DICELLANDRA AND PHJEONEURON. 483 
only the two original ones, on which the genus was based, could 
be retained in this species, the others differing considerably 
in the structure of the stamens. I added, however, that this 
difference did not appear to me sufficient to justify the creation 
of a new genus. Gilg did not share my reluctance, and de- 
scribed a new genus, Pheoneuron, from Schweinfurth’s plant 
mentioned above, and from several specimens which the Berlin 
Herbarium had received from the Cameroons, and which he 
considered identical with the former. The character adduced 
by the author to distinguish Phaoneuron from Dicellandra was 
in the structure of the andreecium. This was, according to him, 
homoeandrous (7. e. consisting of one kind of stamens) in Pheo- 
neuron, and heterandrous (7. e. consisting of two conspicuously 
distinct sets) in Dicellandra. In his monograph of the African 
Melastomace he admits three species for Dicellandra, viz. 3 
D. Barteri, Hook. f.; D. setosa, Hook. f.; and D. liberica, Gilg, 
a new species collected by Dinklage in Liberia; and only one 
species for Pheoneuron, namely Ph. dicellandroides, Gilg. 
At the time of my correspondence with Dr. Gilg, Dicellandra 
was represented at Kew only by the original specimens, and those 
mentioned before as referred to D. Barteri by Cogniaux. Since 
then a specimen of D. liberica and a number of specimens of 
Pheoneuron from the Cameroons have been added to the Kew 
Collections. When incorporating them, it struck me that the 
seeds of D. liberica agreed perfectly with those of Pheoneuron, 
but differed entirely from those which I found in a capsule attached 
to the sheet containing Barter’s specimen of D. Barteri. As 
there was, however, no fruit left actually connected with the 
specimen, it was open to doubt whether those seeds really 
belonged to Barter's specimen. There was apparently then little 
chance of deciding this question. I wrote, however, to Gilg and 
asked him to tell me whether the seeds of D. setosa (of which 
specimens are in the Berlin herbarium) agreed with those of 
D. Barteri or of D. liberica, of both of which I sent him a drawing. 
Dr. Gilg was kind enough to communieate seeds of .D. setosa, 
and a leaf, flower, and seeds of a plant collected by Dinklage 
near Gross Batanga, Cameroons (no. 851). The seeds of 
JD. setosa agreed with those of Gilg’s D. liberica and of his 
Pheoneuron ; whilst Dinklage's specimen proved that the seeds in 
the capsule attached to Barter’s specimen in the Kew Herbarium 
must actually have been taken from this specimen, on which the 
genus Dicellandra was partly established. 
