DE CANDOLLE'S PHYTOGRAPHY. 287 



This for descriptions, except in local Floras, where pojiular 

 use demands the vernacular ; and we interpose the remark 

 that English botanical language, freely incorporating as it 

 does all Latin and Greek terms, comes next to Latin in con- 

 venience, compactness, and facility to all foreign botanists, 

 who, being familiar with Latin, can seldom be at a loss. For 

 discussions and reasonings, the botanists of each nation prefer 

 their vernacular tongue ; but De CandoUe would restrict them 

 to the four modern languages, one or two of which, beside his 

 native tongue, every naturalist is nowadays supposed to be 

 able fairly to read ; English, German, French, and Italian. 

 Indeed, De CandoUe recommends Latin and the technically 

 descriptive style even for (jcneralia^ on the ground of brevity ; 

 and he aptly suggests that the less capable botanists are of 

 handling other than Linnsean Latin, the more brief, senten- 

 tious, and strictly to the point their exposition will be. 



Hints are given as to the best mode of collecting literary 

 material, making and preserving notes (each upon separate 

 slips of paper) ; upon the importance of adding clear explana- 

 tions of drawings at the time they are made ; and upon the 

 desirability of refraining from publication until the work is 

 thoroughly completed, but of then publishing as soon as pos- 

 sible. A manuscript work is said to have its maximum value 

 at the moment of completion. Our author declares that sec- 

 ond and third editions are seldom equal to the first. That 

 depends. He objects also to posthumous publication, citing 

 Eoxburgh's " Flora Indica," published by Wallich, Plumier's 

 plates, published by Burmann, and the wretched figures of 

 Velloso ; and he might have referred to the ill-advised print- 

 ing of Griffith's rough notes and comments. But all depends 

 upon the character of the manuscript and the length of time 

 which has elapsed. 



Chapters IV-XI traverse the whole subject of descriptions, 

 under various aspects and a rather minute division of tojDics. 

 As even a brief analysis would overpass available space, we 

 wall merely touch here and there upon certain points. 



As to the relation of varieties to species, there are two 

 modes of presentation, both of which have been followed by 



