NOTES ON LIMONIUM 7 
ee re pyrami al form, and thes intermediates are sey 
n us. The remark that the « variety seemed to prefer the 
asae’ parts of the marsh is directly opposed to Syme’s statement 
th when salt marsh was drier than usual, the panicle h 
a tendency to revert to the co ne 7 ateness of 
owering re y Long and other observers may 
more extensive panicle, encouraged by some /ocal condition of soil, 
degree of moisture, &c. Pfeffer (Physiology o. wes English ed., 
li. 116 (1903) ) says :—** Hither an excess or d y of food may 
suppress the formation of flowers; in the fi = casey fouitng to the 
excessive development of the vegetative parts, and in the second 
because a starved plant lacks the vigour required for the production 
of reproductive organs. If not too severe, however, partial starva- 
tion usually tani the formation of reproductive organs on a 
previously well-nouris lant.’’ 
The above Slekevakinatt which agree with my own, show that 
L, vulgare, known to be a very variable plant both as regards leaves 
and scape, alters its facies very greatly in the same country under 
local conditions connected with its habitat. Thus, sea lavenders 
grow on banks bordering ae some little way from the sea, or iu 
salt-marshes which vary much in humidity eee to the se age 
me 
submerged at each tide—this in itself causes certain pebiliarities 
(flexuous spikes, &c.). pa rag under the hot southern sun, the 
plant—particularly if growing on drier ground than usual—shows 
a tendency to produce a taller scape, finer in all its parts, than in 
more northern countries ; Syme’s “ pyramidalis”’ is chiefly found in 
riant form or variety appears, the ‘‘ macroclada”’ of Boissier, which 
is another step further away from the normal plant of England and 
Northern Europe. Analogous to this would be Castalia speciosa and 
its var. minor, which is found chiefly in more northern stations. 
Dr. Li. M. Neuman tells me that the luxuriant form of L. vulgare 
sometimes occu urs in Sweden, where it is distinguished as the 
dy ne ” of Statice scanica. 
However, the name ‘‘pyramidalis”’ is of longer standing, and it 
will be convenient to distinguish our tall and luxuriant specimens 
of L. ect as f. pyranidale 
will now examine Boissier’s ‘‘ y mac: stat described in 
DC. Predvonisn, xii. 645, as follows :—‘ Sepe glaucescens, panicule 
ramosissime ramis elongatis patentissimis, ccrwre laxis vel densis 
scorpioideis. Ad littora maris Adriatici, Sicilie, Rumeliz, yrie.” 
I believe this is widely distributed in Southern Europe, stretching 
much further west than the Adriatic—in fact, to Spain and 
Portugal. 
In Gillet and ae a s Nouv. Fl. Frangaise, ed. vii. 406 ia it 
is thus described (under. the synonym S. serotina Reichb): ‘ 
int. 2 fois plus dio que l’extér.; épillets plus petits, forse 
panicule & rameaux étalés, flexueux; feuilles coriaces, oblongues, 
