KTKODUCTIOy. 111 



charactere and affinities, Works wliich contaiu full descriptions aud iUus- 

 trations of plants of the regions in question, should present impoiiant 

 materials for students of phytogeogi"aphy and of systematic botany. Works 

 which deal onlj with the names of plants, perhaps en'oneously classified, 

 as would probably be the case in cLrcumstauees hke om-s, ■would be 

 likely to lead studeuts of phytogeogi'aphy to eiToneous results, and would 

 puzzle students of systematic botany -n-ith an endless confusiou of uomen- 

 clature of plauts. 



In this work, plauts, the identification of which has been satisfactoiily 

 made, are refen-ed to known species. lu tliis case, I usuaUy have made it 

 clear how I have identified them with the species to which they liave been 

 referred, — by means of a comparisou of herbaiiiun specimeus, of figm^es or of 

 descriptions foimd in Iiteratiu'e. But in cases in which ideutification has 

 not been satisfactoiily made, plants have teen described as new witli aflinities. 

 In some cases, my species may be the same as those which I have given 

 as their aUies imder descriptions of plants in question. Therefore, though 

 many of my new species should be reduced to synonyms, still I have little 

 donbt that my work will present important material to students of the floi-a 

 of Eastem coimtiies. No one is mox-e conscious than I that eiToi-s may 

 have crept into my work ; but the errors in most cases are of a nature 

 inevitable iu work of this kind done in the East, far from Westem herbaiia, 

 and wiU doubtless easily detected by those who are famihar with the species, 

 when they see merely my descriptious and figm'es without troubUng them- 

 selves to look at my original tv^pes. 



As stated above, notliiug is more difficult thau to identifj' plants by 

 short descriptions unaccompanied with figiu^es. Original descriptions ought, 

 therefore, to be veiy clear, accm^ate aud exhaustive, and if possible, oiight 

 to be supplemented with figiu'es, so tliat oue cau identify one's plauts easUy 

 and satisfactorily. For one to whom oiiginal types are easUy accessible, it 

 is not difficidt to work with shoii; deseriptions, as one may go to see the 

 types themselves, if one finds original descriptions imavaUable for identifica- 

 tion. But, for one to whom oiiginal types are not accessible, when descrip- 

 tions ai-e too shoii for use, the work cannot but be veiy difficult and 



