Erythrococca and Micrococca. 591 
character included in this which is not also applicable to Claoxylon , is that 
Erythrococca has stipular thorns. 1 
When providing this new criterion for the genus Pax added a fifth 
species, E. mitts, which is certainly an Adenoclaoxylon, and is admittedly 
very nearly allied to Claoxylon Kirkii itself. More recently Rendle has 
added another Adenoclaoxylon to Erythrococca , E. Paxii , Rendle (Journ. 
Linn. Soc. Rot., xxxvii. 213), as nearly allied to E. Fischeri as E. mitis is to 
Claoxylon Kirkii ; while Pax has described yet another, E. rigidifolia , Pax 
(Engl. Rot. Jahrb., xliii. 320), as nearly allied to E. bongensis as E. Paxii 
is to E. Fischeri . Every known Adenoclaoxylon has now been formally 
referred to Erythrococca except Claoxylon Kirkii, Mu ell. arg., the species 
on which the section was based. 
Along with E. rigidifolia ,. Pax added to Erythrococca a species, E. hirta , 
Pax (Engl, Rot. Jahrb., xliii. 321), which is neither an Adenoclaoxylon nor 
a Defersia, but is nearly allied to Claoxylon trichogyne , Muell. arg., one of 
the species added to Athroandra by Mueller which do not, on account of 
their laciniate stigmas, conform with that section as it was originally out¬ 
lined. To this group belong, in addition to E. hirta and C. trichogyne , the 
following species described under Claoxylon :— C. triste , Muell. arg.; C. lasio- 
coccum , Pax (Engl. Rot. Jahrb., xix. 87); C. Menyharthii , Pax (Rull. Herb. 
Roiss., ser. 2, i. 877); £ 7 . Mildbraedii , Pax (Engl. Rot, Jahrb., xliii. 80). All 
six agree in having stipular thorns, densely hirsute ovaries, and plumosely 
laciniate stigmas; where one of them is definitely placed the others must 
go. If Erythrococca be a valid genus, the reference thereto of E. hirta is 
as justifiable as was the reference to that genus of E. abyssinica ; although 
the other five species named have not yet been formally included in Erythro¬ 
cocca, this does not lessen our indebtedness to Pax for having so enlarged 
our conception of the genus as to render it capable of accommodating any 
species, hitherto accounted for under Claoxylon, which combines the charac¬ 
ters of perulate buds, stipular thorns, and plumosely laciniate stigmas. 
The question that has first to be settled, therefore, is as to whether 
Erythrococca be really distinct from Claoxylon. The affinity between the 
two genera has been fully admitted since it was pointed out by A, Jussieu 
in 1824; the settlement of the question involves a scrutiny of the criteria 
which have from time to time been relied upon for their discrimination. 
These criteria have been three in number(1) that formulated by Mueller 
in 1866 ; Erythrococca has an indehiscent, while Claoxylon has a capsular 
fruit: (2) that employed by Pax in 1890; Erythrococca has 3-6 stamens 
and plumosely laciniate stigmas, while Claoxylon has usually many stamens 
and always entire stigmas: (3) that substituted by Pax in 1895; Erythro¬ 
cocca has stipular thorns, while, by implication, Claoxylon is unarmed. 
1 It is not clear that this was the criterion relied on in 1894, because C. lasiococcum is one of the 
species with stipular thorns. 
R r 2 
