards 
mexicana, and another called ornata and pic a, with green leaves with 
great stripes of yellow. I saw one plant of ha latter with leaves 8-9 
feet long and nearly a foot ida at the middle. | 
Group Rieipz. 
A, rigida, Miller. This, the most valuable and most variable of all 
gaves, is common and quite at home in the Riviera gardens, 
flowering freely; an had an opportunity of studying its characters 
and range of variation far better I had ever done 
seeing hd a forms with which I was not previously a 
ommonest form the ere show the characteristic small distant, 
nearly black teeth, and ag ery well with what has been described 
red as Ixtli and "ietlioidis (Bot. Mag. t. à 
t St. i 
a oho “thick at the base) and forming a less dense rosette, 
ts called Cantala and Hampi in the Riviera ado are forms 

