PLatE XCI. 
(A, B, C.) ROMULEA coLtumn#, Seb. et M.; (D, E, F.) R. coLumnaz 
a DIscRETA, Moggridge. 
Natural Order Ir1pAce#. 
GEN. Cuar.—See description of Plate XCIII. below. 
(A, B, C.) Spec. Coar.—Scape 2-3 flowered. Upper valve of spathe 
usually hyalino-membranous, more rarely having a narrow central line of 
herbaceous tissue. Perianth usually very small, nearly white, with one 
purplish central stripe; segments oblongo-lanceolate, very pale yellow- 
green outside with a purplish central stripe. Style slightly shorter than 
the stamens. Leaves 2-3 inches in length, stout, recurved. Romulea 
Columne, Seb. et M. Fl. Romane Prodr. p. 18 ; Woods, Tour. Fl. p. 357 ; 
Ardoino, Fl. Alp. Mar. p. 863; Trichonema Columne, Rchb. Fl. Excurs. 
i. 83; Gren. et Godr. Fl. de Fr. ii. 2388, 
(D, H, F.) Spec. Cuar.—Scape 1-flowered. Upper valve of spathe com- 
pletely membranous. Pertanth from half as large again to twice as large 
as that of &. Columne, white or bluish-white with grey purple tinge, 
and three purplish stripes; segments oblongo-lanceolate, the outer ones 
being yellow-green on the back with three broad purplish stripes. Style 
slightly shorter than the stamens. Leaves lineari-filiform, less curved 
and more slender than in the species. Romulea Columne a discreta, 
Moggridge. 
Hasirats.—( A, B, C.) Cabrolles Valley, Mentone, where I collected the 
flowering specimens on March drd, and the fruiting April 17th, 1870. 
(D,, F.) Cap Croisette, Cannes, collected by my father, March 23rd, 1870. 
Remarks.—I have as yet seen but half a dozen fresh specimens (all 
from the Croisette at Cannes) of 2. Columne var. discreta represented in 
this plate. It has a distinct look, but I doubt whether it will prove to be 
permanently separable as a species. 
I have not been able myself to botanize the neighbourhood of Cannes at 
the time when the species of Romulea are in flower, but I have reason to 
think that several interesting forms might be discovered there, A single 
bulb of the variety discreta cultivated at Mentone reproduced this spring 
the characters of the plant represented at figure D, and yielded the cap- 
sule drawn in the left-hand corner of this plate. I hope to sow the seeds 
next autumn, with a view to observing whether the peculiarities of the 
variety are maintained. Judging from dried specimens there are slight 
differences between the form assumed by R. Column, Seb. et M. at Men- 
tone, and the British plant from Jersey and Dawlish, these latter 
having larger and pale violet and green-tinted flowers, the segments 
of which overlap one another. 
