vi PREFACE. 



descriptions in which work I cannot speak in terms of too high praise. 

 By this method of re-description, whilst I believe I have avoided some 

 errors of my predecessors, I have no doubt made others of my own : such 

 creep into all endeavours to describe most or all of the organs of many 

 species: and if I have made many such blunders, a part may be attri- 

 butable to the fact that various Genera were described amidst constant 

 interruptions, and all under pressure of official duties. 



The adoption of sub-species will, I hope, prove generally acceptable, 

 though open to abundant criticism in details; indeed, in a good many 

 cases this subdivision is almost purely arbitrary. In defining the sub- 

 species I have been much guided by the views held by British Botanist s, 

 and especially as expressed in the London Catalogue. 



The terminology employed is intended to be as simple as I think is 

 attainable with a due regard to precision of language. As a rule, I have 

 been guided in the choice of terms by Oliver's " Lessons in Elementary 

 Botany," and I have especially avoided terras applied to single Orders, 

 or of special signification in single Orders or Genera. It is not easy to 

 be consistent in this matter, in a work extending over so many different 

 Orders, and occupying so many months in its preparation ; and I fear that 

 I may have at times forgotten my own principles. For modifications of the 

 fruit the choice of terms presents great difficulty ; and I have therefore 

 very much confined myself to such as are required to avoid periphrasis, 

 as capsule, drupe, berry, utricle, follicle, pod, &c. (about which there is 

 no ambiguity), and to achene for the dry indehiscent carpels of an apocar- 

 pous fruit. For Grasses, Composites, &c. the term fruit is itself sufficiently 

 explicit, its nature being always explained in the Ordinal description. 

 The term nutlet for the parts of the fruit of Borraginea: and Labiatee 

 1 have borrowed from Asa Gray. 



The Keys to the Genera are naturally arranged, but in Umbellifera 

 I have added an artificial key, as essential for the determination of 

 a Genus before the whole Order has been studied. I have added no 

 keys to the Species, preferring curt diagnoses which embrace the more 



