PREFACE. Vll 



possible. To Dr. Gray is due, not only the thought which 

 grew into this book, but also a constant encouragement and 

 patient criticism which have developed anything of merit it 

 possesses. Mr. Watson has also responded generously to ev- 

 ery demand made upon him : while to Messrs. M. S. Bebb and 

 L. H. Bailey, Jr. is due the relief of some original work, the 

 former being our well-known authority in the difficult genus 

 Salix, and the latter an ardent and most successful student of 

 the perplexing genus Car ex. At the time of his death, Dr. 

 George Engelmann had in preparation the groups with which 

 his name is so closely connected, and their presentation shows 

 the lack of his master hand. 



In general, the ordinal sequence adopted by Bentham and 

 Hooker's Genera Plantarum has been followed, but Gymno- 

 sperms have been transferred to the end of Phsenogams, 

 and Monocotyledons and Dicotyledons subordinated to Angio- 

 sperms. This change has been made simply because it better 

 expresses relationships which have long been recognized. The 

 term "Cryptogam" has been discarded as the correlative of 

 Phaenogam, and Pteridophyta (Vascular Cryptogams) is used 

 as the nama of the second great series of plants. The orders 

 and ordinal sequence of the Pteridophyta are thought best to 

 express their relationships. 



Introduced plants are placed in foot-notes, that they may be 

 separated as far as possible from our native plants, and their 

 relation to the flora thus emphasized. 



To save space, there is no attempt to give any but the most 

 important references and synonymy, while geographical range 

 is reduced to its lowest terms, and collectors' names almost 

 entirely omitted. For the most part no characters have been 

 repeated, and the student is warned that generic characters 

 especially must be sought for through analytical keys. The 

 professional botanist will note a glaring inconsistency in this 

 respect, the genera of some families being grouped by means 

 of a few very salient characters, while those of others are 

 presented with almost full descriptions, only certain supple- 

 mentary statements being left to head the descriptions of 



