ORCHIDACEM 77 



chilum and Platyclinis, respectively. Pfitzer maintained this divi- 

 sion. Later, J. J. Smith, in his review of the genus published in 

 1904, combined them, on the ground that lateral or terminal posi- 

 tion of the inflorescence could not stand as a generic character. 

 When it was learned that the obscure genus Acoridium, which 

 its authors, Nees & Meyen, assigned to the Philydraceae, is con- 

 generic with Platyclinis, Mr. R. A. Rolfe, with the understanding 

 that Dendrochilum and Platyclinis are distinct, renamed some 

 thirty or more species of the latter genus, placing them in Acori- 

 dium, — which name by the rule of priority displaced Platyclinis. 

 Viewed from every point, the generic limits of Blume's Dendro- 

 chilum cannot be fixed with surety until a more exhaustive study 

 of the species now known has been made than has heretofore been 

 attempted. The Philippine species form a remarkably well-con- 

 nected series, both as to vegetative and as to floral parts, and 

 form a natural group which may for convenience be divided into 

 three sections ; namely, Eudendrochilum, with laterally produced 

 inflorescence, Platyclinis, with terminal inflorescence and winged 

 or toothed column, and Acoridium, with terminal inflorescence 

 and wingless column. The last of these sections is perhaps the 

 most remarkable for diversity of habit and floral conformation. 

 With few exceptions, the middle lobe of the labellum is smaller 

 than the lateral lobes, and the plants are for the most part of 

 dwarf habit. Although in my studies of the group I have been 

 inclined to adopt the treatment of Bentham and of Pfitzer, recent 

 investigations have led me to believe that the most rational treat- 

 ment is that advocated by J. J. Smith. 



The section Dendrochilum is placed near Bulbophyllum in Pfit- 

 zer's system of classification, thus being far removed from the 

 section Platyclinis, which is placed near Ccelogyne. As the lateral 

 inflorescence of Eudendrochilum by its rarity seems to be an ano- 

 malous characteristic, as the terminal inflorescence of Platyclinis 

 on the other hand is represented by many species, and seems there- 

 fore to be a normal characteristic, and as the floral character of 

 both sections indicates close affinity, not only with each other, but 

 with the Ccelogyninae, I have assigned the genus as now consti- 

 tuted to the Coelogyninse rather than to the Bulbophyllinae. 



