PREFACE 



drawn and in a measure present original data. Although the part 

 devoted to geographical distribution may be thought unneces- 

 sarily full it is of value from two points of view, — it shows the 

 actual distribution of the species as represented by the great 

 herbaria of the country, and serves as a guide to the material 

 on which the monograph depends. The names of the species are 

 not mere transcripts from herbarium labels, but are the result 

 of an actual identification of the specimens cited. In many cases 

 where these were poorly preserved a flower was soaked off for 

 critical examination. Otherwise identifications were made by 

 careful comparisons with type or authentic material, and are 

 reasonably correct. Although great care has been exercised in 

 copying geographical data from labels, it has proved an extremely 

 difficult task to verify the spelling or exact location of obscure 

 tow^ns, rivers, lakes and mountains, and errors in the geographi- 

 cal lists may have arisen through this difficulty or through un- 

 decipherable writing. It has not proved practicable to compare 

 the proof with the labels, and herein may lie another source of 

 error in spelling and in citation of numbers. 



The illustrations have been prepared with careful attention 

 to accuracy and the floral parts drawn with the aid of the ca- 

 mera lucida. With few exceptions those species have been 

 chosen for illustration which have never been figured before, and 

 in almost every case type or authentic material has been used. 

 In conjunction with the key and the original descriptions the 

 illustrations form the most useful part of the w^ork, and should 

 remove many of the difficulties which attend the identification 

 of obscure or variable species. 



Nomenclatorial problems have been troublesome, although 

 not to such an extent as is experienced in other groups of the 

 Orchidaceee. Absolute priority is a rational basis from which to 



[xii] 



