vi PREFACE. 
anthers. DeCandolle sugzests that the outer series may 
be transformed petals. 
Petals are frequently transformed into stamina in 
Bucklandia, which is exstipulate. The anthers are val- 
vular in Sedgiwickia and Fothergilla. The ovula are 
12 in Bucklandia, the lower two of each cell becom- 
ing perfect seeds. They are indefinite in Sedgiwickia 
in which the lowermost only, and that not often, is 
perfected. Both these genera have the calyx of Fother- 
gilla and the wood of Conifere. Mr. Lindley is 
mistaken in assigning deciduous valves to the anthers ; 
this I think is mentioned by Mr. Brown as occur- 
rng only in Hammallis sinensis. 
The valves of Bucklandia are very unequal, both 
. are permanent. This genus is remarkable for having 
its buds which are always terminal enveloped in two 
scales, resembling much in form and situation stipula. 
Haloragee.—In Mynophyllum tetrandium Roxb., 
the seeds are arcuate and the embryo is necessarily 
curved. 
Circeacee.—One species occurs on the Khasyah 
Rhizophoree.—The stamina in Rhizophora mangle 
are not twice the number of the petals but varying 
from 9 to 12. In Rhizophora as I limit it, they do not 
at all spring from the same points as the petals. The — 
spibend in this genus are multilocellar the cuticle sepa- 
: valves a as s the calyx i is 
and perhaps in. Carallia. In 
Tu PER BIA EIE TOT NE UN NUS IER STREET 
hsdid lm AM f CE Wee 
Es Te PIER OE Re e IM 
rl throughout into anticotus and — 
