AD FLORAM INDICAM (BALSAMINE X). 113 
several having all the leaves opposite may be referred to B, and 
others with them all alternate to D, E, F, and G. There are great 
difficulties attending the discrimination of the allies of Z. latifolia, 
Roylet, and sulcata. I. Goughii may be referred to Umbellate or 
Lateriflore, or even to Racemose. 
D. UwirLomz. Folia omnia alterna. Pedunculi 1-flori. (8р.84- 
47.) 
Another rather artificial assemblage, of which I suspect only 
a few (I. Balsamina, scabriuscula) have invariably 1-flowered pe- 
duncles, and of which the others, when 2-flowered, may be referred 
to Е. І. leptopoda is probably a state of І. latifolia. T. capillipes 
is a curious species, very unlike the others. 
E. LarEkRiFLORX. Folia omnia alterna, subdistantia, non omnia 
approximata v. subverticillata. Pedunculi solitarii v. fascicu- 
lati bracteolati, secus caulem elongatam plerumque axillares 
(non axillis supremis tantum siti), patentes, 2—6-flores. Flores 
racemosi, non umbellati. (Sp. 48-64.) 
A very natural group, if confined to those species which, like 
I. macrophylla, glandulifera, serrata, scabrida, &c., have elongated 
stems and peduncles from axils of many of the rather distant leaves ; 
but with regard to the others, several of them present specimens 
which fall naturally into G, and others when 1-flowered would be 
sought for in D. The group is divisible into several sections. 
F. UmBELLATA et CaPrrATX. Folia omnia alterna, versus apicem 
caulis conferta. Flores apices versus pedunculi elongati bre- 
vissime racemosi v. umbellati v. corymbosi.— (Bractez pleris- 
que conspicue.) (Sp. 65-78.) 
This group embraces certain closely allied peninsular species 
with simple stems and very short, many-flowered inflorescence, of 
which Z. linearis and umbellata are types ; but others are branched 
and few-flowered, as Г. grandis, campanulata, &c., and may be re- 
ferred to E. 
G. RacEMosX. Folia omnia alterna. Pedunculi plures, elongati, 
multiflori, demum folio longiores. Flores racemosi, interdum 
verticillati, v. fasciculati. (Sp. 80-95.) 
A very natural group on the whole, the principal exceptions. 
being Т. longicornu and I. urticifolia, some forms of which may be 
referred to E; I. Walkeri is closely allied to I. Jerdonie; I. cor- 
nigera to certain species in F. 
