260 
A MARYLLID ACE^E. 
ficiently described, I cannot determine whether it 
is allied to the foregoing group, or to 9. Brachyan- 
drum. 
§. 2. Semipatentia. — Half-patent. 
A. Foliis scepius undulatis, umbelld l -'20-flord, laciniis latis, 
germ. sess. vel subsess. — Leaves usually undulated; umbel 
1 -20 -flowered, segments broad ; germen sessile or subsessile. 
29. Broussonetianum. — Bot. Mag. 47. 2121. 
Var. 2. Pluriflorum. — Parad. Londin. 52. C. Yuccse- 
florum. 
Var. 3. Yuccseides. — Foliis non undulatis. Spec, in 
Herb. Soe. Linn. 
The first of these produces usually a solitary flower, with 
a fine purple stripe on each segment, the second two or more; 
the third variety, which is more delicate, has the leaves not 
undulated, the flowers two or more. They are all natives of 
Sierra Leone, and not materially distinguishable. It is ad- 
visable to leave them dry in the winter. Probably from 
some constitutional difference I failed in one or two attempts 
to cross this species with Capense, which shews no disincli- 
nation to breed with any Crinum of Asia or America. 
30. Distychum. — Am. Ornata. a. Bot. Mag. 31. 1253. 
This little plant, a native also of Sierra Leone, 
differs from all the rest of the genus in having 
leaves not multifarious, but alternating to the right 
and left; and, contrary to the prevailing habit of 
this section, they are not undulated. It produces 
usually a solitary flower, striped with red, and is 
allied to Broussonetianum. 
31. Petiolatum . — Foliis latis petiolatis, costosis, nervis vix 
parallelis ex costa oriundis, floribus albis, umbelld sub - 
8 -fiord. 
Var. 1. Spectabile. — A. ornata. ft. Bot. Mag. 923. C. 
Giganteum And. Rep. 169. This fine species with 
broad and large conspicuous white flowers, is a 
native of the tropical west coast of Africa. The 
name, giganteum, has the priority, but it would be 
disgraceful to science to retain the name gigantic 
for a species of inferior size. Such names, as this 
and Thunbergia fragrans for a scentless plant, 
must be altered, being much more objectionable 
