PHILIPPINE URTICACEAE, II. 309 
separated with ease by leaf-ehanu-ters. Indeed, hy forcing' cbaiadors wliicli aie 
ordinarily stable, four species coxild be deliminated. The nature of the material 
itself shows that such a proeeediufr \vould be most inadvisable. The linal 
difficulty conies with Merrill 7030. lO.'/O, collected some (lays apart from nearly 
the same locality on Blount Tonglon. Some of the plants an- as to leaves 
typical K, mriahile, others arc equally typical E. carinoi, some few iiavc the 
upper leaves of the former typo, but the lower (^nes api)roachin<: tho^o of the 
latter. Both of these have staminate receptixcles. On 7C30, these are sessile 
except on one plant, where they are very short, while on "tG-'fO they arc shortly 
peduncled or sessile, with no difference in this or any other respect between 
those in the axils of the two types of leaves. E. carwoi and E. variahiJc are 
therefore very closely allied, liavinnf identical receptacles with the same rau^^e 
of variation in their attaebinent, nor can any snflicient means of se]mralion 
be found in stipules or pubescence. There remains in doubt the question of 
leaf-outline, with which the venation is inseparably connected. I1iis is not 
simplv due to difference in habitat. For example, the two forms were found 
together in the two last collections cited, while my two collections, Ix.th per- 
fectly typical E. carinoi. came from quite different conditions, one growing 
in muddy soil along the edge of a ditch among grasses not" tall eiiough to 
overshadow it, the other was on rock, exposed to the spray of a small waterfall, 
within a small cavern where it was sheltered from direct sunlight except in 
the early m(n-.iing. The link between the species is in the resemblance between 
lower leaves of E, variahUc and n<n-mal ones of E. carinoi: it is not yet quite 
eomplete. 
Elatostrwa cnn„oi is in onltivation at Camp Julm Hay, iR-ar Bagnio, growni? 
luxuiiuntly on rook.s ?n tho .pray of a small fountain, the only known instance 
where a species of tlie genus has been so utilized in the Philippines. 
From rcforenees to pubescence above and its frequent use in my keys an.l 
elsewliere, one Avithout access to the material might suspect that spoc.es Mere 
being segregated with this as the uiain character. It is therefore advisable to 
state that it has been found to be very constant, especially upon the stems, and 
has been ofTered a8 a guide only in the cases where not only this is so but other 
characters are present, less easily or at any rat. less briefly described. 0, e 
leaves, it is J so constant. It does not h,.ld with ,, la,u„r„.e. -'-«■;<•; 
large clump of plants from which the type and other collections ";;-;^J""-^;; 
varies from one individual to another in general c.rre ation "'' ' ^^^ J" 
of spray received from the Los Banos waterfall, but this is a very exceptional 
case. 
10. BOEHMERIA Jacq. 
2a. Boehmeria beyeri sp. nov. 
Arbu.cula raniosa: dioicn, glome.uli, axilkribus, --i'^'"';' f^t 
staminiforis tetramori.; peviantluo pistillifero ^<--Vre^^'^^^. 
apioc cot.tracto, 4-dentato, e.xtus substrigoso ovant™ niuUo .up .^ - 
filiis alternis lei raris.i.ue oppo.itis, -If e niae.puaiagn. h^ ^nu 
baud valde reductis, aliis breviter petio at.s, <^^^'^''^'^'^X 
petiolatiB, ellipticLs apice acute aeuminat.s, margmo dentafs, tnncr.n. 
vol altero minus oonspicuo interjecto. 
DioeoiouB: stannnate gloiuorulos ax.nnry sessile, " 
eter; bvacteoles elliptic, 2.5 mm long, coru.culato. .sl.gbtl, pdo.c 
103750 
