loa 
U/& 
[462] TRANS. OF THE ACAD. OF SCIENCE. 12 
Rota! and also Naples, to the Crimea, and reaching east- 
wardly far into store (Caucasus, Hohenacker! 409 and 1939, 
i nt” Herb. Tournefort !) ‘It varies 
LE 
® 
55 
2s 
6 
me 
e 
the proportion of the sigma and ‘ene nd no perma- 
nent character in them; the style proper is E isoee or shorter 
than the stigmatic portion ; and this part is cylindrical or 
subulate in specimens not otherwise distinguishable ; the stig- 
ma is usually pale polis, Babi or, when dry, dark red, rarely 
d Sweden, is a luxuri nt form, overgrown at the ‘expense of 
the succulent herb, which it destroys. 
Var. 8. MACRANTHERA; C. macranthera, Heldr. & Sart.! in 
sched.; Boiss.! diag. or, II. 3, ch C. Calliopes, Heldr. & 
Sart.! Boiss.! ibid, 128.—Large flowers on very short pedicels ; 
be 
large, often sinless hh n the filam pai ; candies usually shorter 
n 
appearance in green-houses on rica and other evergreen 
shrubs; this garden form is C. xanthonema of the Paris Jar- 
din des Plantes. 
ar. y. ? OBTU pack this very curious form was collected by 
Funk! (Herb. Cosson and Hb. Reichenbach) in the Sierra 
Nevada of Spain on conse shrubby Genista ; the glomerules 
consist of 3-5 flowers, only, on pedicels longer than the calyx; 
lobes of calyx and corolla broadly oval, obtuse, shorter than 
the tube of the corolla; scales large ; ‘styles as in the com- 
mon form. I would, at once, have acknowledged this. pecu- 
liar plant as a distinct species, if a second x ari had not 
ar K 
a small head; scales narrow; styles ordinary, seeds very 
small (0.3 lines diam.) The shi rmer may be pi Atce  Se as 
var. — a, the latter as v a. 
Var. 6. ? SAGITTANTHERA*; allied to var. angustiloba, distin- 
* Philologists will blame this “ vox — but daily experience 
teaches us and philological research confirms, that words are not formed 
