THE GENUS EBIA IX THE PHILIPPINE ISLANDS. 225 



cm longae. Pedunculus circ. 18 cm longus, bracteis paucis adpressis. 

 Bracteae refiexae, lineari-laneeolatae, acutae, circ. 0.40 em longae. Flares 

 patentee, circ. 0.80 cm longi. Sepalum dorsah oblongum, obtusum, 

 valde concavum, Snervium, circ. 0.45 cm longum, 0.25 cm latum, extus 

 disperse pilosum. Sepala lateralia ad pedem gynostemii decurrentia, 

 mentuni magnum, latum, saccatum, obtusum formantia, late oblique 

 ovato-2angularia, obtusa, 5 nervia, circ. 0.40 cm longa, 0.50 cm lata, 

 extua disperse pilosa. Petala oblonga, obtusa, 3nervia, circ. 0.46 cm 

 longa, 0.23 cm lata. Labellum erectum, concavo-cymbiforme, basi fere 

 saccatum, apice breviter fere aequaliter 31obatum, intus costis 3 alaeform- 

 ibus, media duplici apicem versus evanescenti, lateralibus ad basim lobi 

 tnedii in lobulum brevem, obtusum, liberum terminantibus, expansum 

 fere quadrangulum, circ. 0.47 cm longum, 0.425 cm latum; lobi hi I cm I ex 

 anguli, obtusi, convexi, vix 1.10 cm longi; lobus median paulo brevior, 

 late 3angulus, obtusus, coneavus, marginibus reflexis. Gynostemium 

 breve. Stigma magnum, profundum, rotundum. Pes gynostemii cum 

 ovario angulum acutum, fere rectum formans, rectum, circ. <>.:>(> cm 

 longum. Ovarium pedicellatum circ. 0.60 cm longum." Smith, /. c. 



LUZON, Province of Jiataan, .Mount Mariveles. Merrill $757, January 1, 1904. 

 \|[.\i>oiu>, Mount Haloon, Merrill 5847, November 26, 1!)0(>. epiphytic in ravine 

 forest, altitude 1,200 m. 



SYNOPSIS OF SPECIES. 



In a large genus tbe forming of distinct sections is often difficult. 

 Yet with a local treatment, the number of species being limited and 

 the confusing species absent, the matter is arranged much more readily. 

 This is but saying that the rule for comprehensive treatises may be 

 different from the rule for provincial floras. It is permitted to draw 

 lines rather more sharply in the latter than in the former. And in fact 

 it is necessary, if one is to be guided by practical rather than theoretical 

 considerations, to draw them somewhat differently in the local case. 

 With the definitions of sections in the present conspectus I have had 

 regard to the Philippine species alone, feeling free to modify, for the 

 convenience of the student of the Philippine flora in especial, the form 

 given the groups by their authors. 



As to the arrangement of the sections, it is obvious that no linear 

 arrangement will represent the natural affinities, llymeneria (in a 

 restricted sense) seems to me to be the most generalized, or central form, 

 introduced perhaps by Urostachya — with which it merges — and giving off 

 directly or indirectly Aeridostachya and Dendrolirium. Mycaranthes, 

 Cylindrolobus, and Trichotosia are evidently allied, but the exact con- 

 nection with the rest of the genus is not clear. 



-.Measurements in the following descriptions are taken from dried specimens, 

 and when single measurements are given they are the maxima rather than the 

 averages. Lengths of leaves are taken from the disarticulation and do not include 

 sheathing hases. Length of racemes are exclusive of the peduncle. The height 



