6 MR. J. MIERS ON NAPOLEONA, 
Those botanists who have advocated the amalgamation of the species, place little 
reliance on the colour of the flowers, which they say is variable; but Beauvais, Jussieu, 
Whitfield, and Welwitsch assert that the colours are constant; and this appears con- 
firmed by the dried specimens. In Napoleona imperialis the corolla is of a beautiful 
ultramarine blue, the corona of a roseate colour. In NW. Hewdelotii the corolla is of a 
reddish purple, the corona of a pale rose-colour. In №. Vogelii the corolla is yellow, 
with a tinge of red about the base, the corona being yellowish white. In N. Whitfieldii 
the eorolla is of a uniform apricot-colour for more than half its length, then suddenly 
red towards the base, the corona is yellow, the filaments of a rich crimson colour. In 
N. Mannii the corolla is only half ihe usual size; this, as well as the corona, is of a 
pale yellow, the filaments quite white. In М. cuspidata the corolla in the dried state 
appears of the same colours as in №. Whitfieldii, but the corona is of a deeper roseate 
hue. In М. angolensis the corolla is white, the corona is white, but the campanular 
base of the 2nd whorl is of a violet apricot-colour, and the staminal whorl of the 
same hue. 
The fruit varies in size in the several species. In М. imperialis it is said to be spherical, 
with a softish pericarp about the size of a small apple, with many seeds imbedded in 
fleshy pulp. In №. Heudeloti it is figured as spherical, has а soft coriaceous pericarp 
1 or 2 lines in thickness, with 5 thin persistent dissepiments, each cell containing 4 
seeds in superposed pairs, of a shape very different from all others. In N. Vogelii the 
fruit is globular, somewhat depressed, obsoletely 5-lobed in the periphery; the pericarp 
is 24 inches in diameter, and 12 line in thickness, externally of a dark maroon 
colour, with pale spots, is about 10-celled, contains about 12 seeds, some of them slant- 
ingly superposed; and there is no appearance of pulp in the specimen preserved in 
spirits. In N. Whitfieldii the fruit is 24 in. in diameter, of a more depressed globular 
form, with a pericarp not thicker than a common card, of a pale rose-colour, with white 
spots, is very brittle when dry: it has apparently been 5-celled ; but no remains of disse- 
piments were visible in the specimen in the British Museum. Whitfield says that the 
seeds are enveloped in an edible pulp. In М. angolensis the fruit much resembles that 
of the preceding in size and colour, but the pericarp is somewhat stouter, more firm, 
obsoletely 5-lobed on the periphery; in the dry state it seems to have been 5-celled, and 
the seeds have probably been enveloped in pulp. 
These many differential characters, though long disregarded, are positive, and fully 
justify the maintenance of several good species. Up to 1845, indeed, three of them 
were acknowledged, when in that year Dr. Lindley proposed a fourth, upon Whitfield’s 
specimens. In 1848, however, Sir Wm. Hooker, in figuring and describing the latter 
from a living plant, pronounced decidedly that it did not differ specifically from the 
original species, .N. imperialis, or from Heudelot’s plant; for he could discern no dissimi- 
larity between the three plants, except that of colour, which was due solely to their fading 
in drying. This broad assertion so puzzled Van Houtte, that he wrote immediately to 
Prof. Decaisne, asking him to examine the typical specimens of Beauvais and Heudelot’s 
plants in the Paris collections, in order to determine this point. The learned Professor ' 
Flore des Serres, vol. iv., sub tab. 386, 387 (1848). 
