A NEW GENUS OF ACANTHACEJE. 7 
the whole group, which is fairly well represented at Kew. Through the 
courtesy of Professor Radlkofer I had, further, an opportunity of seeing the 
originals of Hallier’s Ptyssiglottis so far as they are preserved in the Munich 
Herbarium, whilst the late M. Drake del Castillo sent me a fragment of 
Zollinger’s original of Rostellularia sarmentosa, of which, moreover, an 
excellent duplicate specimen was subsequently discovered by Mr. C. B. Clarke 
in the Kew Herbarium, where it had been placed in Justicia. 
Rostellularia sarmentosa was easily recognized as a species of Rungia, and 
named R. sarmentosa by the late С. B. Clarke. All the Borneo plants described 
by Hallier f. as species of Ptyssiglottis were found to be closely connected 
members of a very homogeneous group, which, for reasons stated below, could 
not well be retained in Ptyssiglottis. On the other hand, the only non-Borneo 
Ptyssiglottis described by Hallier, P. picta, a native of North-east Sumatra, 
exhibited differences in the pollen, which suggested other affinities. The 
material of this species in the Munich Herbarium was, however, too scanty to 
warrant a more definite statement. But when last year Mr. C. B. Clarke 
examined the Calcutta set of Acanthaces from the Malayan Peninsula, he 
found among them several species which he comprised under the new genus 
Polytrema, and I have no doubt that Ptyssiglottis picta will have to be referred 
to it. If we exclude therefore Ptyssiglottis radicosa and P. picta, all of 
Hallier's Ptyssiglottis appear confined to Borneo, and more especially to the 
Kapuas Basin of West and Central Borneo. That area is considerably 
extended through the accession of not less than ten additional species from 
Sarawak, including Brunei and British North Borneo, and one species from 
Mindanao. For this group, which thus includes eight species described by 
Hallier and eleven to be described below, I propose the name Hallieracantha. 
Compared with many other genera of Acanthacez, it has the merit of being 
very homogeneous and also of. having a continuous and uniform distribution 
area. It is more difficult to define its exact position in the family. Its 
relationship with Polytrema is manifest, and botanists who prefer larger 
genera with subgenera as subdivisions may feel inclined to reduce Polytrema 
to a subgenus of Hallieracantha, as the two genera differ only in the shape of 
the corolla, the attachment of the anther-cells on the connective, and the 
structure of the pollen. Оп the other hand, the relationship of Hallieracantha 
and Ptyssiglottis seems to be much more remote. Ptyssiglottes has been 
plaeed by Lindau' near Codonticanthus in Pseuderanthemee, and I would 
leave it there for the present; whilst НаШетасат а seems to have more in 
common with the Justiciee (Lindau) than with the Pseuderanthemeew ; and 
Bentham?, who placed two of Motley's plants which I deseribe below as 
Hallieracantha Motleyi and Н. laxa in Dianthera, was not so far wrong. А 
closer examination of the very heterogeneous genus Justicia (heterogeneous 
1 In Engl. & Prantl, Natürl. Pflanzenf. vol. iv. З b (1895) p. 33 
2 In Bentham & Hook. f. Gen. Plant. vol. ii, p. 1114. 
