Erythrococca and Micrococca . 603 
aculeatnm assists us in interpreting the three types of ‘ disc ’ met with in 
the female flowers of Erythrococca. From a condition such as that occa¬ 
sionally seen in Claoxylon Menyharthii , in which the hypogynous bodies 
alternating with the carpels are staminodes which are accompanied by an 
incomplete ring of small free scales, to a condition in which these additional 
scales fuse with the adjacent staminodes to form as many discrete flattened 
hypogynous scales as there are carpels, is only a step. From a condition 
such as that characteristic of Poggeophyton aculeatnm , in which the hypo¬ 
gynous bodies alternating with the carpels are staminodes which are accom¬ 
panied by a complete ring of small free scales, to a condition in which these 
accessory scales fuse with the adjacent staminodes to form as many con¬ 
tiguous flattened hypogynous scales as there are carpels, is again only 
a step. P'rom this latter condition to one where these contiguous glands 
are fused in a complete urceolate ‘ disc ’ is but another step in the same 
direction. That this may have been the history of the evolution of the 
6 disc ’ in Erythrococca is suggested by another consideration. In E. africana 
the disc at first is complete and entire ; in fruit it is distinctly subequally 
5-lobed ; the number of lobes suggests the possibility of some relationship 
to the fact that in the male flowers there are jo stamens in the outermost 
(marginal) series. Another species where there is the suggestion of a similar 
retrogression is E. Molleri ; here the *' disc ’ in the female flower sometimes 
consists of two free scales with contiguous margins, sometimes is complete ; 
when it is complete this disc at first has a uniform, even or 2-lobed margin, 
later on it splits along the edge into a number of crenulate lobules. 
Summary. 
The result of a detailed consideration of the characters exhibited by 
these various groups of species with anther cells that remain erect even 
when empty, is to confirm the conclusion to which Bentham came in 1849: 
Erythrococca , Benth., and Micrococca , Benth., though very closely allied to 
Claoxylon , A. Juss., are easily distinguishable and natural genera. But while 
Bentham’s instinct was in each case right, neither genus is monotypic, as 
he believed them to be ; to Hooker we are indebted both for one of the 
criteria which enable us to distinguish Erythrococca from its two allies, 
and for the demonstration of the fact that Micrococca is wider in its limits 
than Bentham supposed. For the proof of the necessity for an enlargement 
of the limits of Erythrococca we are indebted to Pax. The characters which 
enable us to discriminate the three genera may be summarized as follows:— 
I. Erythrococca, Benth. ampl. Buds perulate. Racemes inter¬ 
rupted or uniformly floriferous. Capsule subglobose or didymous, coria¬ 
ceous ; cocci dehiscing loculicidally: subdivided into :— 
