AND STERCULIACES. 105 
genus in the more clavate branches of the style with less strictly 
terminal stigmas ; but,as far as I am aware, the character is scarcely 
sufficiently marked to form even a good section. 
SPIUERALCEA, A. de St. Hil. 
Harvey proposes to distinguish under the name of Spheroma 
two species which differ from the others, as Lavatera from Malva, 
by the bracts connate at the base, and which has appeared in 
the Cape species to be confirmed by a difference in habit. But 
when the American species come to be examined, it will be found 
that the free and connate bracts pass gradually the one into the 
other, without any relation to habit or other characters. As to 
the rule that if a character separates two good genera in one part 
of a natural order, it must be considered as generic throughout 
the order, it is a very unsafe one, and the attempted strict ad- 
herence to it has been one of the causes of the raising so enormous 
a number of isolated species to bad genera, and of the consequent 
confusion, in Crucifera, Umbellifere, Composite, &c. 
Meliphlea, Zucc., a single Mexican species, has been distinguished 
from Spheralcea by its connate bracts, by the calyx marked inside 
at the base by a smooth five-lobed portion scarcely thickened 
enough to be called a disk, and by clavate styles with the stigmas 
less strictly terminal; but all these characters may be observed, 
although in a much less degree, in other species passing gradually 
into the typical form. It is probable, however, that, when better 
known, the red-flowered species, such as S. umbellata, S. rosea, 
&c., may, with Zuccarini’s Meliphlea, form a good section of 
Spheralcea, whilst Spheroma would constitute a third section. 
URENA, Linn. 
As no character has been found to separate this genus from 
Pavonia except the glochidiate points covering the fruit, the U. 
speciosa, Wall. must be transferred to Pavonia, in which many 
species have connate bracts. 
Pavonta, Cav. 
There are about 60 species known of this genus, varying con- 
siderably in habit and in several minor characters derived chiefly 
from the bracts and the shape and degree of dehiscence of the 
cocci; and it would require a careful monographie examination of 
the whole to determine how far the genus is divisible into good 
sections, and what are the limits to be assigned to it with reference 
to the closely allied genera Urena, Goethea, and Malvaviscus. Le- 
