PREFACE 



tention to cultivated orchids must realize that while slight dis- 

 tinctions are very serviceable for horticultural uses, they are im- 

 practicable for the purposes of systematic botany. Undoubtedly 

 many of these horticultural forms will breed true to type and 

 thus act like species. But should they be given specific rank the 

 orchid flora of the world would be increased tenfold. 



Several Habenarias are characterized by a strong tendency to 

 produce variations. The temptation to recognize these variations 

 as distinct species may be strong, but I think the best interests 

 of the science are subserved by regarding them as the compo- 

 nents of compound species. Slightly modified, the Linnasan con- 

 ception of a species forms a convenient basis for systematic work. 

 For economic purposes the subspecies is necessary. When it is 

 propagated from pure cultures and carefully labelled, it is of 

 undoubted value. 



The greatest difficulties in the present monograph were ex- 

 perienced among recent segregates. In my treatment of these I 

 may have reduced some which further studies of fresh material 

 will reinstate, but at present they appear to be exceptional forms 

 unworthy of specific rank. In several of these segregates the dis- 

 tinguishing characteristics may be found in the flowers of one 

 inflorescence, so that a single specimen will exhibit the floral 

 peculiarities of two described species. Then again, the vegetative 

 characters in the group to which Hahenaria hyperborea and 

 H. saccata belong are extremely variable and unsafe criteria for 

 specific distinction. 



In 1904 I made special efforts to obtain in a fresh state, from 

 several localities, large collections of H. hyperhorea and H. dila- 

 tata. The specimens were taken from different kinds of soil, 

 from sunny and shady places, and the range of variation was re- 

 presented as completely as possible. The conclusions based on 



[ viii ] 



