550 PROCEEDINGS OF SECTION D. 
manipulation, as is required if we employ the anthereal system or 
that based on the histology of the petiole, or that on the chemical 
properties of the kinos and oils. 
Bentham, “Fl. Austr.,” ii, made only a minor use of the fruit 
character, but Baron von Mueller, ‘‘ Syst. Victorian Plants,” gave 
it a higher value. In my “Flora of South Australia” (1890) the 
thirty-three species inhabiting the province are primarily arranged, 
on the form of the fruit, into four sections, but the number of the 
divisions will admit of increase when dealing with the genus in its 
entirety. 
I would now review the nature and value of the component 
elements embraced by a carpological scheme of classification :— 
1. Shape of Fruit.—The shape to be described is that of a fully- 
ripe specimen, as immature states may prove delusive when testing 
the carpological system. In this connection Bentham, “ Fl. Aust.,” 
ili, p. 187, writes :—“ It (the calyx-tube) often alters so much that 
it neither indicates the form it had in flower nor yet that which it 
will assume in the fruit.” Thus in Z. pyriformis the calyx-tube, 
on the fall of the operculum, is obconic, with a horizontal summit, 
but in the adult state it becomes biconic. Again, #. cosmophylla 
ranges from ovoid-conic in the early stages to hemispheric when 
mature. 
The calycine portion of the fruit may extend beyond the cap- 
sular portion to varying heights ; its rim or margin may be acute, 
as in £. Foelscheana, or it may be of varying width, remain hori- 
zontal, as in L.goniocalyx,or become ascending with a convex slope, 
as in LZ. capitellata, or with a concave slope, as in #. longifolia, 
EL. pyriformis, &e. 
In the appended carpological schedule I have set forth the 
leading geometric forms assumed by the Eucalyptine fruits, so 
there is no need to describe them in this place ; but a fundamental 
investigation is that of the persistency of shape for each species. 
From my own experience, the shape of the fruit is constant within 
natural and reasonable limits. Thus #. Foelscheana is usually 
globosely urn-shaped ; but by contraction at the summit becomes 
globosely-oval—a natural transition. . capitellata is usually 
biconic, but may become roundly depressed atop, thus passing to 
globulose or ovoid-conic. I do not find any variation of shape 
that does violence to a geometric development. If we except 
£. incrassata, following Mueller’s interpretation of the species, here 
there is undoubted geometric discordance ; but the disassociation 
of ZL. dumosa, as adopted by Bentham, at once removes the 
anomaly ; indeed, the difference of fruit in conjunction with 
that of habit justify specific distinctions between them. 
2. External Sculpture and Ornament.—Though the outer sur- 
face of the fruit is usually smooth, with a more or less circular 
