CLASSIFICATION OF THE EUCALYPTS. 549 
ANTHERS. 
Bentham, in “FI. Aust., 1866,” arranged the species into five 
ptimary sections according to the shape and dehiscence of the 
anthers, thus :— 
1. Renantherze.—Anthers reniform, large ; cells divergent at 
the base. 
2. Heterostemones.—Outer stamens sterile; anthers globular, 
small, opening by pores or slits. 
3. Porantherze.—Anthers globular, smal], opening by circular 
pores. 
4. Micrantherze.—Anthers globular, small, opening by lateral 
slits. 
5. Normales.—Anthers globular, small, the cells distinct and 
parallel, opening by longitudinal slits. 
Mueller, in his “Zucalyptographia,” adopted the anthereal system 
of Bentham, but reduced the sections to four, and in the second 
edition (1889) of the “Census” three only are retained, viz. :— 
(1) Renantherz, (2) Poranthere, and (3) Parallelanthere. 
The species of Series 2 of Bentham are mostly transferred to 
Porantherze. The elimitation of the characters dependent on the 
occurrence of anantherous stamens simplified the classification, 
and was desirable because the tendency to sterility of the exterior 
stamens belongs to all Eucalypts as to other polyandrous flowers. 
The presence or absence of staminodia may, therefore, be a feature 
of each individual flower. 
The species of series 4 and 5 are comprised under Parallel- 
anthere. 
Though it must be confessed that the anthereal system of clas- 
sification brings into juxtaposition species allied in other respects, 
more so, perhaps, than any other system, yet persistency of anther- 
type is not always a certainty, and the manner of determination 
is a bar to its employment for primary classification. Moreover, 
the species are unequally distributed in the three sections. Thus, 
in the 1889 edition of the “Syst. Census Austr. Plants,” series 1 
contains 23 species, series 2 contains 14 species, and series 3 con- 
tains 96 species. 
FRUIT. 
There is obviously the need of an organ which exhibits greater 
diversity of form and structure and admits of a greater number 
-of combinations than is afforded by the anther, or indeed any 
single structure as yet considered. The requirements seem to me 
to be best fulfilled by the fruit offering, for the most part, macro- 
scopic characters, and the special advantage that it is nearly always 
possible to obtain them, whilst the flowering season is of limited 
duration and is not always of annual recurrence. At the same 
time, the characteristics are readily interpreted, needing no special 
