296 
ROBINSON. 
only to the pistillate. Only one receptacle is present on the Sarawak Museum 
material, but it is staminate, borne upon a peduncle 3 mm long, pubescent like 
the stem, slender in contrast to the general thickened nature of the stems and 
leaves. The bracts are cornieulate, the flowers tetramerous. All this is in line 
with Stapf’s suggestion, except that other Philippine species come closer to 
E. bulbothrix than does E. glaucescens, although the texture of the leaves would 
keep it distinct from any of them even if there were no other characters. Probably 
its nearest allies are E. heterophyllum 0. B. Rob., whose leaves have quite a 
different outline and margin, and E. obovatum Wedd. which has much larger and 
differently shaped leaves. 
Elatostema lineare Stapf 1. c. 228. 
Mount Kinabalu, Kinitaki River, at 5,000 fee tv elevation, Haviland 1271, type 
collection. This at once suggests E. pinnatinervium Elmer, by reason of the 
pinnate leaf-venation, but the leaf-outline is very different, and the staminate 
receptacles of E. pinnatinervium are peduncled while those of E. lineare are 
sessile. The leaf resemblance is much closer with an undetermined specimen 
from Polillo, Bur. Sci. 9195, but the leaf -teeth of the latter are more deeply 
cut, and the stipules of the Bornean plant are longer than those of either the 
Polillo plant or of E. pinnatinervium. 
Elatostema lithoneurum Stapf 1. e. 230, PI. XIX, fig. C, 9-16. 
Mount Kinabalu, at 11,000 feet elevation, Haviland 1206, type collection. The 
specimen has much external resemblance to E. palaioanense C. B. Rob. The 
leaves are similar in outline and margin, but wider on the Philippine plant, 
whose narrowest are of about the width of the widest of the Bornean, they are 
also more definitely oblanceolate or obovate than the .latter. Further, the upper 
surface of the leaves of E. palaioanense is scabrous, while the pubescence of E. 
lithoneurum is much more scattered, the hairs somewhat longer, the surface not 
at all scabrous. The under surface of the Palawan leaves is also much more 
densely pubescent, with this pubescence glaucescent, much paler than on the 
Bornean plant. The lateral veins of the latter leave the costa at a smaller angle. 
Finally, the staminate receptacles of E. lithoneurum have rather long and slender 
peduncles, but some young ones in the upper axils are sessile, and this was the 
state in which the Palawan plant was found. 
Several of these differences bring E. lithoneurum close to E. lagunense Merr., 
the nearest Philippine alliance of E. palaioanense, but that is a much coarser 
plant, with the staminate receptacles sessile or on short stout peduncles, the 
leaf-teeth are coarser, and the terminal acumen much more prolonged. 
The series of E. lithoneurum, E. palaioanense , E. lagunense, links through the 
last with E. paludosum Miq., as represented by Hallier 396, Tjibodas, Java, this 
species in its turn having been reduced by Weddell 5 to the status of a variety 
of E. macrophyllum Brongn. The ease between E. lagunense and E. paludosum 
is very critical indeed, but the former has much more continuous leaf-nerves, 
the leaves themselves are larger, and the stipules somewhat longer. It is diffi- 
cult to decide positively upon a single collection, and our specimen of the 
Javan plant is sterile, but the two species may prove to be incapable of 
separation. Between E. paludosum and E. macrophyllum, the differences are 
quite sufficient for them to be held apart specifically, if the .former is adequately 
represented by the specimen above cited, and the latter by Hallier 371f, Buitenzorg, 
Java. E. macrophyllum is described as penninerved, which agrees well with the 
specimen, but the same can not be said of Hallier 396, as the basal nerves are 
5 DC. Prodr. 16 1 (1869) 174. 
