514 
ROBINSON. 
the opposite for the pistillate. They have further completely linked together 
what had seemed to be two distinct species. A further possible result of Ramos’ 
find is that this may solve the puzzle of Dorstenia pubescent Blanco, as Malapad- 
nabato is only separated from Pasig by the river: the habit stated by Blanco is 
still an obstacle. 
The species approaches the polymorphic Indian E. surculosum Wight, many of 
whose variations it parallels. But the subopposed reduced leaves so characteristic 
of that species are very rare in E. luzonense, being found on about 5 per cent of 
perhaps 400 plants examined for the purpose, and even in these it was normal 
on three stems ( not plants ) only, on the others confined to one or two nodes : 
moreover the leaves are more often not glabrous, although this is best seen on 
fresh material. As E. luzonense seemed to come even closer to E. sikkimense . 
Clarke, material then considered very representative was sent to Calcutta, and 
Mr. W. W. Smith, to whom I am greatly indebted for this and much other 
valuable assistance, considered it sufficiently distinct, E. sikkimense being two to 
three times larger, with the leaves larger, quite glabrous, and much more dis- 
tinctly acuminate, and with very long peduncles. The more recent collections 
have shown E. luzonense to be still nearer to E. sikkimense than originally sup- 
posed, but I still consider it distinct. 
A collection not above cited, Bur. Sci. 1095 Ramos, Bosoboso, Rizal, has the 
stems densely substrigose, but has no staminate receptacles, so that -a definite 
opinion is deferred : so far there seems nothing but the pubescence to prevent its 
inclusion in E. luzonense. • E. lanaense is very similar in habit, but differs in 
several ways. ■ 
2. Elatostema variabile sp. nov. (Plate I, Yol. VI.) 
Receptaculis pistilliferis sessilibus, mediocribus, brapteis marginem 
versus solum liberis, triangulari-lanceolatis vel triangulari-ovatis, corni- 
culatis vel aeumiuatis; floribus typicis: foliis membranaceis, difformibus, 
saepissime anguste oblohgo-lanceolatis, sed etiam oblanceolatis vel oblongo- 
ovatis, margine integris, vel irregulariter obscureque undulatis, vel 
obscure vel subgrosse dentatis, saepe basim versus lobatis, subpinnatinerviis 
vel triplinerviis. 
Pistillate receptacles sessile, attaining 1 cm in diameter, the bracts fused 
except toward the margins, the outer nearly similar to the others, trian- 
gular-ovate to triangular-lanceolate, free for about 2 -mm, ciliate on the 
margins, bracteoles, linear-oblanceolate, 2 mm long, densely ciliate; 
perianth minute, 3-lobed; achene ellipsoid, about 0.8 mm long, about 
8-striate. 
Plants erect except at base, 20 to 50 cm high, the vegetative parts 
glabrous : leaves subsessile, the lamina when dry thinly membranaceous, 
from 10 cm long and 15 him wide to 2.5 cm by 13 mm, extremely 
variable, most often narrowly oblong-lanceolate, but also narrowly lance- 
olate, oblong-ovate, or obovate, the costa straight or curved, the base 
distinctly ineqifilateral, farther produced on the wider side, rounded, 
obtuse, or subobtuse, the margins entire, or obscurely and irregularly 
wavy, shallowly or for the size of the leaf coarsely serrate, the teeth 
numbering from 1 to at. least 10, the distance of the lowest tooth from 
